Home |
Search |
Today's Posts |
#41
|
|||
|
|||
"Phil Allison" wrote
But what were the signs of saturation? You said you observed them, but you haven't said what they are. ** As the voltage increases above a certain threshold the current waveform becomes distorted in a characteristic way. There is firstly increasing 3H - then peaks of both polarities develop in the waveform. After this the rms value of the current rises almost exponentially with increasing applied voltage. No Phil. As I ramarked to Patrick Turner, I know what saturation looks like. What I asked for are the signs of saturation you observed in this case. You are wriggling now in two ways. Firstly, you are avoiding the specific question about your claimed observation. Secondly, you have excluded the word "saturation" from your response. Only the last sentence of your description applies to saturation. If your core is ductile iron, then distortion will begin way before saturation. Even with GOSS there is some distortion before saturation. Why not just admit that your 160VA core, with 40 turns and 5V @ 50 Hz was not showing signs of saturation. There is distortion, and hysteresis, but not saturation. You made that up to look clever, but you failed. You seem to have fooled Patrick Turner though. But he doesn't know about how the source impedance is effectively zero because of your negative feedback. But then I remember neither of you know what feedback is. cheers, Ian |
#42
|
|||
|
|||
"Adam Stouffer" ** No brain but still moves = another hotmail Zombie. ............ Phil |
#43
|
|||
|
|||
Ian Iveson wrote: "Phil Allison" wrote But what were the signs of saturation? You said you observed them, but you haven't said what they are. ** As the voltage increases above a certain threshold the current waveform becomes distorted in a characteristic way. There is firstly increasing 3H - then peaks of both polarities develop in the waveform. After this the rms value of the current rises almost exponentially with increasing applied voltage. No Phil. As I ramarked to Patrick Turner, I know what saturation looks like. What I asked for are the signs of saturation you observed in this case. You are wriggling now in two ways. Firstly, you are avoiding the specific question about your claimed observation. Secondly, you have excluded the word "saturation" from your response. Only the last sentence of your description applies to saturation. If your core is ductile iron, then distortion will begin way before saturation. Even with GOSS there is some distortion before saturation. Why not just admit that your 160VA core, with 40 turns and 5V @ 50 Hz was not showing signs of saturation. There is distortion, and hysteresis, but not saturation. You made that up to look clever, but you failed. With only 5v rms across a 40 turn winding the B max could indeed be up where saturation is beginning. Phil didn't state the Afe for the material but he did say it was a 160 VA toroidal core, and I assume GOSS. But non oriented will also saturate at the same F, but distortion is much more up to saturation. We could guess the AFe from a simple calculation, and guess a near value for the Bmax at 5v, and its probably around 1.3 Tesla. if the onset of saturation was occuring. Phil did explain the manifestation of the onset of saturation, and the resulting distortions. It is a little more complex than what he said, or I said, but I suggest you take a look yourself with a small tranny and a CRO with a dual trace. You seem to have fooled Patrick Turner though. But he doesn't know about how the source impedance is effectively zero because of your negative feedback. WTF has NFB got to do with this issue? I don't think too much of what Phil said would fool anyone. a large % of his posts are just insult. A small % wouldn't fool anyone though. He could do a lot better to can the demon, and describe in more detail, but he's too busy complaining about ppl not doing their research on their own. Being in a group like this **is** the research many folks are doing. He should remember that and maybe you could to. I have assumed the source Z to be low forthe experiments for core losses, regardless of whether NFB is used or not. But then I remember neither of you know what feedback is. Well that makes me think you have not read enough. But then I think you are just taking a shot, and a very poorly aimed one it is. Have a great saturday night. Patrick Turner. cheers, Ian |
#44
|
|||
|
|||
"Ian Iveson" "Phil Allison" But what were the signs of saturation? You said you observed them, but you haven't said what they are. ** As the voltage increases above a certain threshold the current waveform becomes distorted in a characteristic way. There is firstly increasing 3H - then peaks of both polarities develop in the waveform. After this the rms value of the current rises almost exponentially with increasing applied voltage. No Phil. As I ramarked to Patrick Turner, I know what saturation looks like. What I asked for are the signs of saturation you observed in this case. ** Posted by me to PT on the 25th in this thread: " ** Only the 50 Hz test showed any visible distortion on the scope - about 10%, 3H " You are wriggling now in two ways. Firstly, you are avoiding the specific question about your claimed observation. ** See above - you asinine pommy ****head. Why not just admit that your 160VA core, with 40 turns and 5V @ 50 Hz was not showing signs of saturation. ** Because that would be false. There is distortion, and hysteresis, but not saturation. ** If the input is increased to 7 volts rms at 50 Hz - then the current waveform takes on the distinctive peaked appearance of core saturation. By reducing it to 5 volts that effect was visibly almost gone. Hence my comment stands. You and any other pedantic pommy ****s can go to ****ing hell if you don't like it . You made that up to look clever, but you failed. ** The one trying to be too clever by half is YOU - ****head. You seem to have fooled Patrick Turner though. But he doesn't know about how the source impedance is effectively zero because of your negative feedback. But then I remember neither of you know what feedback is. ** The Iveson lunatic is right off his tiny head tonight. ............. Phil |
#45
|
|||
|
|||
"Patrick Turner" After testing a Luxman OPT and 2 other power trannies with a dual trace CRO for phase angle at 50 Hz, and at low levels, all measured in the vicinity of 30 degrees of phase shift, so a high % of the V x I was causing the core to heat up. ** It is utter madness to infer full voltage core loss performance from a low voltage test. When run at rated voltage, both the current waveform and phase angle of a mains tranny change dramatically !! ** PLUS ** the simple: " PF = cos phi " formula *cannot* be used where non sine waves exist. And I have assumed all the input V x I when unloaded was power that contributed to heating the core, but it was a safe wrong assumption to make. In actual fact only about 80% of the V x I is power converted to heat at 50 Hz. ** It is NOTHING like 80% in most cases. EG A 1000VA, 240/120 volt step-down in my workshop draws 0 .74 A rms off load or 178 VA. The peaks of the current waveform coincide exactly in time with AC zero crossings - suggesting a 90 degree phase angle and zero loss !! After many hours off load it is only warm to touch - so the heat loss is no more than 25 watts. The primary R is 0.8 ohms so copper loss is only 0.43 watts. PF is also given by: PF = true watts / VA So in this case the PF = 0.14 - corresponding to an effective phase angle of 82 degrees. One needs a "watt-meter" to measure the power dissipated by an off load mains transformer - then calculate the I squared R loss for the primary and *subtract* it from the watt-meter reading to get total core loss. Another very simple way is to allow about 2 watts per kilo of core for toroidals and about 3.5 for E cores. ............... Phil |
#46
|
|||
|
|||
"Phil Allison" wrote
** If the input is increased to 7 volts rms at 50 Hz - then the current waveform takes on the distinctive peaked appearance of core saturation. Now we know it shows signs of saturation at 7V, not 5V. That is a big difference...a 40% error. By reducing it to 5 volts that effect was visibly almost gone. No. If it "takes on" the the appearance at 7V, then at 5V it has not taken it on. At 5V you saw distortion, not saturation. You seem to have fooled Patrick Turner though. But he doesn't know about how the source impedance is effectively zero because of your negative feedback. But then I remember neither of you know what feedback is. ** The Iveson lunatic is right off his tiny head tonight. And still *neither* of you know what feedback is. cheers, Ian in message ... "Ian Iveson" "Phil Allison" But what were the signs of saturation? You said you observed them, but you haven't said what they are. ** As the voltage increases above a certain threshold the current waveform becomes distorted in a characteristic way. There is firstly increasing 3H - then peaks of both polarities develop in the waveform. After this the rms value of the current rises almost exponentially with increasing applied voltage. No Phil. As I ramarked to Patrick Turner, I know what saturation looks like. What I asked for are the signs of saturation you observed in this case. ** Posted by me to PT on the 25th in this thread: " ** Only the 50 Hz test showed any visible distortion on the scope - about 10%, 3H " You are wriggling now in two ways. Firstly, you are avoiding the specific question about your claimed observation. ** See above - you asinine pommy ****head. Why not just admit that your 160VA core, with 40 turns and 5V @ 50 Hz was not showing signs of saturation. ** Because that would be false. There is distortion, and hysteresis, but not saturation. Hence my comment stands. You and any other pedantic pommy ****s can go to ****ing hell if you don't like it . You made that up to look clever, but you failed. ** The one trying to be too clever by half is YOU - ****head. ............ Phil |
#47
|
|||
|
|||
"Patrick Turner" wrote
I have assumed the source Z to be low forthe experiments for core losses, regardless of whether NFB is used or not. Why assume when you have been told it is true? Phil said he used feedback to reduce the effective source impedance to zero. He said he adjusted the input voltage until the test voltage was 5V in each case. That is voltage-applied, voltage-derived nfb. Why don't you know some of the simplest things? But then I remember neither of you know what feedback is. Well that makes me think you have not read enough. Only because you don't know some of the simplest things. For the rest of your rambling, it is idle speculation about a particular experiment that you have not done. But then I think you are just taking a shot, and a very poorly aimed one it is. Absulutely not. I responded constructively to John's original post, but some ppl are too busy posturing in this misguided thread to remember what the original issue was, and too busy posing to even read each other's posts properly. But perhaps you don't realise that the shape of the BH curve is important when considering core loss. Or that honest reporting of precision is important when posting test results. cheers, Ian |
#48
|
|||
|
|||
Phil Allison wrote: "Patrick Turner" After testing a Luxman OPT and 2 other power trannies with a dual trace CRO for phase angle at 50 Hz, and at low levels, all measured in the vicinity of 30 degrees of phase shift, so a high % of the V x I was causing the core to heat up. ** It is utter madness to infer full voltage core loss performance from a low voltage test. When run at rated voltage, both the current waveform and phase angle of a mains tranny change dramatically !! I didn't make the tests at maximum full voltage levels for either power or OPTs, but I did use levels more akin to actual levels used. ** PLUS ** the simple: " PF = cos phi " formula *cannot* be used where non sine waves exist. Yes, there is some variation in the shift, but it is hard to pin down exactly how much Using the XY scope function to view the elipse to indicate the phase angle gave constant shape elipses at low levels with the power trannies, but the elipse became more like a square at high levels, since the current wave has so much distortion. The only transformer of the 3 which I tested which didn't exhibit serious distortion above a fraction of the max operating voltage was a Luxman OPT primary. The phase angle didn't change much with voltage input. And I have assumed all the input V x I when unloaded was power that contributed to heating the core, but it was a safe wrong assumption to make. In actual fact only about 80% of the V x I is power converted to heat at 50 Hz. ** It is NOTHING like 80% in most cases. Not so. Its obvious that in many power trannies, the power factor is 80% of a rather too high V x I. As I said, I try to limit the VI with no load to no more than 5% of the rated maximum input VA for the transformer. If the power factor is 0.5, not 0.8, or whatever, good, I am happy. The power factor isn't something I really have to know, since my designs are so conservative. They run quiet, and cool, and relaiable, and can take a real over load for hours, and 20% rise in mains voltage won't stuff the tranny. Too much emphasis is being focused on power factor. EG A 1000VA, 240/120 volt step-down in my workshop draws 0 .74 A rms off load or 178 VA. I have 120 mm stacks of 51 mm tongue GOSS material in E&I lams which measured U = 17,000 at the Bmax of about 0.8 Tesla. The cores are about rated for 1.9 Kw, and used for power supplies to provide 550 watts to a substantially class A amplifier. The VI input with no load measures 4 watts. I repeat, 4 watts. No load current is 17 mA. I don't care a hoot about power factors in this case because the trannies I have wound don't have any seriously important amount of core heating. After being on for 4 hrs, these trannies only get about 10C above ambient. I have never bothered to measure the power factor for such transformers. The power factor could be 1.0, and still I wouldn't give a **** about it. The peaks of the current waveform coincide exactly in time with AC zero crossings - suggesting a 90 degree phase angle and zero loss !! In many power trannies, the low level wave form soon becomes distorted, like a limited wave form, then the peaks increase on one side and there seems to be an increasing phase angle as level is increased, and from about 40% of the mains voltage the current wave becomes very distorted with peaking 3H, and exactly what the power factor is is hard to estimate. So I am not wrong in saying lets keep the VI to the tranny when not loaded to less than 5% of the working VA. Or put another way, when using good material like I do, run the stuff at no higher than 0.85 tesla, about where the primary inductance will be at its maximum, and VI unloaded will be very small, and whatever the heating losses are they will be small After many hours off load it is only warm to touch - so the heat loss is no more than 25 watts. Fair enough, you have a reasonable transformer. Alas I have lots of gear with trannies running at 40C above ambient. The primary R is 0.8 ohms so copper loss is only 0.43 watts. PF is also given by: PF = true watts / VA So in this case the PF = 0.14 - corresponding to an effective phase angle of 82 degrees. One needs a "watt-meter" to measure the power dissipated by an off load mains transformer - then calculate the I squared R loss for the primary and *subtract* it from the watt-meter reading to get total core loss. I don't have a watt meter. Somehow I manage to get cooler and quieter trannies than all the gear including high end stuff from Conrad Johnson. Another very simple way is to allow about 2 watts per kilo of core for toroidals and about 3.5 for E cores. You could do that, but I prefer to build extremly low stressed components. In a core weighing 16 Kgs, I have only 4 watts of VI, and I don't know the PF is, I have never bothered to measure it, or what the phase angle is. I don't know what % of the 4 watts is heating the core. I don't care about it, it works, and there is SFA core heating. The regulation is fine for a tube class AB1 amplifier where the first 100 watts is class A and the watts up to 350 max in AB1 do not result in any more than a couple of volts of change with music signals using rock and roll when the peaks in the somewhat low dynamic range music are just beginning to clip. Sure there is a slight drop in B+ with asine wave or square wave test. Its inconsequential. In the 300 watt rated OPTs, the core losses in the OPTs are also utterly inconsequential. I recall measuring 600H of inductance at about 600vrms a-a at 50 Hz.. This voltage makes 300 watts into 1.2 kohms. Unloaded current at 600v and 50 Hz is around 4 mA. The core losses are entirely negligible. Anyone can pillory me for not being very aware of power factors and cos ph angles, but really they are just charging at windmills. Patrick Turner. .............. Phil |
#49
|
|||
|
|||
Ian Iveson wrote: "Phil Allison" wrote ** If the input is increased to 7 volts rms at 50 Hz - then the current waveform takes on the distinctive peaked appearance of core saturation. Now we know it shows signs of saturation at 7V, not 5V. That is a big difference...a 40% error. By reducing it to 5 volts that effect was visibly almost gone. No. If it "takes on" the the appearance at 7V, then at 5V it has not taken it on. At 5V you saw distortion, not saturation. Methinks you are being a trifle picky Ian. Much plain non oriented Si Fe materials show a gradual onset of saturation. Anyway, I know the sort of thing Phil has described; he's right about his descriptions. You seem to have fooled Patrick Turner though. But he doesn't know about how the source impedance is effectively zero because of your negative feedback. But then I remember neither of you know what feedback is. ** The Iveson lunatic is right off his tiny head tonight. And still *neither* of you know what feedback is. Well, we sure are getting feedback from you in this discussion. I mean that in the general sense of discussion FB rather than the electronic type. Phil has called you a lunatic for your feedback, but I would rather not call you anything. I could say its rash to be so fast to tell us what we don't know.... But I won't say you are rash.... Patrick Turner. cheers, Ian in message ... "Ian Iveson" "Phil Allison" But what were the signs of saturation? You said you observed them, but you haven't said what they are. ** As the voltage increases above a certain threshold the current waveform becomes distorted in a characteristic way. There is firstly increasing 3H - then peaks of both polarities develop in the waveform. After this the rms value of the current rises almost exponentially with increasing applied voltage. No Phil. As I ramarked to Patrick Turner, I know what saturation looks like. What I asked for are the signs of saturation you observed in this case. ** Posted by me to PT on the 25th in this thread: " ** Only the 50 Hz test showed any visible distortion on the scope - about 10%, 3H " You are wriggling now in two ways. Firstly, you are avoiding the specific question about your claimed observation. ** See above - you asinine pommy ****head. Why not just admit that your 160VA core, with 40 turns and 5V @ 50 Hz was not showing signs of saturation. ** Because that would be false. There is distortion, and hysteresis, but not saturation. Hence my comment stands. You and any other pedantic pommy ****s can go to ****ing hell if you don't like it . You made that up to look clever, but you failed. ** The one trying to be too clever by half is YOU - ****head. ............ Phil |
#50
|
|||
|
|||
Ian Iveson wrote: "Patrick Turner" wrote I have assumed the source Z to be low forthe experiments for core losses, regardless of whether NFB is used or not. Why assume when you have been told it is true? Phil said he used feedback to reduce the effective source impedance to zero. So? He said he adjusted the input voltage until the test voltage was 5V in each case. That is voltage-applied, voltage-derived nfb. So? Low impedance is low impedance. The mains offeres low impedance from the wal socket, mainly because there are thousands of other folks connected making a low load resistance. But any amplifer capable of 20vrms output into 8 ohms, and with Ro = 1 ohm, and with 100 kHz of bw will do. Many SS amps exceed the capability need for Phil's tests. Why don't you know some of the simplest things? such as? But then I remember neither of you know what feedback is. Well that makes me think you have not read enough. Only because you don't know some of the simplest things. such as? For the rest of your rambling, it is idle speculation about a particular experiment that you have not done. Me rambling? c'mon, your'e the one to ramble on about Phil and I not understanding what FB is. Phil reckons you are insane. Don't let it worry you. But then I think you are just taking a shot, and a very poorly aimed one it is. Absulutely not. I responded constructively to John's original post, but some ppl are too busy posturing in this misguided thread to remember what the original issue was, and too busy posing to even read each other's posts properly. Finally, overall I got the message about power factor and phase angles. It turns out that heat losses and powe factors and phase angles are of little practical value in my everyday practice of tube craft because the core losses in anything I wind are so low. The original issue was core losses. But perhaps you don't realise that the shape of the BH curve is important when considering core loss. Or that honest reporting of precision is important when posting test results. The BH curves do affect core losses, and I'd worry more about them if they really had a large effect on what I was trying to make. But they damn well don't. We are stuck with whatever curves there are. We know we have to use turns. OK, select turns and core size to get cool quiet trannies. Once achieved, there is no need for me to feel guilty for not comprehending power factors deeply enough to impress the tranny cognescenti who don't seem to be winding much of anything as we speak. After this discussion I know more, but it won't change the design methods I use. I don't have a been counter on my staff that I have to feed. Patrick Turner. cheers, Ian |
#51
|
|||
|
|||
"Patrick Turner" wrote
I have assumed the source Z to be low forthe experiments for core losses, regardless of whether NFB is used or not. Why assume when you have been told it is true? Phil said he used feedback to reduce the effective source impedance to zero. So? So... He said he adjusted the input voltage until the test voltage was 5V in each case. That is voltage-applied, voltage-derived nfb. So? So... Low impedance is low impedance. The mains offeres low impedance from the wal socket, mainly because there are thousands of other folks connected making a low load resistance. But any amplifer capable of 20vrms output into 8 ohms, and with Ro = 1 ohm, and with 100 kHz of bw will do. Many SS amps exceed the capability need for Phil's tests. ....so you still haven't got the message. The open loop impedance of the source, including the test resistance, *does not matter*. As long it can produce enough current for the test (not much) and maintain a reasonable sine wave. With the application of feedback, the source impedance is effectively *zero*. i.e. 0, as in none. The output of the driver doesn't have to be 1 ohm, or 20V output into 8 ohms. It doesn't need 100kHz of bandwidth either. In fact it needs absolutely *none* of the things you say it needs. That is because it is in a *closed loop* and, because of the *-nfb* its output impedance doen't matter. No, not at all. Not one bit. ....so you can stop babbling on about what you think you might know about it. Why don't you know some of the simplest things? such as? What feedback is, what it does, what PF is, and why it is significant, and trigonometry and complex numbers and transient response and... It is easier to list what you *do* know. Although you may not be able to understand it, I do give you credit for knowing what is written in RDH4. ...After this discussion I know more, but it won't change the design methods I use. If you really knew more, it *would* change the design methods you use. Perhaps after a while this new knowledge might find a slot to fit into, and it will seem then like a revelation rather than a mere detail waiting to be forgotten. ...OK, select turns and core size to get cool quiet trannies. Once achieved, there is no need for me to feel guilty for not comprehending power factors deeply enough to impress the tranny cognescenti who don't seem to be winding much of anything as we speak. I don't have a been counter on my staff that I have to feed. Quite right. If you only build brick ****houses, you don't need to know much about bricks. And if you build brick ****houses for those with more money than sense, you don't need an accountant. cheers, Ian |
#52
|
|||
|
|||
"Ian Iveson" Why assume when you have been told it is true? Phil said he used feedback to reduce the effective source impedance to zero. ** No he bloody did not !!!!!!!! He said he adjusted the input voltage until the test voltage was 5V in each case. ** Right - to compensate for HF roll off in the SS amp. Otherwise the level would start to fall above 25 kHz. That is voltage-applied, voltage-derived nfb. ** ROTFLMAO !!!!!!!! Never considered my left hand on a level control to be NFB before !!!! I reckon it is more like feed forward error correction. You see, I was eating a peanut butter sandwich at the time. ................ Phil |
#53
|
|||
|
|||
"Patrick Turner" Phil Allison wrote: After testing a Luxman OPT and 2 other power trannies with a dual trace CRO for phase angle at 50 Hz, and at low levels, all measured in the vicinity of 30 degrees of phase shift, so a high % of the V x I was causing the core to heat up. ** It is utter madness to infer full voltage core loss performance from a low voltage test. When run at rated voltage, both the current waveform and phase angle of a mains trannie change dramatically !! I didn't make the tests at maximum full voltage levels for either power or OPTs, but I did use levels more akin to actual levels used. ** Pure madness - power trannies operate at a fixed voltage - 240 AC / 50 Hz in Aussie. Your low voltage tests on them were utterly WORTHLESS !!!!!!!!! ** PLUS ** the simple: " PF = cos phi " formula *cannot* be used where non sine waves exist. And I have assumed all the input V x I when unloaded was power that contributed to heating the core, but it was a safe wrong assumption to make. In actual fact only about 80% of the V x I is power converted to heat at 50 Hz. ** It is NOTHING like 80% in most cases. Not so. ** The Turneroid ****wit is not paying attention. Its obvious .... ** From doing the WRONG tests.......... that in many power trannies, the power factor is 80% of a rather too high V x I. ** More utter Turneroid BULL**** follows. I have never bothered to measure the power factor for such transformers. The power factor could be 1.0, and still I wouldn't give a **** about it. ** Pig ignorant ARSEHOLE !!!!! One needs a "watt-meter" to measure the power dissipated by an off load mains transformer - then calculate the I squared R loss for the primary and *subtract* it from the watt-meter reading to get total core loss. I don't have a watt meter. ** Nor I - so I use estimates from the current wave shape and experience. The PF for commercial mains trannies ,when off load, is only 0.1 to 0.3 No ****ing way is it 0.8 !!!! Anyone can pillory me for not being very aware of power factors and cos ph angles, but really they are just charging at windmills. ** More like ****ing on a pile of broken bricks. ................ Phil |
#54
|
|||
|
|||
Ian Iveson wrote: "Patrick Turner" wrote I have assumed the source Z to be low forthe experiments for core losses, regardless of whether NFB is used or not. Why assume when you have been told it is true? Phil said he used feedback to reduce the effective source impedance to zero. So? So... He said he adjusted the input voltage until the test voltage was 5V in each case. That is voltage-applied, voltage-derived nfb. So? So... Low impedance is low impedance. The mains offeres low impedance from the wal socket, mainly because there are thousands of other folks connected making a low load resistance. But any amplifer capable of 20vrms output into 8 ohms, and with Ro = 1 ohm, and with 100 kHz of bw will do. Many SS amps exceed the capability need for Phil's tests. ...so you still haven't got the message. The open loop impedance of the source, including the test resistance, *does not matter*. As long it can produce enough current for the test (not much) and maintain a reasonable sine wave. With the application of feedback, the source impedance is effectively *zero*. i.e. 0, as in none. The output of the driver doesn't have to be 1 ohm, or 20V output into 8 ohms. It doesn't need 100kHz of bandwidth either. In fact it needs absolutely *none* of the things you say it needs. That is because it is in a *closed loop* and, because of the *-nfb* its output impedance doen't matter. No, not at all. Not one bit. ...so you can stop babbling on about what you think you might know about it. I don't think you have a clue awhat you are talking about. Why don't you know some of the simplest things? such as? What feedback is, what it does, what PF is, and why it is significant, and trigonometry and complex numbers and transient response and... Utter BS. It is easier to list what you *do* know. Although you may not be able to understand it, I do give you credit for knowing what is written in RDH4. No need to discuss matters further with *you*. ...After this discussion I know more, but it won't change the design methods I use. If you really knew more, it *would* change the design methods you use. Perhaps after a while this new knowledge might find a slot to fit into, and it will seem then like a revelation rather than a mere detail waiting to be forgotten. You are bull****ting on and on for the sake of wasting all our time we could be focusing on someone with something to say. You are saying nothing. ...OK, select turns and core size to get cool quiet trannies. Once achieved, there is no need for me to feel guilty for not comprehending power factors deeply enough to impress the tranny cognescenti who don't seem to be winding much of anything as we speak. I don't have a been counter on my staff that I have to feed. Quite right. If you only build brick ****houses, you don't need to know much about bricks. And if you build brick ****houses for those with more money than sense, you don't need an accountant. And you sure don't like the fact that I am sure of myself, my methods, and that they are fine products, and you meanwhile make SFA to offer anyone. You are not even offering any ideas which can be utilised favourably in the design process. You denigrate with no justification behind it. It makes you a pathetic pomme windbag. Patrick Turner. cheers, Ian |
#55
|
|||
|
|||
Phil Allison wrote: "Patrick Turner" Phil Allison wrote: After testing a Luxman OPT and 2 other power trannies with a dual trace CRO for phase angle at 50 Hz, and at low levels, all measured in the vicinity of 30 degrees of phase shift, so a high % of the V x I was causing the core to heat up. ** It is utter madness to infer full voltage core loss performance from a low voltage test. When run at rated voltage, both the current waveform and phase angle of a mains trannie change dramatically !! I didn't make the tests at maximum full voltage levels for either power or OPTs, but I did use levels more akin to actual levels used. ** Pure madness - power trannies operate at a fixed voltage - 240 AC / 50 Hz in Aussie. Your low voltage tests on them were utterly WORTHLESS !!!!!!!!! Not for the purposes of examining the current phase shifts at low levels. ** PLUS ** the simple: " PF = cos phi " formula *cannot* be used where non sine waves exist. And I have assumed all the input V x I when unloaded was power that contributed to heating the core, but it was a safe wrong assumption to make. In actual fact only about 80% of the V x I is power converted to heat at 50 Hz. ** It is NOTHING like 80% in most cases. Not so. ** The Turneroid ****wit is not paying attention. Its obvious .... ** From doing the WRONG tests.......... that in many power trannies, the power factor is 80% of a rather too high V x I. ** More utter Turneroid BULL**** follows. I have never bothered to measure the power factor for such transformers. The power factor could be 1.0, and still I wouldn't give a **** about it. ** Pig ignorant ARSEHOLE !!!!! I don't need to worry about phase angles or power factors in the design process in anything I wind. I have come to learn more about all this during the discussion, but I won't be changing my design methods. One needs a "watt-meter" to measure the power dissipated by an off load mains transformer - then calculate the I squared R loss for the primary and *subtract* it from the watt-meter reading to get total core loss. I don't have a watt meter. ** Nor I - so I use estimates from the current wave shape and experience. The PF for commercial mains trannies ,when off load, is only 0.1 to 0.3 No ****ing way is it 0.8 !!!! Anyone can pillory me for not being very aware of power factors and cos ph angles, but really they are just charging at windmills. ** More like ****ing on a pile of broken bricks. ............... Phil I found that to get quiet trannies, especially when feeding a rectifier supply in a class A amp where the rectifier power is the majority power component, I set up the primary to run at about 0.85 tesla max. Usually this makes the VI input a lot lower than all the noisy hot commercial trannies I have come to loathe. The cores also run as cool as a cucumber. I have never bothered to worry about PF because the core heating power is always going to be less than the VI input with the tranny unloaded. But why worry anyway? Using the GOSS I use gives me very decent transformers. I had to wind a new power tranny for an EAR 509 to match the good one in another mono amp. So I was forced to use the size of core chosen by Paravicini. But the one I wound ran cooler and quieter than the existing one. Criticise my methods if you want, but I do know enough to get by real well. Your'e the one with the broken lance from charging at windmills. I think your horse is also a bit lame. Patrick Turner. |
#56
|
|||
|
|||
"Patrick Turner" Phil Allison ** The Turneroid ****wit is not paying attention. Its obvious .... ** From doing the WRONG tests.......... that in many power trannies, the power factor is 80% of a rather too high V x I. ** More utter Turneroid BULL**** follows. I have never bothered to measure the power factor for such transformers. The power factor could be 1.0, and still I wouldn't give a **** about it. ** Pig ignorant ARSEHOLE !!!!! I don't need to worry about phase angles or power factors in the design process in anything I wind. ** Then kindly do not ****ing continue misinforming the whole NG about PF. Or I will have to tell everyone what a criminal **** you are. .................. Phil |
#57
|
|||
|
|||
Hi Phil,
I put your numbers into a spreadsheet and calculated the equivalent R that would dissipate the power shown in the last column of your measurements. I'm assuming that the core losses in the inductor can be modelled as an R in parallel with an L. The R is the real part of the impedance which I have labeled Rl(Z). Then using phi I calculated the imaginary part, i.e., XL, which I labeled Im(Z). Finally from Im(Z) I calculated L and saw that it decreased quite a bit with increasing frequency. Is that correct? Joe Pcore[W] Rl(Z) Im(Z) L 4.05E-01 40.5 29.4 9.4E-02 3.48E-01 48.2 33.7 5.4E-02 2.38E-01 70.6 49.4 2.6E-02 1.84E-01 104.1 57.7 9.2E-03 1.54E-01 133.3 62.1 3.3E-03 1.25E-01 171.7 69.4 1.1E-03 9.67E-02 219.2 93.0 4.9E-04 7.77E-02 268.7 119.6 1.9E-04 |
#58
|
|||
|
|||
"Joseph Meditz" Hi Phil, I put your numbers into a spreadsheet and calculated the equivalent R that would dissipate the power shown in the last column of your measurements. I'm assuming that the core losses in the inductor can be modelled as an R in parallel with an L. The R is the real part of the impedance which I have labeled Rl(Z). Then using phi I calculated the imaginary part, i.e., XL, which I labeled Im(Z). Finally from Im(Z) I calculated L and saw that it decreased quite a bit with increasing frequency. Is that correct? Joe Pcore[W] Rl(Z) Im(Z) L 4.05E-01 40.5 29.4 9.4E-02 3.48E-01 48.2 33.7 5.4E-02 2.38E-01 70.6 49.4 2.6E-02 1.84E-01 104.1 57.7 9.2E-03 1.54E-01 133.3 62.1 3.3E-03 1.25E-01 171.7 69.4 1.1E-03 9.67E-02 219.2 93.0 4.9E-04 7.77E-02 268.7 119.6 1.9E-04 ** You have gone haywire somewhere with your calcs. EG: For the 50Hz case, I = 100mA, P = 400mW and V = 5 volts. R must be a resistance that dissipates 400mW at 5 volts. That R = 5 x 5 / 0.4 = 62.5 ohms So the current through R = 5/62.5 = 80mA. The L component in parallel increases the current flow to 100mA and shifts the phase angle by 36 degrees. Yep - the L component acts *weird* and changes to follow the frequency. Specifying the equivalent parallel L of an iron transformer's primary means also specifying the frequency and the voltage level. ............. Phil |
#59
|
|||
|
|||
"Phil Allison" wrote
** Right - to compensate for HF roll off in the SS amp. Otherwise the level would start to fall above 25 kHz. Yup, that's what voltage feedback does, amongst other things. That's why Patrick was wrong to say you needed an amp with 100kHz of bandwidth. ** ROTFLMAO !!!!!!!! Never considered my left hand on a level control to be NFB before !!!! Glad to make a contribution to your understanding of feedback. Not just your hand, but your eyes and brain. You sense a voltage difference (hence voltage-derived), you adjust the input voltage until the test voltage is correct again (voltage-applied). Your time period is long, so you can only respond to an aggregate of the output, but that's OK coz the signal is simply repetitive. OTOH, you can't null the voltage distortion (unless you can watch and wiggle your fingers at 100kHz), which is a weakness of the method. You compensated for *low* f voltage drop across the test piece also, thus eliminating the output impedance of the combination of amp and sense resistor as the current through the test piece rose. Wondering if these two threads have been worthwhile so far. Two people have done interesting tests and posted data, with the net result of helping Patrick Turner to understand PF and making strange Gerald grumpy. John's data is not accurate but we don't know how imprecise. Your data is derived from reading peak V rather than true rms and may also be inaccurate. We know about John's core but we don't know what yours is made of. John's is an EI whereas yours is a torroid. John's has a primary winding and yours doesn't. John used a high open-loop source resistance but you used a much lower value leading to much less (but still maybe significant) voltage distortion. I have said that I doubt the primary winding is the only reason why John's trend differs from yours at HF, but we don't really know, and I thought that was what you were trying to prove. Do you think you have done so? It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. I can see that it would cause a phase shift at HF because of Cp, but how would this translate into a loss that would be revealed by your test method? Can you estimate what the theoretical difference would be, for comparison? Perhaps you could wind a primary and re-test :-) cheers, Ian |
#60
|
|||
|
|||
"Ian Iveson" "Phil Allison" ** Right - to compensate for HF roll off in the SS amp. Otherwise the level would start to fall above 25 kHz. Yup, that's what voltage feedback does, amongst other things. ** **** off pommy ****wit. ** ROTFLMAO !!!!!!!! Never considered my left hand on a level control to be NFB before !!!! Glad to make a contribution to your understanding of feedback. ** **** off pommy ****wit. Not just your hand, but your eyes and brain. You sense a voltage difference (hence voltage-derived), you adjust the input voltage until the test voltage is correct again (voltage-applied). ** Nothing to do with NFB - ****head. Your time period is long, so you can only respond to an aggregate of the output, but that's OK coz the signal is simply repetitive. OTOH, you can't null the voltage distortion (unless you can watch and wiggle your fingers at 100kHz), which is a weakness of the method. ** **** off pommy ****wit. Wondering if these two threads have been worthwhile so far. Two people have done interesting tests and posted data, with the net result of helping Patrick Turner to understand PF and making strange Gerald grumpy. ** **** off pommy ****wit. John's data is not accurate but we don't know how imprecise. Your data is derived from reading peak V rather than true rms and may also be inaccurate. ** Huh ????? Where did you get that idiot idea ?? Plucked straight out of you fat stupid arse like all the others. We know about John's core but we don't know what yours is made of. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. John's is an EI whereas yours is a torroid. John's has a primary winding and yours doesn't. John used a high open-loop source resistance but you used a much lower value leading to much less (but still maybe significant) voltage distortion. ** There was no visible distortion above 50 Hz . All voltages were measured with a 4.5 digit true rms meter. I have said that I doubt the primary winding is the only reason why John's trend differs from yours at HF, ** **** off pommy ****wit. It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. ** Posted that already. Perhaps you could wind a primary and re-test :-) ** No need for that. I also tested a similar, complete toroidal mains transformer - the results were just like the results with valve OTs. There was a loss minima and zero phase angle around 1kHz - then a capacitive phase angle and rising HF losses above that. The loading effect of the primary is the culprit all right. Prove otherwise if you can - arsehole. BTW Learn to spell "toroid" some day - if that is not to torrid for a pommy turd like you. .............. Phil |
#61
|
|||
|
|||
** Pig ignorant ARSEHOLE !!!!! I don't need to worry about phase angles or power factors in the design process in anything I wind. ** Then kindly do not ****ing continue misinforming the whole NG about PF. Or I will have to tell everyone what a criminal **** you are. ................. Phil Be my guest to call me any kind of entity you deem appropriate. This will make utterly no difference to when/where/how/if I may respond in a news group. The discussion on PF has taught me I have nothing to worry about re my present design methods and standards of construction of all power and output transformers and iron cored chokes. Threats of calling me names will only have one result; the group will continue to think you are possessed by a demon. Your antics will never destroy us. Patrick Turner. |
#62
|
|||
|
|||
"Phil Allison" wrote
Not just your hand, but your eyes and brain. You sense a voltage difference (hence voltage-derived), you adjust the input voltage until the test voltage is correct again (voltage-applied). ** Nothing to do with NFB - ****head. Perhaps if you sleep on it. Read it again in the morning when you're feeling less vulnerable and defensive. John's data is not accurate but we don't know how imprecise. Your data is derived from reading peak V rather than true rms and may also be inaccurate. ** Huh ????? Where did you get that idiot idea ?? Plucked straight out of you fat stupid arse like all the others. You said you used an oscilloscope to monitor the voltage and current waveforms. Perhaps it has a dual rms measuring or conversion facility...you didn't say so it is reasonable to assume you used your eyes to measure the peak, and then converted to rms. Unless you say how else you measured it. I have a nice arse, BTW. Not too fat, not too thin, and *very* clever. We know about John's core but we don't know what yours is made of. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. You didn't say, and no they are not...not by a very long way, silly. ** There was no visible distortion above 50 Hz . How many traces have you got? If a dual trace then you could be comparing a distorted current to a distorted voltage. You don't say whether the visible distortion at 50Hz was of voltage or current. What percentage is visible anyway? All voltages were measured with a 4.5 digit true rms meter. Now you say... It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. ** Posted that already. Where? I just got some handwaving about "tuned circuit". Have you said where the power is lost? Winding or core, or both? In what proportion? Which winding? The loading effect of the primary is the culprit all right. Prove otherwise if you can - arsehole. You're the one who is making the assertion. You haven't proven anything so far. I also tested a similar, complete toroidal mains transformer - the results were just like the results with valve OTs. There was a loss minima and zero phase angle around 1kHz - then a capacitive phase angle and rising HF losses above that. You've made this up. What angle, in what direction, and how much loss? If you'd really done it you would have jubilantly reported the results. That would be useful because it is an experiment that many could repeat. If the maths works it would also clinch your case. And if you can explain *where* the power is lost, you will have made some progress. Learn to spell "toroid" some day - if that is not to torrid for a pommy turd like you. Today will be fine thanks. Wasn't in my MS dictionary. At least it wasn't a grammatical error: us English hate those. The singular of "minima" is "minimum", BTW. cheers, Ian |
#63
|
|||
|
|||
Phil Allison wrote: "Ian Iveson" "Phil Allison" ** Right - to compensate for HF roll off in the SS amp. Otherwise the level would start to fall above 25 kHz. Yup, that's what voltage feedback does, amongst other things. ** **** off pommy ****wit. ** ROTFLMAO !!!!!!!! Never considered my left hand on a level control to be NFB before !!!! Glad to make a contribution to your understanding of feedback. ** **** off pommy ****wit. Not just your hand, but your eyes and brain. You sense a voltage difference (hence voltage-derived), you adjust the input voltage until the test voltage is correct again (voltage-applied). ** Nothing to do with NFB - ****head. Your time period is long, so you can only respond to an aggregate of the output, but that's OK coz the signal is simply repetitive. OTOH, you can't null the voltage distortion (unless you can watch and wiggle your fingers at 100kHz), which is a weakness of the method. ** **** off pommy ****wit. Wondering if these two threads have been worthwhile so far. Two people have done interesting tests and posted data, with the net result of helping Patrick Turner to understand PF and making strange Gerald grumpy. ** **** off pommy ****wit. John's data is not accurate but we don't know how imprecise. Your data is derived from reading peak V rather than true rms and may also be inaccurate. ** Huh ????? Where did you get that idiot idea ?? Plucked straight out of you fat stupid arse like all the others. We know about John's core but we don't know what yours is made of. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. John's is an EI whereas yours is a torroid. John's has a primary winding and yours doesn't. John used a high open-loop source resistance but you used a much lower value leading to much less (but still maybe significant) voltage distortion. ** There was no visible distortion above 50 Hz . All voltages were measured with a 4.5 digit true rms meter. I have said that I doubt the primary winding is the only reason why John's trend differs from yours at HF, ** **** off pommy ****wit. It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. ** Posted that already. Perhaps you could wind a primary and re-test :-) ** No need for that. I also tested a similar, complete toroidal mains transformer - the results were just like the results with valve OTs. There was a loss minima and zero phase angle around 1kHz - then a capacitive phase angle and rising HF losses above that. The loading effect of the primary is the culprit all right. Prove otherwise if you can - arsehole. BTW Learn to spell "toroid" some day - if that is not to torrid for a pommy turd like you. ............. Phil Regardless of difficulties with pomme torrid toroidal turds, perhaps you could give us the strip width x build up dimension of the toroid you used for your tests so we could work out for ourselves what the Bmax would be at the F and V you used for testing. The Bmax may explain whether the core was/wasn't near saturation. Patrick Turner. |
#64
|
|||
|
|||
Ian Iveson wrote: "Phil Allison" wrote Not just your hand, but your eyes and brain. You sense a voltage difference (hence voltage-derived), you adjust the input voltage until the test voltage is correct again (voltage-applied). ** Nothing to do with NFB - ****head. Perhaps if you sleep on it. Read it again in the morning when you're feeling less vulnerable and defensive. John's data is not accurate but we don't know how imprecise. Your data is derived from reading peak V rather than true rms and may also be inaccurate. ** Huh ????? Where did you get that idiot idea ?? Plucked straight out of you fat stupid arse like all the others. You said you used an oscilloscope to monitor the voltage and current waveforms. Perhaps it has a dual rms measuring or conversion facility...you didn't say so it is reasonable to assume you used your eyes to measure the peak, and then converted to rms. Unless you say how else you measured it. I have a nice arse, BTW. Not too fat, not too thin, and *very* clever. We know about John's core but we don't know what yours is made of. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. You didn't say, and no they are not...not by a very long way, silly. ** There was no visible distortion above 50 Hz . How many traces have you got? If a dual trace then you could be comparing a distorted current to a distorted voltage. You don't say whether the visible distortion at 50Hz was of voltage or current. What percentage is visible anyway? All voltages were measured with a 4.5 digit true rms meter. Now you say... It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. ** Posted that already. Where? I just got some handwaving about "tuned circuit". Have you said where the power is lost? Winding or core, or both? In what proportion? Which winding? The loading effect of the primary is the culprit all right. Prove otherwise if you can - arsehole. You're the one who is making the assertion. You haven't proven anything so far. I also tested a similar, complete toroidal mains transformer - the results were just like the results with valve OTs. There was a loss minima and zero phase angle around 1kHz - then a capacitive phase angle and rising HF losses above that. You've made this up. What angle, in what direction, and how much loss? If you'd really done it you would have jubilantly reported the results. That would be useful because it is an experiment that many could repeat. If the maths works it would also clinch your case. And if you can explain *where* the power is lost, you will have made some progress. Learn to spell "toroid" some day - if that is not to torrid for a pommy turd like you. Today will be fine thanks. Wasn't in my MS dictionary. At least it wasn't a grammatical error: us English hate those. The singular of "minima" is "minimum", BTW. Youse really ought to learn to read proper Ian and take a look at the CRO with a tranny hooked up and duplicate Phil and John's tests and thus not have to ask so many questions most of which Phil has ansered satisfactorillily and it would save us from so much swearin and a cursin and my granpa said if youse don't look youse won't know Patrick Turner. cheers, Ian |
#65
|
|||
|
|||
"Ian Iveson"
"Phil Allison" Not just your hand, but your eyes and brain. You sense a voltage difference (hence voltage-derived), you adjust the input voltage until the test voltage is correct again (voltage-applied). ** Nothing to do with NFB - ****head. Perhaps if you sleep on it. ** Perhaps if you took your hand off it for one second, ******. John's data is not accurate but we don't know how imprecise. Your data is derived from reading peak V rather than true rms and may also be inaccurate. ** Huh ????? Where did you get that idiot idea ?? Plucked straight out of you fat stupid arse like all the others. You said you used an oscilloscope to monitor the voltage and current waveforms. Perhaps it has a dual rms measuring or conversion facility...you didn't say so it is reasonable to assume you used your eyes to measure the peak, and then converted to rms. ** Perfectly valid method to get rms value when the waves are not distorted. We know about John's core but we don't know what yours is made of. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. You didn't say, and no they are not...not by a very long way, silly. ** OK you pig arrogant pommy arsehole - point us all to all the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. The onus is on YOU to prove your mad assertions.!!! ** There was no visible distortion above 50 Hz . How many traces have you got? If a dual trace then you could be comparing a distorted current to a distorted voltage. You don't say whether the visible distortion at 50Hz was of voltage or current. ** Hey ****wit - a SS amp is not gonna distort driving a 50 ohm load at 50 Hz. What percentage is visible anyway? ** The threshold of visible 3H on a decent analogue scope is about 2 % . All voltages were measured with a 4.5 digit true rms meter. Now you say... ** **** off - pommy ****. It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. ** Posted that already. Where? I just got some handwaving about "tuned circuit". ** That is just what it is. Have you said where the power is lost? Winding or core, or both? In what proportion? Which winding? ** Not part of the original question about CORE losses. In any case - when power is being transferred to a secondary, core magnetisation ( and hence loss) is reduced. The loading effect of the primary is the culprit all right. Prove otherwise if you can - arsehole. You're the one who is making the assertion. You haven't proven anything so far. ** The figures I posted say otherwise and are consistent with theory. Nothing will convince a mad pommy **** like you Ian. I also tested a similar, complete toroidal mains transformer - the results were just like the results with valve OTs. There was a loss minima and zero phase angle around 1kHz - then a capacitive phase angle and rising HF losses above that. You've made this up. ** Get ****ed you evil pommy ****. What angle, in what direction, and how much loss? ** Learn to read - ****head. If you'd really done it you would have jubilantly reported the results. ** Rot - they only give the same picture that has been posted already by JS. That would be useful because it is an experiment that many could repeat. ** Anyone can drive a transformer with 5 volts and measure the current and relative phase. Time for YOU to do a bit of work - arsehole. Learn to spell "toroid" some day - if that is not to torrid for a pommy turd like you. Today will be fine thanks. Wasn't in my MS dictionary. ** ****ing ******. At least it wasn't a grammatical error: us English hate those. The singular of "minima" is "minimum", BTW. ** There was no grammar error - you posturing ass. " a loss minima " = one of the loss minima. Don't forget get this: " OK you pig arrogant pommy arsehole - point us all to al the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. " The onus is on you to prove your MAD assertions. .............. Phil |
#66
|
|||
|
|||
"Patrick Turner" = congenital criminal LIAR perhaps you could give us the strip width x build up dimension of the toroid you used for your tests so we could work out for ourselves what the Bmax would be at the F and V you used for testing. ** A common 160 VA core, 31 x 20 mm ( 620 sq mm ) The Bmax may explain whether the core was/wasn't near saturation. ** It was no where near saturation from 100 Hz up. The test was of HIGH FREQUENCY core losses - you ASS !!!!!!!! Now **** the hell off and stop posting dumb lies that mislead everyone - you ****. ................ Phil |
#67
|
|||
|
|||
"Phil Allison" wrote Where did you get that idiot idea ?? Plucked straight out of you fat stupid arse like all the others. You said you used an oscilloscope to monitor the voltage and current waveforms. Perhaps it has a dual rms measuring or conversion facility...you didn't say so it is reasonable to assume you used your eyes to measure the peak, and then converted to rms. ** Perfectly valid method to get rms value when the waves are not distorted. So why say it's an "idiot idea"? I didn't criticise it, beyond pointing out that distortion makes it an approximation. Distortion is moot here, of course, so we need to be careful. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. You didn't say, and no they are not...not by a very long way, silly. ** OK you pig arrogant pommy arsehole - point us all to all the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. Twisting and turning in the wind again. You said *all toroidal transformers*. You mean "sell", BTW. Spell checker won't often help with grammar. The onus is on YOU to prove your mad assertions.!!! To who? You are probably the only one who thinks *all toroidal transformers* are GOSS. You didn't say it is a core intended for a mains transformer either, although you did say it is "standard". Perhaps some ppl know that "standard" means "GOSS", but I didn't, and still don't. Are you really sure of this? Perhaps I should take your word? ** There was no visible distortion above 50 Hz . How many traces have you got? If a dual trace then you could be comparing a distorted current to a distorted voltage. You don't say whether the visible distortion at 50Hz was of voltage or current. ** Hey ****wit - a SS amp is not gonna distort driving a 50 ohm load at 50 Hz. Perhaps not, but the drive to the test piece will distort as a result of the current sensing resistor. You say you observed distortion, which I assume was to the current waveform. That means there is *definitely* voltage distortion too, again because of the sense resistor and also because of the output resistance of the amp, which is significant wrt distortion in your set up. What percentage is visible anyway? ** The threshold of visible 3H on a decent analogue scope is about 2 % . OK. Assuming it is 3H, but I won't quibble if you can't analyse it further. It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. ** Posted that already. Where? I just got some handwaving about "tuned circuit". ** That is just what it is. But that in itself doesn't explain power loss. Have you said where the power is lost? Winding or core, or both? In what proportion? Which winding? ** Not part of the original question about CORE losses. The original question was about transformer losses. John pointed that mistake out to you. So are you saying all the losses you have measured are core losses? In any case - when power is being transferred to a secondary, core magnetisation ( and hence loss) is reduced. Wouldn't be a 160VA core otherwise, would it? But how is this relevant please? At HF, some power *is* transferred to an open primary. That small amount would add to winding losses, but detract from core losses. But you appear to be saying that the open primary leads to increased *core* loss. Given a bit more data these losses can be disentangled. So far what you have presented will only concince weak and grovelling minds prone to preconception. The loading effect of the primary is the culprit all right. Prove otherwise if you can - arsehole. You're the one who is making the assertion. You haven't proven anything so far. ** The figures I posted say otherwise and are consistent with theory. No they don't, although yuo are getting closer slowly as you release more details of what you did and how you did it. Nothing will convince a mad pommy **** like you Ian. Appropriate and comparable data are generally a requirement of proof. You have made some progress but you are still not there yet. Could be though, quite easily, if... I also tested a similar, complete toroidal mains transformer - the results were just like the results with valve OTs. There was a loss minima and zero phase angle around 1kHz - then a capacitive phase angle and rising HF losses above that. You've made this up. ** Get ****ed you evil pommy ****. What angle, in what direction, and how much loss? ** Learn to read - ****head. I have, but I can't see where you have posted a table of data from a test of a toroid with an open secondary winding. You should be able to tell that I am reading...you disappoint me. If you'd really done it you would have jubilantly reported the results. ** Rot - they only give the same picture that has been posted already by JS. You mean you haven't posted them? You had me looking all over... That would be useful because it is an experiment that many could repeat. ** Anyone can drive a transformer with 5 volts and measure the current and relative phase. Not everyone. I can't do it with your, or any, 160VA GOSS toroid with 40 turns on its secondary winding unless I buy a core. You have two similar cores, one populated and one not. You are in a much better position than nearly everyone *and you are already set up* so much experimental error can be eliminated if you do both tests, making them as comparable as possible. Time for YOU to do a bit of work - arsehole. But you say you have already done it, and you are already set up. Won't take long to post the data. It won't be quite comparable because it won't have 40 evenly spaced turns on the secondary, but it would be better evidence than what you have posted so far. " a loss minima " = one of the loss minima. I see. Like "a kangaroos " = one of the kangaroos. Isn't Australian english quaint. Don't forget get this: " OK you pig arrogant pommy arsehole - point us all to al the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. Why? I don't need to buy a toroidal mains transformer. They won't generally say what they are made of anyway. And who's "us"? You said "all toroidal transformers" are made of GOSS. Do you really need me to show you otherwise? Can't you look yourself? The onus is on you to prove your MAD assertions. What assertions? cheers, Ian |
#68
|
|||
|
|||
Thanks Phil. I fixed the error in my R calculation.
Pcore I RI(Z) Rl(Z) Im(Z) L 4.05E-01 8.09E-02 61.8 44.9 1.4E-01 3.48E-01 6.96E-02 71.8 50.3 8.0E-02 2.38E-01 4.75E-02 105.2 73.7 3.9E-02 1.84E-01 3.67E-02 136.1 75.4 1.2E-02 1.54E-01 3.08E-02 162.3 75.7 4.0E-03 1.25E-01 2.50E-02 199.7 80.7 1.3E-03 9.67E-02 1.93E-02 258.7 109.8 5.8E-04 7.77E-02 1.55E-02 322.0 143.3 2.3E-04 Joe |
#69
|
|||
|
|||
"Ian Iveson" "Phil Allison Your data is derived from reading peak V rather than true rms and may also be inccurate. Where did you get that idiot idea ?? So why say it's an "idiot idea"? ** Because it is a wild assumption and it is false. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. You didn't say, and no they are not...not by a very long way, silly. ** OK you pig arrogant pommy arsehole - point us all to all the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. Twisting and turning in the wind again. You said *all toroidal transformers*. ** The core I used was ***stated*** to be a 160 VA type. That makes it a mains tranny so it uses GOSS. Audio grade toroidals use the same or similar GOSS. So it is FALSE to say I did not indicate the core material. The onus is on YOU to prove your mad assertions.!!! To who? You are probably the only one who thinks *all toroidal transformers* are GOSS. ** See above - ****head. Find a 160 VA toroidal that is NOT made using GOSS. The onus is on YOU to prove your mad assertions.!!! You didn't say it is a core intended for a mains transformer either, ** What do you think 160 VA refers to - ****head ?????????????? ** Hey ****wit - a SS amp is not gonna distort driving a 50 ohm load at 50 Hz. Perhaps not, but the drive to the test piece will distort as a result of the current sensing resistor. You say you observed distortion, which I assume was to the current waveform. That means there is *definitely* voltage distortion too, again because of the sense resistor and also because of the output resistance of the amp, which is significant wrt distortion in your set up. ** ****ING BULL****. At 100mA the sense resistor had 200mV rms across it - ie 4% of 5 volts. The distortion was about 10% - so only 0.4% of the applied voltage. The figures are sufficiently accurate for the purpose. Have you said where the power is lost? Winding or core, or both? In what proportion? Which winding? ** Not part of the original question about CORE losses. The original question was about transformer losses. ** WRONG !!! The original claim disputed by JS was from Pat Turner: " Core losses in most OPTs, even in Quad II, are negligible, because the inductance of the primary is a high impedance at most signal F. " **** IT WAS ABOUT CORE LOSSES ONLY !!! In any case - when power is being transferred to a secondary, core magnetisation ( and hence loss) is reduced. Wouldn't be a 160VA core otherwise, would it? But how is this relevant please? At HF, some power *is* transferred to an open primary. That small amount would add to winding losses, but detract from core losses. But you appear to be saying that the open primary leads to increased *core* loss. ** You are one confused pommy ****wit. Nothing will convince a mad pommy **** like you Ian. Appropriate and comparable data are generally a requirement of proof. ** With no secondary in place - the previously rising power loss at HF disappears and becomes a falling power loss PLUS the phase angle stays lagging all the way. That was all I intended to prove and have. END of GAME ** Learn to read - ****head. I have, but I can't see where you have posted a table of data from a test of a toroid with an open secondary winding. ** I summarised it as very similar to JS's figures for his OT. ** Anyone can drive a transformer with 5 volts and measure the current and relative phase. Not everyone. ** Bad luck - arsehole. Don't forget get this: " OK you pig arrogant pommy arsehole - point us all to al the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. Why? ** Because you claimed they existed. I don't need to buy a toroidal mains transformer. ** Get ****ed. They won't generally say what they are made of anyway. ** They usually do - and they all say it is GOSS. And who's "us"? ** RAT readers. You said "all toroidal transformers" are made of GOSS. ** No I did not. You failed to consider the context - just like all autistic mental defectives are prone to do. The onus is on you to prove your MAD assertions. What assertions? ** All of them are mad. ............... Phil |
#70
|
|||
|
|||
"Joseph Meditz" Thanks Phil. I fixed the error in my R calculation. Pcore I RI(Z) Rl(Z) Im(Z) L 4.05E-01 8.09E-02 61.8 44.9 1.4E-01 3.48E-01 6.96E-02 71.8 50.3 8.0E-02 2.38E-01 4.75E-02 105.2 73.7 3.9E-02 1.84E-01 3.67E-02 136.1 75.4 1.2E-02 1.54E-01 3.08E-02 162.3 75.7 4.0E-03 1.25E-01 2.50E-02 199.7 80.7 1.3E-03 9.67E-02 1.93E-02 258.7 109.8 5.8E-04 7.77E-02 1.55E-02 322.0 143.3 2.3E-04 ** Shame about the L calc. The formula is L = R / ( 2 . pi . F . tan phi ) ............. Phil |
#71
|
|||
|
|||
Ian Iveson wrote: "Phil Allison" wrote Where did you get that idiot idea ?? Plucked straight out of you fat stupid arse like all the others. You said you used an oscilloscope to monitor the voltage and current waveforms. Perhaps it has a dual rms measuring or conversion facility...you didn't say so it is reasonable to assume you used your eyes to measure the peak, and then converted to rms. ** Perfectly valid method to get rms value when the waves are not distorted. So why say it's an "idiot idea"? I didn't criticise it, beyond pointing out that distortion makes it an approximation. Distortion is moot here, of course, so we need to be careful. ** GOSS - just like all toroidal transformers - you ****ing pommy tenth wit. You didn't say, and no they are not...not by a very long way, silly. ** OK you pig arrogant pommy arsehole - point us all to all the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. Twisting and turning in the wind again. You said *all toroidal transformers*. You mean "sell", BTW. Spell checker won't often help with grammar. Just exactly what some folks use in their toroidal cores could be a moot point. I don't see why all toroidal cores would have to be GOSS. Its more expensive than non oriented material, NOSS. If we wind a spiral of NOSS, there is a major increase in the U of the iron. This alone will reduce the losses for a given Np, Afe, F and V, ie Bmax, over what the losses would be for an E&I laminated transformer where NOSS measures a U of maybe 3,500. The GOSS I get here in Oz in E&I lams measures with a U = 17,000, which gives astonishly low core losses, so low in fact I just design for B = 0.85 and I know the core losses can then be neglected, much in the same way as the core losses can be neglected with a GOSS troidal core with the same B, because the U of the GOSS in a toroid is even higher than my nice Os made lams. With C-cores I ahve purchased in Oz the U of the material has varied from 4,500 to anywhere up to 11,000+ depending how well I have lapped the cut faces on a flat oil stone. But some material when lapped well only went up to the 4,500, which suggests makers didn't use very good GOSS, in fact it probably is NOSS. Lotsa BS goes on with core materials to save a penny and fool ppl The onus is on YOU to prove your mad assertions.!!! To who? You are probably the only one who thinks *all toroidal transformers* are GOSS. You didn't say it is a core intended for a mains transformer either, although you did say it is "standard". Perhaps some ppl know that "standard" means "GOSS", but I didn't, and still don't. Are you really sure of this? Perhaps I should take your word? snip, ** The threshold of visible 3H on a decent analogue scope is about 2 % . OK. Assuming it is 3H, but I won't quibble if you can't analyse it further. Core distortion without a steady DC flow one way gives mainly all odd order distortion. It might help if you were to say quite *how* you believe the primary winding is responsible for *power loss*. ** Posted that already. Where? I just got some handwaving about "tuned circuit". ** That is just what it is. But that in itself doesn't explain power loss. Have you said where the power is lost? Winding or core, or both? In what proportion? Which winding? ** Not part of the original question about CORE losses. The original question was about transformer losses. John pointed that mistake out to you. So are you saying all the losses you have measured are core losses? They have to be core losses Ian because Phil has wound 40 turns onto a toroid where he deliberately used thick wire to avoid winding losses and spaced the turns to avoid self capacitances. The winding and capacitance losses are therefore negligible. Core losses are thus easily measured. In any case - when power is being transferred to a secondary, core magnetisation ( and hence loss) is reduced. Wouldn't be a 160VA core otherwise, would it? But how is this relevant please? At HF, some power *is* transferred to an open primary. That small amount would add to winding losses, but detract from core losses. But you appear to be saying that the open primary leads to increased *core* loss. He is saying the maximum core losses are with no load on any secondary. Then when a load is connected, core losses decrease, and winding losses increase. Given a bit more data these losses can be disentangled. So far what you have presented will only concince weak and grovelling minds prone to preconception. Well we won't be grovelling to try to share your view on the matter. The loading effect of the primary is the culprit all right. Prove otherwise if you can - arsehole. You're the one who is making the assertion. You haven't proven anything so far. ** The figures I posted say otherwise and are consistent with theory. No they don't, although yuo are getting closer slowly as you release more details of what you did and how you did it. Nothing will convince a mad pommy **** like you Ian. Appropriate and comparable data are generally a requirement of proof. You have made some progress but you are still not there yet. Could be though, quite easily, if... I also tested a similar, complete toroidal mains transformer - the results were just like the results with valve OTs. There was a loss minima and zero phase angle around 1kHz - then a capacitive phase angle and rising HF losses above that. You've made this up. ** Get ****ed you evil pommy ****. What angle, in what direction, and how much loss? ** Learn to read - ****head. I have, but I can't see where you have posted a table of data from a test of a toroid with an open secondary winding. You should be able to tell that I am reading...you disappoint me. If you'd really done it you would have jubilantly reported the results. ** Rot - they only give the same picture that has been posted already by JS. You mean you haven't posted them? You had me looking all over... That would be useful because it is an experiment that many could repeat. ** Anyone can drive a transformer with 5 volts and measure the current and relative phase. Not everyone. I can't do it with your, or any, 160VA GOSS toroid with 40 turns on its secondary winding unless I buy a core. You have two similar cores, one populated and one not. You are in a much better position than nearly everyone *and you are already set up* so much experimental error can be eliminated if you do both tests, making them as comparable as possible. I tested with two randomly picked E&I lam PTs from the junk store, and a Luxman OPT, model OY15, not quite in the junk class. I got similar results with all 3 with regard to phase angle shift at low levels, and similar results to Phil and JS. There is a point where we shouldn't give in to petulant demands for more proof of what the observations we witnessed were. To do that we should have had a camera man filming out efforts. Time for YOU to do a bit of work - arsehole. But you say you have already done it, and you are already set up. Won't take long to post the data. It won't be quite comparable because it won't have 40 evenly spaced turns on the secondary, but it would be better evidence than what you have posted so far. " a loss minima " = one of the loss minima. I see. Like "a kangaroos " = one of the kangaroos. Isn't Australian english quaint. Don't forget get this: " OK you pig arrogant pommy arsehole - point us all to al the sites that sells toroidal mains transformers that DO NOT use GOSS strip for the core. Why? I don't need to buy a toroidal mains transformer. They won't generally say what they are made of anyway. And who's "us"? Everyone in the group. Its not safe to assert that all toroidal cores are GOSS of best quality, or even grain oriented at all. Some would have higher losses than others. You said "all toroidal transformers" are made of GOSS. Do you really need me to show you otherwise? Can't you look yourself? You really should do a search to find out if totoids are all GOSS or not. It'd lend weight to your argument, rather than argue all day about whose turn it is to proove something, where both of you have not checked out the facts on toroidal materials used. The onus is on you to prove your MAD assertions. What assertions? cheers, Ian Meanwhile, let us take our minds back to the Leak PT I tested the other day, and over which I was roundly condemned for all sort of travesties of criminal codes for asserting wrongly that core losses = the measured VI at the primary of an unloaded tranny. Left on the bench unloaded, I measured VI = 24 watts, but that's not the heat in the core because after 4 hrs it only got warm, and as we know now, the power factor isn't 1.0 But over the last few days I have got the amp working really well, and the tranny gets so hot I cannot keep a hand on it after 3 hrs where the ambient T = 30C, so a lotta heating must be going on in the copper. The VA draw is about 90, and I estimate total losses of 20%, so maybe 18 watts of heating are occuring, perhaps 10 watts of copper heat and 8 watts of Fe heat. I also tested the TL12 OPT for the winding losses, and these confirmed my recent general criticisms of 1955 amplifiers and especially Olde British Amplifiers and their methods of connecting secondary windings, their use of hardly any interleaving, their use of wire so thin its hardly there, and the absense of enough iron. There are 4 secondaries of 1 ohm and this allows 1,4,9,and 16 ohms. The ability to get 9 ohms involves two seriesed one ohm windings with a pair of parallel 1 ohm windings. The equivalent winding resistance including the reflected primary winding resistance when all measured at the sec = 2.66 ohms, so wind losses totalled 23%. I also measured Ra-a for the trioded KT66 and got 3.2k, and because the "9" ohm turn ratio was 18.2:1 the Ro of the triode amp without NFB was a whopping 12 ohms including the winding R. The reflected load is 4ka-a to the tubes when a 12 ohm load is connected to the 9 ohm outlet, so good load matching with triodes must have also escaped Harold Leak's mind at the time. The series + parallel windings with different current densities increased the leakage inductance so much that it was impossible to make the amp unconditionally stable without seriously reducing BW below 20 kHz. I changed the load match to suit 4 ohms, with a symetrical arrangement, and the turn ratio gave a 6k to 8 ohm load match, and marginally lower winding losses. Hey, but I thought 10k was about the right load for triodes? But the Leak TL12 has UL taps, so its meant for UL, but even so anything below 8 ohms connected to the "4 ohm" connection gives a tube RLa-a load below 6k, although the wind losse add considerably to the load actually seen by the tubes. when 4 ohms is connected to the OPT set for 4 ohms, the actual load to the tubes is about 3k plus about 600 ohms winding R. For UL, there should be 6.6k a-a for UL with 4 ohms on the 4 ohm outlet. The triodes put up with such poor respect for class A and load matching because ppl only want a watt most days. And Leak claimed he had 26 dB of applied global NFB. This makes his TL12 very fussy about loads. Without a load the TL12 oscillated badly at LF. I have made several subtle changes to include networks to make the amp at least unconditionally stable, and to add 50 uF between the OPT CT and 0V. FB has been reduced to 12 dB. Ro with FB = 0.9 ohms. With 8 ohms on the 4 ohm outlet, performance is tolerable and the amp sounds quite nice, and not doing much wrong. Basically only the 4 ohm outlet match should ever be used since the 1, 9, and 16 ohm outlets are BS, imho..... Patrick Turner. |
#72
|
|||
|
|||
"Patrick Turner" I don't see why all toroidal cores would have to be GOSS. ** There is no other way to make a continuous strip of thin silicon steel, maybe 250 metres long, EXCEPT by rolling it along the direction of the strip. You have failed to take account to context too - just like all autistics do. ( No-way was anyone alluding to Mumetal or Permalloy. ) If we wind a spiral of NOSS, there is a major increase in the U of the iron. ** WRONG - surely you meant GOSS. The GOSS I get here in Oz in E&I lams measures with a U = 17,000, which gives astonishly low core losses, so low in fact I just design for B = 0.85 and I know the core losses can then be neglected, ** With a single strip of GOSS in a toroid - the benefits are 10 - 50 times more. They have to be core losses Ian because Phil has wound 40 turns onto a toroid where he deliberately used thick wire to avoid winding losses and spaced the turns to avoid self capacitances. ** Precisely. He is saying the maximum core losses are with no load on any secondary. Then when a load is connected, core losses decrease, and winding losses increase. ** Ditto. Its not safe to assert that all toroidal cores are GOSS of best quality, ** No one did. or even grain oriented at all. ** I reckon that one is a safe bet. ............. Phil |
#73
|
|||
|
|||
I've been patient with you in the hope that you might post the
comparable evidence that you say you have observed, because it might have been useful. Time to put up or shut up. cheers, Ian |
#74
|
|||
|
|||
Pcore I RI(Z) Rl(Z) Im(Z) L
4.05E-01 8.09E-02 61.8 44.9 1.4E-01 3.48E-01 6.96E-02 71.8 50.3 8.0E-02 2.38E-01 4.75E-02 105.2 73.7 3.9E-02 1.84E-01 3.67E-02 136.1 75.4 1.2E-02 1.54E-01 3.08E-02 162.3 75.7 4.0E-03 1.25E-01 2.50E-02 199.7 80.7 1.3E-03 9.67E-02 1.93E-02 258.7 109.8 5.8E-04 7.77E-02 1.55E-02 322.0 143.3 2.3E-04 ** Shame about the L calc. The formula is L = R / ( 2 . pi . F . tan phi ) I believe there is an error in your formula for L. I think the tan phi term should be upstairs. tan phi = OPP/ADJ = IM/RE = XL/R You have L = R/(w.tan phi) wL = R/(tan phi) XL = R/(tan phi) tan phi = R/XL (?) Joe |
#75
|
|||
|
|||
Phil Allison wrote: "Patrick Turner" I don't see why all toroidal cores would have to be GOSS. ** There is no other way to make a continuous strip of thin silicon steel, maybe 250 metres long, EXCEPT by rolling it along the direction of the strip. You have failed to take account to context too - just like all autistics do. They roll huge wide rolls of Si Fe at BHP Sankey and some ends up as GOSS after annealing. Not all the material rolled is GOSS afaik, but I might do a check on that. The rolls of material are further sliced to the widths needed for toroidal cores, but I am not sure of all the processes involved. Some gets stamped into E&I and further annealed to make the GOSS E&I lams which I like, and then there is Lycore150, which isn't grain oriented, ( NOSS ), and I do't know what processes are used, but obviously less because it is half the price of GOSS. I really think you need to be more certain than you are to assume say 95% of all toroidal cores are in fact GOSS with a high max U of say above 20,000. ( No-way was anyone alluding to Mumetal or Permalloy. ) Indeed not. If we wind a spiral of NOSS, there is a major increase in the U of the iron. ** WRONG - surely you meant GOSS. No. If you take plain old NOSS, such as Lycore 150 sold by O.H.O'brien it has U = 3,500 in E&I form. There is a grain reversal when the E&Is are placed together, effectively reducing the U. If you used the same material and wound a spiral core with the grain all the same way, the effective U is raised. Same with GOSS material like the Sankey stuff which measures 17,000 even in E&I form when assembled. It too would have a higher U if spiralled. The GOSS I get here in Oz in E&I lams measures with a U = 17,000, which gives astonishly low core losses, so low in fact I just design for B = 0.85 and I know the core losses can then be neglected, ** With a single strip of GOSS in a toroid - the benefits are 10 - 50 times more. The differences between toroidal transformers I have altered by hand to increase the P turns to get 0.85T and the Samkey E&I lams is only marginal. There is not a 10 fold increase in U just because the same GOSS is used as a spiral. I did all the measurements, and sure, toroids have low losses, but not that much lower. I am not sure how much the U would rise for a given material. Keen audio amateurs in the 1950s would get raw strips of GOSS material from the metal makers ( often used for C-cores ) and cut the strip into 4 metre lengths, and wind each strip through the wound bobbins in their home workshop, and thus wind the core into the bobbin. This gave then a U increase over using E&I, or C-cores with their two cuts, but they still had to observe laws for saturation because all these varieties of iron begin to saturate at around 1.2 T to 1.4T, and with OPTs it is saturation F that deteremines the Np, not the required amount of inductance. If Fsat = 14 Hz for a given Afe, Np and V, the inductance Lp will always be sufficient for a given OPT design and core losses will always be SFA for a PP design. For SE designs, B max shouldn't be above 0.3T at 50 Hz for a given Afe, V and Np. Then the gap should be adjusted to allow ZLp = RL at 16 Hz. That's a bit of a juggling act. Nowdays, the home DIYer can buy Unicore from AEM in Sth Aust which are rectangles of strips bent to fit perfectly inside each other concentrically, and each has a cut join which is near perfect. Each "C" can be sprung open and snapped into position around a completed conventional bobbin so that the cut join is on opposite sides of the bobbin and the core lams act like a continual strip with about 95% of the U of an uncut toroidal core. AEM sent me some samples and sure enough the losses were low, and U was above 20,000, but no more than a toriod. I found it impossible to get this stuff to be quiet with a rectifier supply despite a lotta impregataion. There is no way to clamp the lams tightly together, and the varnish penetration has to be depended on to make the tranny quiet. But with a pure R load, the lams were quite silent without any varnish or clamping. I explained the problems I had to AEM who had no answers for me. AEM boast that their Unicores give the lowest of all losses, but the Sankey E&I lams I use give virtually the same benefits, and are quiet with rectifiers, so that's what I use now. AEM used to stock C-cores, maybe they still make them to order which I have had them do, but the first batch went U = 4,500, and the next batch went 7,000, and then I lapped their rough cut surfaces and I got 11,000, still less than Sankey E&I. O.H.O'Brien's have Eilor from Israel, a far better looking C-core product but I don't know what the U max of those cores are. So there are cores and there are cores, and makers deliberately make it difficult to compare their materials with somebody eles's, and the glosssy brochures are sometimes BS. Not all the same, there are variations in material and processes. They have to be core losses Ian because Phil has wound 40 turns onto a toroid where he deliberately used thick wire to avoid winding losses and spaced the turns to avoid self capacitances. ** Precisely. He is saying the maximum core losses are with no load on any secondary. Then when a load is connected, core losses decrease, and winding losses increase. ** Ditto. Its not safe to assert that all toroidal cores are GOSS of best quality, ** No one did. or even grain oriented at all. ** I reckon that one is a safe bet. ............ Phil GOSS is now routinely used in even cheap electronics in the form of E&I materials in say microwave oven trannies and much other junk Its manufacture is so vast, its cheap to do. Most toroidals would be some type of GOSS. GOSS is commonly used in E&I lams. The E&I lams are rarely fully interleaved and bolted up because this process is more expensive than simply welding inserted lumps of E with a lump of I held tight together by machine into completed bobbins of transformers. The bobbins themselves have been rationalised with concentric plastic sliding bobbins made with the primaries in the inner one to suit all the different mains around the world, and the outer windings to suit the item being exported. This allows a dreaded thermal fuse to be concealed between P&S windings. ( if it goes open, maybe you have to buy a new tranny if the ends of the windings are not accessable ). The welded E&I assembly gives an effective air gap with some amount of very rough gap shunting by the welding, and the final U is high enough to give Lp a high enough VI figure so that losses are still less than NOSS low grade EI when fully interleaved. The same technique is used in thousands of lamp ballasts for flurescent lights, and sodium street lights. Toroidals have never been universally accepted as a replacement for E&I lammed trannies because of the cheaper manufacturing costs of the E&I tranny. I still don't much like toroidals since most with Bmax = 1.25 Tesla whch is typical always makes them noisy in a large tube amp. The low losses are welcome, but there is more to a good tranny than low losses. Almost nobody is willing to use perforated or woven plastic insulation netween wires and between core and the wires so that vacuum varnishing can penetrate fully, and thus thouroughly solidify the completed tranny when varnished with epoxy in a can made to fit neatly around the tranny. All the winders I have dealt with cannot make such wonders. There is no demand for it, apart from nutters like me who expect ppl to work their arse off to make perfect things like I do for almost no income. They also wouldn't like to have to provide me with a circuit board on top of the finished tranny with all the taps I want neatly labled. Providing a gurantee for the specified parameters is out of the question. I won't repeat my many gross dissapointments with the buffoons I have met in the transformer winding industry. I have good reasons to wind all my own. Patrick Turner. |
#76
|
|||
|
|||
Ian Iveson I've been patient with you... ** You have in fact been stalking me. You have been posting utterly insane garbage and pretending it is fact. You are nothing but a usenet pest, troll and a criminal stalker. You are a seriously mentally ill person. in the hope that you might post the comparable evidence that you say you have observed, because it might have been useful. ** Useful only for your continued criminal stalking and insane nit picking. Time to put up or shut up. ** Someone should put a bullet in your head to put you out of your misery. ................ Phil |
#77
|
|||
|
|||
"Patrick Turner" = utter ****wit Phil Allison I don't see why all toroidal cores would have to be GOSS. ** There is no other way to make a continuous strip of thin silicon steel, maybe 250 metres long, EXCEPT by rolling it along the direction of the strip. They roll huge wide rolls of Si Fe at BHP Sankey and some ends up as GOSS after annealing. ** Grain orientation is a direct result of the rolling process - fool. The grain structure orients itself in the rolling direction. Some gets stamped into E&I and further annealed to make the GOSS E&I lams which I like, and then there is Lycore150, which isn't grain oriented, ( NOSS ), and I do't know what processes are used, but obviously less because it is half the price of GOSS. ** The subject is toroidal cores. I really think you need to be more certain than you are to assume say 95% of all toroidal cores are in fact GOSS with a high max U of say above 20,000. ** Iron toroidal cores are all GOSS. If you take plain old NOSS, such as Lycore 150 sold by...... If you used the same material and wound a spiral core with the grain all the same way, the effective U is raised. ** Then you have made a GOSS spiral. How asinine the Turneroid is. Same with GOSS material like the Sankey stuff which measures 17,000 even in E&I form when assembled. It too would have a higher U if spiralled. ** Since a GOSS EI core has one limb oriented the wrong way ( the back of the E stamping) - that is an obvious truism. ............... Phil |
#78
|
|||
|
|||
Your compatriot is gambling that you know even less about
steelmaking than he does. Don't give up. How do they get non-oriented steel flat, do you think? And how would that process be cheaper, do you think? What makes rolling GOSS more expensive? What can make steel easier to roll, and at the same time randomise its structure? cheers, Ian |
#79
|
|||
|
|||
"Joseph Meditz" Pcore I RI(Z) Rl(Z) Im(Z) L 4.05E-01 8.09E-02 61.8 44.9 1.4E-01 3.48E-01 6.96E-02 71.8 50.3 8.0E-02 2.38E-01 4.75E-02 105.2 73.7 3.9E-02 1.84E-01 3.67E-02 136.1 75.4 1.2E-02 1.54E-01 3.08E-02 162.3 75.7 4.0E-03 1.25E-01 2.50E-02 199.7 80.7 1.3E-03 9.67E-02 1.93E-02 258.7 109.8 5.8E-04 7.77E-02 1.55E-02 322.0 143.3 2.3E-04 ** Shame about the L calc. The formula is L = R / ( 2 . pi . F . tan phi ) I believe there is an error in your formula for L. I think the tan phi term should be upstairs. ** Nope. For the phase angle to approach 90 degrees, all the current must go via L - so L must have a low impedance value compared to R. Hence a low L value. Tan phi increases dramatically as phi approaches 90 degrees - so it must be "downstairs" to make L diminish. ........... Phil |
#80
|
|||
|
|||
For the phase angle to approach 90 degrees, all the current must go
via L - so L must have a low impedance value compared to R. Hence a low L value. Tan phi increases dramatically as phi approaches 90 degrees - so it must be "downstairs" to make L diminish. Yes Phil, you are right. And boy is my face red! I was careless in going from XL + R i.,e. a series CKT to a parallel one. When I worked it out for a || CKT I got your formula of tan phi = R/wL The darndest thing about this is that when the parallel L R CKT is that as the angle approaches 90 degrees the small L approaches a wire! Joe |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
James Randi gets clarified on audio biz | High End Audio | |||
A comparative versus evaluative, double-blind vs. sighted control test | High End Audio | |||
Mechanic blames amplifier for alternator failing?? Help>>>>>>>>>>> | Car Audio | |||
hearing loss info | Car Audio | |||
Reversing a Power Transformer | Vacuum Tubes |