Home |
Search |
Today's Posts |
#1
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote: As some of you no doubt think I sometimes do strange things mixing transistors with tubes and I got another one for your amusement. The motivation stemmed from wanted to slap something together with parts at hand and, well, not having much at hand. Any how, it boiled down to trying a 12u7; not normally very useful with a decent B+ and not a tremendous amount of gain either. But I did have an MPSA42 so I tried using it as a protective cascode for the low voltage 12u7, an amusing roll reversal as I normally use them to protect low voltage, low noise, transistors from the tube circuits (like in the Looking Glass phase splitter mirror). But, rather than just use the MPSA42 to fix a low B+ for the 12u7 I put the load R above it. That fixes the plate, at 23.3V in this case (24V minus a Vbe drop), and isolates it from the output signal. The upshot is a gain of 130.6 (120k load resistor above the cascode with AF in 8.262mV, AF out 1.079V). 2'nd harmonic is -63dB with the rest under -110dB. (in spice) I decided to do a comparison (all in spice) with a 'normal' anode R configuration and figured same plate, same AF in, same load R and same Rk (240Ohm) was fair. That meant, for a 'normal' R straight on the anode, using whatever B+ would result in 23.3V on the 12u7 plate. That turned out to be 201V (with ~1.47mA through it, matching the ~1.47mA in the fixed plate configuration). Gain was 15.08 with 2'nd down -71.9dB. That means my 'trick' increased gain by 9.38dB while increasing 2'nd harmonics 8.5dB. And while the MPSA14 was 'in the signal path" spice showed no detectable influence on the harmonic distribution. Second comparison was to set equal gains, selecting whatever load R in the fixed plate configuration would result in 15.08 (and all else equal) since there's no way to get the 'normal' configuration to 130.6. That turned out to be ~13.75k with 2'nd harmonics still at -63dB, virtually equal to the high gain case. So, little value to it, other than protecting a low voltage device, unless one is going for high gain as it sacrifices 8.5dB of distortion. On the other hand, try getting a gain of 130.5 out of a 12u7 the 'proper' way What is the exact schematic for what you have tried? I have never heard of a 12U7. Is it actually a 12AU7? Patrick Turner. |
#2
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
I have never heard of a 12U7. Is it actually a 12AU7?
Patrick Turner. http://www.nostalgiaair.org/tubes/12u7.htm Peter Wieck Wyncote, PA |
#3
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Just for chuckles I tried spicing that one into the cascode and it performed surprisingly well too, with a fixed 30V plate. So did a 6DJ8. If the spice models are good. I have a spice 3f4 model of the 12U7 at http://home.netcom.com/~wb2jia/tubes/spice4.txt Copy and paste into your spice SW. |
#4
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote: On 6 Jun 2006 13:25:44 -0700, " wrote: I have never heard of a 12U7. Is it actually a 12AU7? Patrick Turner. http://www.nostalgiaair.org/tubes/12u7.htm That's an interesting substitute suggestion they've got since the 12AE7 is double dissimilar triodes while the 12u7 is twin identicals. Different AF, gm, bias, plate resistance, and heater current. Same basing though. Peter Wieck Wyncote, PA 12U7 is like a 12AU7 but meant to operate with Ea =12.6V. Gm = 1.6mA/V, Ra = 12.5k, µ = 20, all at Ia = 1mA. Maybe they used them in car radios. I have no idea of the cascode schematic you spoke at length about, and I guess nobody else did either. Patrick Turner. |
#5
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote: On Wed, 07 Jun 2006 04:14:07 GMT, robert casey wrote: Just for chuckles I tried spicing that one into the cascode and it performed surprisingly well too, with a fixed 30V plate. So did a 6DJ8. If the spice models are good. I have a spice 3f4 model of the 12U7 at http://home.netcom.com/~wb2jia/tubes/spice4.txt Copy and paste into your spice SW. Thanks! I thought that looked familiar Yeah, I got that set, plus the other three, from your site and that's the 12u7 model I'm currently using in circuitmaker Btw, In case you're curious I put up a quick copy of two approaches I'm tinkering with under the title "The Strange Case of Mixing Low Voltage Tubes and Transistors in High Voltage Circuits or, what's that 12u7 doing in there?" http://flipperhome.dyndns.org/ Ah, now I see what you are doing at http://flipperhome.dyndns.org/LowHighMix.htm What on earth is the advantage of using a 12u7 low gm triode to power an emitter circuit of an MPSA42? If the bjt voltage gain was 100, then RL for the triode is only 1.2k, so the triode gain = 1.8 approx so all that you achieve is a triode buffer with hardly any triode gain. But it may sound well so all is not lost. The normal way to cascode is to have the high gm device powering a cathode of a triode with its grid taken to a fixed bias. Perhaps there is an advantage I cannot see... The mpsa42 has a fairly high amount of source resistance in series with the base, so since the base current = 1/hfe x Ic, a considerable signal voltage will exists at the base as well as the emitter, and in fact the gains are somewhat different.... BTW, where you have a mosfet of j-fet driving a tube cathode with fixed grid bias the overall gain is simply gm of the fet device x the above triode anode RL, and this is true regardless of whether you have a 12AT7 or 12AU7 as the triode. Patrick Turner. |
#6
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
12U7 is a low voltage (design center max 30V plate) twin medium mu triode originally intended for 12V car radio service. I had picked up a couple, along with the 6GM8s, for the low voltage wall wart stuff I was playing with (guitar to stereo preamp and the PC Speaker amp) but ended up using the 6GM8s. Here's a spice 3f4 model of the 6GM8: * 6gm8 low voltage twin triode low mu Spice 3F4 model ..SUBCKT X6GM8 P G K *the following 3 lines should be one very long line Bp P K I=((0.0652280007e-3)+(0.02616312706e-3)*V(G,K))*uramp((12.19320993)*V(G,K)+V(P,K)+(5.22 5386744)) ^ 1.5 * V(P,K)/(V(P,K)+(0.599401572)) Cgk G K 1.7P Cgp G P 1.3P Cpk P K 0.5 ..ends X6GM8 Also you can find this at the bottom of http://home.netcom.com/~wb2jia/tubes/spice4.txt |
#7
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
"Patrick Turner" wrote in message ... I have no idea of the cascode schematic you spoke at length about, and I guess nobody else did either. RDHB page 533.xi Iain |
#8
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Iain Churches wrote: "Patrick Turner" wrote in message ... I have no idea of the cascode schematic you spoke at length about, and I guess nobody else did either. RDHB page 533.xi Iain Eh? No illustration xi on p533 of the RDH4. Is the RDHB different from the RDH4? Andre Jute Visit Jute on Amps at http://members.lycos.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review |
#9
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote: On Wed, 07 Jun 2006 15:18:27 GMT, Patrick Turner wrote: flipper wrote: On Wed, 07 Jun 2006 04:14:07 GMT, robert casey wrote: Just for chuckles I tried spicing that one into the cascode and it performed surprisingly well too, with a fixed 30V plate. So did a 6DJ8. If the spice models are good. I have a spice 3f4 model of the 12U7 at http://home.netcom.com/~wb2jia/tubes/spice4.txt Copy and paste into your spice SW. Thanks! I thought that looked familiar Yeah, I got that set, plus the other three, from your site and that's the 12u7 model I'm currently using in circuitmaker Btw, In case you're curious I put up a quick copy of two approaches I'm tinkering with under the title "The Strange Case of Mixing Low Voltage Tubes and Transistors in High Voltage Circuits or, what's that 12u7 doing in there?" http://flipperhome.dyndns.org/ Ah, now I see what you are doing at http://flipperhome.dyndns.org/LowHighMix.htm What on earth is the advantage of using a 12u7 low gm triode to power an emitter circuit of an MPSA42? Hehe. Well, the 12u7 wasn't picked for it's gm, it was picked because I have a couple. The 6GM8 would be 'better' in that regard, however, after I did some modeling I decided I didn't need that much gain as I only wanted 'just enough' NFB to lower pentode output impedance. Btw, I originally mentioned a gain of 300, and it was, but I knocked it down a bit in the one posted to 166 (increased the Rk). It's not as easy as it might seem to get a lot more gm as the voltages available, and my wanting to use 1 gain stage, force a rather high load R, which means rather low idle currents, and that cuts way into the gm of high gm tubes. The 6GM8 does give me more gain, though, and I also have one of those left, but, as I mentioned above, I decided I didn't need that much gain (could always lower the load R though, for less, but I'm not sure that buys me anything) and I would kind of like to keep it for a future low voltage project where it's high gm could be put to better use. If the bjt voltage gain was 100, then RL for the triode is only 1.2k, so the triode gain = 1.8 approx But that isn't where the stage voltage gain takes place and, ideally, it would be great if the effective RL there was 0. so all that you achieve is a triode buffer with hardly any triode gain. But it may sound well so all is not lost. The normal way to cascode is to have the high gm device powering a cathode of a triode with its grid taken to a fixed bias. Well, yes. I realize that's the normal tube cascode. The MPSA14 is doing essentially the same thing. Setting aside, for the moment, real world issues of internal impedance and such, the bjt fixes the 12u7 plate at 23 volts and passes signal current on up to the 120k load R. Which is the same thing a tube cascode does except transistors are quite happy working with no more than a volt across them so I get more V available to drop idle current across the load R. I did spice it up with a tube cascode but didn't see any advantage to it. The MPSA14 isn't doing any gain but it is in the signal path so I suppose that makes it bad audiophile JuJu. Whoa, the majority of the gain is due to the mpsa14? doncha mean mpsa42? Although, rather moot if I use the current mirror phase splitter since they're 'in the path' too. Not really. The mirror tries to keep the dc currents the same but essentially any sonic effect they have is well bypassed by the 330uF cathode C. On the other hand, the current mirror phase splitter should be a near perfect power supply noise eliminator since the exact same noise would be imposed on both (I think, anyway, as I haven't analyzed the effect of the MPSA92 cascode on that one) and then negated in the PP output stage. The remaining 'noise point' would be the MPSA14 cascode Vref. I managed to eliminate all the caps in the gain stage, though Perhaps there is an advantage I cannot see... The gain, mainly. Although the origins of it was to keep the 12u7 plate under 30V. Things 'evolved' The mpsa42 has a fairly high amount of source resistance in series with the base, so since the base current = 1/hfe x Ic, a considerable signal voltage will exists at the base as well as the emitter, and in fact the gains are somewhat different.... Yes, there is some signal 'loss' to the base-emitter drive. Spice shows that to be about 121nA out of a 9.01uA signal, about 1.3% . A FET could be used to eliminate that but, overall, the loss is imperceptible and akin to having a 1% load R tolerance, or the gain being 164 vs 166, or 166 vs 168. Just doesn't seem to matter all that much. BTW, where you have a mosfet of j-fet driving a tube cathode with fixed grid bias the overall gain is simply gm of the fet device x the above triode anode RL, and this is true regardless of whether you have a 12AT7 or 12AU7 as the triode. What would be the purpose of doing that? A j-fet such as 2sk369 has gm = 40mA/V so if the RL at the triode anode is 20k, the gain is 800 regardless of what triode you use. The voltage for the fet to operate need only be so Ek of the triode is +8V, and all the other voltage is across the triode. Linearity depends on the transfer curcve for the fet since the triode effectively has the Rd of the fet as its Rk so its own thd is subject to reduction by current NFB. A mosfet with gm higher than 2sk369 would give even more gain and this gain can all be reduced by a source resistor to tailor the gain and linearize the whole stage. But a pair of mpsa42 connected as a darlington pair driving a triode would also take advanatge of the highgm needed for current drive into the cathode and the tube provides most of the voltage gain... In your circuit you need more than the output stage grid drive voltage because you only have a concertina driver stage. Your circuit reminds me of so many from the old tube era where a couple of 6BM8 were used similarly. Dynaco ST70 went further with a triode-pentode which has the pentode anode load bootstrapped to give heaps more open loop gain. Patrick Turner. Got a question, though. The design center max plate for the 12u7 is 30V but do you know what that's based on? I mean, I can't believe the thing would arc at, say, 100V so I imagine it's because of the specially treated cathode. I have never been interested in tubes meant for car battery power supplies. I don't exactly know what they did to the cathode; something about the space charge was different. The question is, what's 'acceptable' prior to conduction, I.E. prior to warmup? Say I did a tube cascode using the 12u7. The operating points are within 30V but before they come on there would be, say, 280 on the top triode, which would then swing down through warmup to 54V (27 on each), or so. Use a neon tube to limit the high turn on voltage if you are worried. if you are only experimenting, and nothing blows up, by the time you change it to another circuit maybe the tubes will be undamaged.... Patrick Turner. |
#10
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Andre Jute wrote: Iain Churches wrote: "Patrick Turner" wrote in message ... I have no idea of the cascode schematic you spoke at length about, and I guess nobody else did either. RDHB page 533.xi Iain Eh? No illustration xi on p533 of the RDH4. Is the RDHB different from the RDH4? I think our Iain meant RDH4 page 533 where at the page bottom a dissertation (xi) starts on cascode which continues into page 534 where there is a sketch of an incorrectly biased dodgy cascode circuit which isn't ever actually used like cascodes actually are used. ( the top tube has its grid drive directly connected to the bottom tube grid.) Probably the author is trying to dodge trouble around patents or some other reason. But otherwise considerable math is presentented about cascode. But cascode gain is easy to work out as well as Effective Rout from the top anode. The load on the bottom tube anode = anode load of the top tube / gain of the top tube. This is where top tube has a fixed grid voltage. Patrick Turner. Andre Jute Visit Jute on Amps at http://members.lycos.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review |
#11
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Patrick Turner wrote: Andre Jute wrote: Iain Churches wrote: "Patrick Turner" wrote in message ... I have no idea of the cascode schematic you spoke at length about, and I guess nobody else did either. RDHB page 533.xi Iain Eh? No illustration xi on p533 of the RDH4. Is the RDHB different from the RDH4? I think our Iain meant RDH4 page 533 where at the page bottom a dissertation (xi) starts on cascode which continues into page 534 where there is a sketch of an incorrectly biased dodgy cascode circuit which isn't ever actually used like cascodes actually are used. ( the top tube has its grid drive directly connected to the bottom tube grid.) Probably the author is trying to dodge trouble around patents or some other reason. But otherwise considerable math is presentented about cascode. But cascode gain is easy to work out as well as Effective Rout from the top anode. The load on the bottom tube anode = anode load of the top tube / gain of the top tube. This is where top tube has a fixed grid voltage. Patrick Turner. No, no, no, Patrick, this won't do at all. I (and all the other illiterates on RAT) demand a piccie. How can we be expected to understand anything without a piccie? This Iain Churches is *oppressing* by his elite learning when he demands we turn a page to discover further information. Contrast, for instance, you own patience, again demonstrated above, when we innumerates of RAT, by far the majority (and therefore in a democracy by definition always *in the right*), demand you do our math for us. I think, and of course everyone who knows what is good for him will agree, that Iain should draw us a picture. QED. Andre Jute Visit Jute on Amps at http://members.lycos.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review ....with plenty of piccies of course! PS I think the illustration in the RDH is one of those thingies that says loud and clear to those who have burned their fingers a few times "here's the principle but if you build it too bad it won't work until you read a few pages further and instead build the actual working schematic". The authors of a book like the RDH, and particularly in its period, before the socialization of education, when even engineers were usually correctly presumed to be gentlemen, had absolutely no worries about frivolous suits for patent infringement. |
#12
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Andre Jute wrote: Patrick Turner wrote: Andre Jute wrote: Iain Churches wrote: "Patrick Turner" wrote in message ... I have no idea of the cascode schematic you spoke at length about, and I guess nobody else did either. RDHB page 533.xi Iain Eh? No illustration xi on p533 of the RDH4. Is the RDHB different from the RDH4? I think our Iain meant RDH4 page 533 where at the page bottom a dissertation (xi) starts on cascode which continues into page 534 where there is a sketch of an incorrectly biased dodgy cascode circuit which isn't ever actually used like cascodes actually are used. ( the top tube has its grid drive directly connected to the bottom tube grid.) Probably the author is trying to dodge trouble around patents or some other reason. But otherwise considerable math is presentented about cascode. But cascode gain is easy to work out as well as Effective Rout from the top anode. The load on the bottom tube anode = anode load of the top tube / gain of the top tube. This is where top tube has a fixed grid voltage. Patrick Turner. No, no, no, Patrick, this won't do at all. I (and all the other illiterates on RAT) demand a piccie. How can we be expected to understand anything without a piccie? This Iain Churches is *oppressing* by his elite learning when he demands we turn a page to discover further information. Contrast, for instance, you own patience, again demonstrated above, when we innumerates of RAT, by far the majority (and therefore in a democracy by definition always *in the right*), demand you do our math for us. I think, and of course everyone who knows what is good for him will agree, that Iain should draw us a picture. QED. LOL. I'll have to do. Time and patience is not someting blokes who are tree kwaters irish always have an abundance of. I don't have time to exactly check out the math for cascode in RDH 4 for a pair of triodes. But to get the numbers for a cascode where the top triode grid is grounded, bottom Rk bypassed, then as I said above top tube gain = µ x RL / (RL + Ra ) or simply Va / ( Vg to Vk ) . RL of the bottom tube anode looking into the top tube cathode = top tube RL / top tube gain. Gain of the bottom tube = µ x Rkin of top tube / Ra + Rkin of top tube. Gain of total overall = gain of top tube x gain of bottom tube. The above first principles logic flow allows the use of dissimilar triodes, say 12AU7 on the top and 6DJ8 on the bottom, or vice versa. For ECC88/6922/6DJ8, with typical total RL = 40k load for top triode, gain top = 25 approx, Rkin top = 40 / 25 = 1.375k, gain of the bottom = approx 7. Total gain = 25 x 7 = 175. Rout at the top tube anode = RL of top anode in parallel with effective Ra of top tube. Ra of top tube + [( µ + 1) x Ra of bottom tube.] In this case with 6DJ8 its 40k // 160k approx = 32k approx, in other words marginally lower than the RL, but very much above Ra for the top tube. The cascode thus works like a poor man's pentoad. The miller effect is reduced because bottom tube gain is lower than if you had the tubes cascAded rather than casOded. Where the cascode is good is in an RF circuit where the top tube RL is only maybe 3k, and total gain is maybe only 25, but with no miller effect since the bottom tube gain is very low. The cascode circuit makes an excellent input stage for an MM amp. And with a j-fet on the bottom its excellent for MC. Allen Wright made a big deal of cascode in line stages in FVP and in some secret input LTPs for PP amps in a "Forced Symmetry" idea, ( to which I muttered "why force anything Allen, why not just let it be?") The Hedge circuit of the 1960s also used cascoded pairs of triodes in a cross coupled LTP gain block. But I don't like them in power amps; too slow and not enough voltage headroom, and the Rout is high like a pentode, and I like driving output tubes with the low Ra of triodes. Patrick Turner. Andre Jute Visit Jute on Amps at http://members.lycos.co.uk/fiultra/ "wonderfully well written and reasoned information for the tube audio constructor" John Broskie TubeCAD & GlassWare "an unbelievably comprehensive web site containing vital gems of wisdom" Stuart Perry Hi-Fi News & Record Review ...with plenty of piccies of course! PS I think the illustration in the RDH is one of those thingies that says loud and clear to those who have burned their fingers a few times "here's the principle but if you build it too bad it won't work until you read a few pages further and instead build the actual working schematic". The authors of a book like the RDH, and particularly in its period, before the socialization of education, when even engineers were usually correctly presumed to be gentlemen, had absolutely no worries about frivolous suits for patent infringement. |
#13
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
AH, AH. I found it. Cpk P K 0.5 is missing the p for picofarad Now it works. Thanks. corrected my spice4.txt file per above, thanks Don't happen to have a 6GU7 model, do you? Try this: * 6GU7 Spice 3F4 model ..SUBCKT X6GU7 P G K Bp P K I=(0.06810278442e-3)*uramp(V(P,K)*ln(1.0+(-0.07171045201)+exp((2.250885686)+(2.250885686)*((2 2.31757643)+(-193.4922211e-3)*V(G,K))*V(G,K)/sqrt((48.95241165)^2+(V(P,K)-(24.3417218))^2)))/(2.250885686))^(1.260156564) Cgk G K 3.4P Cgp G P 3P Cpk P K 0.44p ..ends X6GU7 http://home.netcom.com/~wb2jia/tubes/spice4.txt at the bottom |
#14
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
AH, AH. I found it. Cpk P K 0.5 is missing the p for picofarad Now it works. Thanks. Thanks, corrected it in http://home.netcom.com/~wb2jia/tubes/spice4.txt Don't happen to have a 6GU7 model, do you? Added it to the bottom of http://home.netcom.com/~wb2jia/tubes/spice4.txt * 6GU7 Spice 3F4 model ..SUBCKT X6GU7 P G K * next few lines below should be one line Bp P K I=(0.06810278442e-3)*uramp(V(P,K)*ln(1.0+(-0.07171045201)+exp((2.250885686)+(2.250885686)*((2 2.31757643)+(-193.4922211e-3)*V(G,K))*V(G,K)/sqrt((48.95241165) ^2+(V(P,K)-(24.3417218)) ^2)))/(2.250885686)) ^(1.260156564) *the few lines above should be one line Cgk G K 3.4P Cgp G P 3P Cpk P K 0.44p ..ends X6GU7 |
#15
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote:
Super! Thanks I've got 4 of those but didn't have a model. Do you have a quick and simple way of creating new models? One that I could use to do the same thing? And pentodes? Reason I ask is I get oddball tubes from time to time and it would be great to have a simple means to whip up a spice model for them. I've been using curvecaptor for triodes and diodes. But no support for pentodes unless they are triode connected. See http://sourceforge.net/projects/curvecaptor/ and http://sourceforge.net/project/showf...roup_id=138442 This program is a bit hard to learn. What you need as source data is a gif image file (jpg I think works too) of the plate curves. From the program you pick out which gif file, then with the mouse you start plotting data points onto the image file. First are voltages (Vp axis), then mouse up to a pull down menu to get into the plate current mode Ip axis), and after that "plate chasterictics curve". You can change the numbers in the little white boxes to the right of the menu selections. Vp and Ip increment, but Vg does not. To start on Vg set it to 0v and plant crosses (with clicking the mouse) at short intervals along the 0v grid curve. Then for say grid voltage at -2v you have to edit the Vg voltage in the little white box up top. The increment number doesn't do anything here. Then start clicking on the -2V grid curve. "Save Markers" fairly often, as I haven't figured out how to correct a misplanted + yet. When I screw it up I revert to teh last batch of saved markers. Eventually you get all the curves done, and then save markers and then at the bottom hit "capture curves" and then you see a list of numbers. Then hit "build model" and after about a half minute you'll see the program's best fit to one of about 8 template models. Then hit "see spice model" to get the model. On this screen, down below is a selection of several varieties of spice model code (3f4, ORCAD, Circuitmaker, and such). At this point just cut and paste (ctrl C and ctrl V) into your spice simulator. There are a few syntax sillyness you'll have to edit (like adding the X in front of the tube name). |
#16
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote:
On Fri, 09 Jun 2006 22:11:36 GMT, robert casey wrote: AH, AH. I found it. Cpk P K 0.5 is missing the p for picofarad Now it works. Thanks. Thanks, corrected it in http://home.netcom.com/~wb2jia/tubes/spice4.txt Don't happen to have a 6GU7 model, do you? Added it to the bottom of http://home.netcom.com/~wb2jia/tubes/spice4.txt Super! Thanks I've got 4 of those but didn't have a model. Do you have a quick and simple way of creating new models? One that I could use to do the same thing? And pentodes? Reason I ask is I get oddball tubes from time to time and it would be great to have a simple means to whip up a spice model for them. Also read http://www.diyaudio.com/forums/showthread/t-56327.html Mentions that you also need Tcl/Tk to run the GUI - get it from ActiveState http://www.activestate.com/Products/...x?id=ActiveTcl Then run the curvecaptor.tcl (if you did associations properly when installing Tcl/Tk, you'll just need to double-click it). Read this thread for more info. 'make' is a build tool of Unix origins. It is used for automatic compilation of source code. You don't have to do anything, though, as you already have the binary distribution. |
#17
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
But is there some requirement about the graph dimensions? My first one was apparently too big (I figured the bigger the better to pick points) and there doesn't seem to be any 'fit' option. Guess you have to edit the size of the image file using your favorite image editor, like Photoshop. The setting of the size (resolution, ie, 800x600) on your VGA display That also made me wonder about alignment as some of mine are apparently manual scans of a book page and are either rotated a bit I figured the Vp and Ip axises point plots are used by the program to compensate out any rotation errors. or, even worse, distorted from a good orthogonal. DOn't know if the program will remove non linear image distortions |
#18
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Mucho thanks. Too bad there isn't a similar one for pentodes. But this will help a LOT since, as I mentioned, I often end up with oddball tubes. Thanks. Be sure to post the models you make. Thanks again |
#19
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
OK, but NEWBIE ALERT, NEWBIE ALERT This is my first shot at it after carefully reading the instructions that don't exist grin All I claim is they 'do something' in my circuitmaker Which, in case anyone missed that, means they're circuitmaker models. My models made with curvecaptor ************************************************** ******************************* After adjusting for some cut and paste stupidity between my newsreader to notepad and then to circuitmaker, your models look quite good. I have a simple curve tracer simulation I use to show plate curves that I can compare to published curves. Below is a cut and paste of my triodecurv.ckt file, and hopefully you can create a copy of this in your circuits directory and hopefully circuitmaker can make sense of it. Start to cut and paste at the line "CircuitMaker Test" and end at the last zero at the bottom. triodecurv.ckt: CircuitMaker Text 5.6 Probes: 5 v1#branch AC Analysis 0 130 22 65280 v1#branch Operating Point 0 130 22 65280 v1#branch Transient Analysis 0 130 22 65280 v1#branch Fourier Analysis 0 130 22 65280 v1#branch DC Sweep 0 131 22 65280 41 35 62 4 97 108 86 46 82 16 73 113 113 59 105 107 0 5 0 1e+09 0.001 2 33 13 10 13 12 18 10 20 20 12 20 10 13 13 10 20 13 46 20 14 20 18 17 14 16 20 20 20 20 10 13 20 18 11 0 0 30 100 10 176 75 1020 719 7 5.000 V 7 5.000 V 3 GND 8.75e+06 10 24 100 0 1 0 20 Package,Description, 30 C:\PROGRAM FILES\CM60S\BOM.DAT 0 7 2 4 0.500000 0.500000 344 171 1188 493 144179216 0 0 0 0 0 0 0 8 7 Triode~ 219 129 114 0 3 7 0 4 3 2 0 0 0 336 0 4 6GU7 30 -5 58 3 2 V2 44 -26 58 -18 0 0 14 %D %1 %2 %3 %S 0 0 0 10 0 1 2 3 1 2 3 6 7 8 27928028 88 0 0 0 0 0 0 0 1 V 5130 0 0 2 38877.7 0 0 7 Ground~ 168 62 126 0 1 3 0 2 0 0 0 53360 0 0 4 GND2 -14 -26 14 -18 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 391 0 0 2 38877.7 1 0 5 SAVE- 218 130 22 0 10 11 0 0 0 0 0 0 0 0 0 0 1 0 0 0 57552 0 1 A 3 -26 10 -18 0 0 0 14 *DC 6.65 153 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 4 SAVE 3124 0 0 2 38877.7 2 0 7 Ground~ 168 239 141 0 1 3 0 2 0 0 0 53360 0 0 0 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 3421 0 0 2 38877.7 3 0 9 V Source~ 197 130 40 0 2 5 0 5 4 0 0 0 16752 0 2 0V 15 -2 29 6 2 V1 15 -12 29 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 8157 0 0 2 38877.7 4 0 9 V Source~ 197 239 95 0 2 5 0 5 2 0 0 0 17008 0 4 300V 9 -2 37 6 3 VDS 12 -12 33 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 5572 0 0 2 38877.7 5 0 9 V Source~ 197 64 72 0 2 5 0 3 2 0 0 0 17008 0 3 10V 12 -2 33 6 3 VGS 12 -12 33 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 8901 0 0 2 38877.7 6 0 7 Ground~ 168 116 164 0 1 3 0 2 0 0 0 53360 0 0 4 GND1 -14 -26 14 -18 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 7361 0 0 2 38877.7 7 0 6 2 1 3 0 0 4224 0 1 7 0 0 3 103 114 103 51 64 51 3 1 2 0 0 4096 0 1 8 0 0 3 118 137 118 158 116 158 2 1 4 0 0 12416 0 5 1 0 0 4 130 61 130 74 129 74 129 88 2 1 2 0 0 8320 0 7 2 0 0 3 64 93 62 93 62 120 2 1 2 0 0 0 0 6 4 0 0 2 239 116 239 135 1 1 5 0 0 8320 0 5 6 0 0 4 130 19 130 12 239 12 239 74 0 0 4 0 0 0 0 3 VDS 0 500 2 3 VGS 0 -30 -2 3 0 1 4 0 8e-07 1e-08 1e-08 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 -1 10 10 10 0 10 10 0 3368 2259520 100 100 0 0 77 66 617 396 0 97 640 568 617 66 77 66 617 66 617 396 0 0 0 0 0 0 0 0 12401 0 4 100 200 0 0 0 100 100 0 0 77 66 767 186 0 0 0 0 767 66 77 66 767 66 767 186 0 0 0 0 0 0 0 0 12401 0 0 1 100 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 |
#20
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
robert casey wrote:
OK, but NEWBIE ALERT, NEWBIE ALERT This is my first shot at it after carefully reading the instructions that don't exist grin All I claim is they 'do something' in my circuitmaker Which, in case anyone missed that, means they're circuitmaker models. My models made with curvecaptor ************************************************** ******************************* After adjusting for some cut and paste stupidity between my newsreader to notepad and then to circuitmaker, your models look quite good. I have a simple curve tracer simulation I use to show plate curves that I can compare to published curves. Below is a cut and paste of my triodecurv.ckt file, and hopefully you can create a copy of this in your circuits directory and hopefully circuitmaker can make sense of it. Start to cut and paste at the line "CircuitMaker Test" and end at the last zero at the bottom. triodecurv.ckt: CircuitMaker Text 5.6 Probes: 5 snip Looks like it works, I posted it and then cut and pasted from my post back to circuitmaker, so nothing got munged. However, to get it to run right, go into "Simulation" on the top toolbar, then pick "Analysis setup". Enable the checkbox to the left of the DC button, and then click the DC button. Unckeck all other buttons. When you do click the DC button you'll see a window with primary and secondary (enable both). You want VDS as primary and VGS as secondary. (now I remember what I did to make this, I modified the FET curvetracer cir file, which I cannot find right now). VDS is the plate B+ supply that circuitmaker will vary, and VGS is the grid voltage that also gets varied. Start, stop and step size I think are self explanatory, remember to make the grid voltages negative. |
#21
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
"Andre Jute" wrote in message oups.com... No, no, no, Patrick, this won't do at all. I (and all the other illiterates on RAT) demand a piccie. How can we be expected to understand anything without a piccie? This Iain Churches is *oppressing* by his elite learning when he demands we turn a page to discover further information. Contrast, for instance, you own patience, again demonstrated above, when we innumerates of RAT, by far the majority (and therefore in a democracy by definition always *in the right*), demand you do our math for us. I think, and of course everyone who knows what is good for him will agree, that Iain should draw us a picture. QED. Hi Andre. I drew it out, with a big stick in the sand. Then, while I was away looking for my digicamera, a 6ft wave came in........:-(( regards to all Iain |
#22
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Hi Andre. I drew it out, with a big stick in the sand. Then, while I
was away looking for my digicamera, a 6ft wave came in........:-(( regards to all Iain Are you suggesting that Mr. McCoy go "pound sand"? Peter Wieck Wyncote, PA |
#23
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
"Andre Jute" wrote in message oups.com... Iain Churches wrote: "Patrick Turner" wrote in message ... I have no idea of the cascode schematic you spoke at length about, and I guess nobody else did either. RDHB page 533.xi Iain Eh? No illustration xi on p533 of the RDH4. Is the RDHB different from the RDH4? Hello Andre I am referring to RDH4 Classic ed. The text re cascode begins page533. There is a fundamental circuit fig 12.51B on page 544. Morgan Jones "Valve Amplifiers" has a good description of the cascode page 90-96 with a full circuit diagram on page 91 (fig 2.17) Strangely enough, as Patrick points out, comparison of the RDH4 diagram with schematics found elsewhere shows up some remarkable discrepancies. People seem to have favoured the cascode for its low noise floor. It was said to have the advantages of the pentode with none of the drawbacks. Regards to all Iain |
#24
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Yeah, it cut and pasted just fine and the DC was already set with the appropriate parameters. The only 'bleep' was the 6GU7 model it couldn't find because your name for it was a bit different. In short, it works. Great. Below is a ckt file for diode curves: CircuitMaker Text 5.6 Probes: 5 v1#branch AC Analysis 0 130 22 65280 v1#branch DC Sweep 0 130 22 65280 v1#branch Operating Point 0 130 22 65280 v1#branch Transient Analysis 0 130 22 65280 v1#branch Fourier Analysis 0 130 22 65280 41 35 62 4 97 108 86 46 82 16 73 113 113 59 105 107 0 5 0 1e+09 0.001 2 33 13 10 13 12 18 10 20 20 12 20 10 13 13 10 20 13 46 20 14 20 18 17 14 16 20 20 20 20 10 13 20 18 11 0 0 30 100 10 176 75 1020 719 7 5.000 V 7 5.000 V 3 GND 8.75e+06 10 24 100 0 1 0 20 Package,Description, 30 C:\PROGRAM FILES\CM60S\BOM.DAT 0 7 2 4 0.500000 0.500000 344 171 1188 493 144179216 0 0 0 0 0 0 0 6 10 Vac Diode~ 219 138 114 0 2 5 0 3 2 0 0 0 336 0 3 5U4 42 -5 63 3 2 V2 51 -26 65 -18 0 0 11 %D %1 %2 %M 0 0 0 5 0 1 2 1 2 -1610612676 88 0 0 0 0 0 0 0 1 V 5130 0 0 2 5.89246e-315 0 0 7 Ground~ 168 129 175 0 1 3 0 2 0 0 0 53360 0 0 4 GND1 -14 -26 14 -18 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 391 0 0 2 5.89246e-315 5.26354e-315 0 5 SAVE- 218 130 22 0 10 11 0 0 0 0 0 0 0 0 0 0 1 0 0 0 57552 0 1 A 3 -26 10 -18 0 0 0 14 *DC 6.65 153 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 4 SAVE 3124 0 0 2 5.89246e-315 5.30499e-315 0 7 Ground~ 168 239 141 0 1 3 0 2 0 0 0 53360 0 0 0 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 3421 0 0 2 5.89246e-315 5.32571e-315 0 9 V Source~ 197 130 40 0 2 5 0 4 3 0 0 0 16752 0 2 0V 15 -2 29 6 2 V1 15 -12 29 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 8157 0 0 2 5.89246e-315 5.34643e-315 0 9 V Source~ 197 239 95 0 2 5 0 4 2 0 0 0 17008 0 4 300V 9 -2 37 6 3 VDS 12 -12 33 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 5572 0 0 2 5.89246e-315 5.3568e-315 0 4 2 1 2 0 0 8320 0 1 2 0 0 3 127 137 129 137 129 169 2 1 3 0 0 4224 0 5 1 0 0 3 130 61 130 88 138 88 2 1 2 0 0 0 0 6 4 0 0 2 239 116 239 135 1 1 4 0 0 8320 0 5 6 0 0 4 130 19 130 12 239 12 239 74 0 0 4 0 0 0 0 3 VDS 0 30 0.1 0 0 0 0 3 0 1 4 0 8e-07 1e-08 1e-08 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 -1 10 10 10 0 10 10 0 2908 2259520 100 100 0 0 77 66 617 396 0 97 640 568 617 66 77 66 617 66 617 396 0 0 0 0 0 0 0 0 12401 0 4 10 20 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 |
#25
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
The only 'bleep' was the 6GU7 model it couldn't find because your name for it was a bit different. In short, it works. And this one is for pentodes, one pentode connected, and another triode connected. Note that this ckt assumes that the G3 is tied to the cathode, thus the use of the tetrode symbol. CircuitMaker Text 5.6 Probes: 5 v1#branch AC Analysis 0 130 22 65280 v1#branch Operating Point 0 130 22 65280 v1#branch Transient Analysis 0 130 22 65280 v1#branch Fourier Analysis 0 130 22 65280 v1#branch DC Sweep 0 132 23 65280 41 35 62 4 97 108 86 46 82 16 73 113 113 59 105 107 0 5 0 1e+09 0.001 2 33 13 10 13 12 18 10 20 20 12 20 10 13 13 10 20 13 46 20 14 20 18 17 14 16 20 20 20 20 10 13 20 18 11 0 0 30 100 10 176 75 1020 719 7 5.000 V 7 5.000 V 3 GND 0 0 24 100 0 1 0 20 Package,Description, 30 C:\PROGRAM FILES\CM60S\BOM.DAT 0 7 2 4 0.500000 0.500000 344 171 1188 493 144179218 0 0 0 0 0 0 0 13 7 Ground~ 168 95 269 0 1 3 0 2 0 0 0 53360 0 0 4 GND3 -14 -26 14 -18 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 8901 0 0 2 5.89289e-315 0 0 8 Tetrode~ 219 107 232 0 4 9 0 4 4 3 2 0 0 0 336 0 4 6AU6 26 -30 54 -22 2 V6 47 -26 61 -18 0 0 17 %D %1 %2 %3 %4 %S 0 0 0 9 0 1 2 3 4 1 2 3 4 0 88 0 0 0 0 0 0 0 1 V 7361 0 0 2 5.89289e-315 5.26354e-315 0 9 V Source~ 197 200 33 0 2 5 0 6 4 0 0 0 16752 0 2 0V 15 -2 29 6 2 V5 15 -12 29 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 4747 0 0 2 5.89289e-315 5.30499e-315 0 9 V Source~ 197 351 102 0 2 5 0 7 5 0 0 0 16752 0 2 0V 15 -2 29 6 2 V4 15 -12 29 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 972 0 0 2 5.89289e-315 5.32571e-315 0 2 +V 167 447 177 0 1 3 0 7 0 0 0 53744 0 4 100V -14 -22 14 -14 2 V3 15 -12 29 -4 0 0 13 %D %1 0 DC %V 0 0 0 3 0 1 1 0 86 0 0 0 1 0 0 0 1 V 3472 0 0 2 5.89289e-315 5.34643e-315 0 8 Tetrode~ 219 140 119 0 4 9 0 8 5 3 2 0 0 0 336 0 4 6AU6 26 -30 54 -22 2 V2 47 -26 61 -18 0 0 17 %D %1 %2 %3 %4 %S 0 0 0 9 0 1 2 3 4 1 2 3 4 0 88 0 0 0 0 0 0 0 1 V 9998 0 0 2 5.89289e-315 5.3568e-315 0 7 Ground~ 168 62 126 0 1 3 0 2 0 0 0 53360 0 0 4 GND2 -14 -26 14 -18 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 3536 0 0 2 5.89289e-315 5.36716e-315 0 5 SAVE- 218 130 22 0 10 11 0 0 0 0 0 0 0 0 0 0 1 0 0 0 57552 0 1 A 3 -26 10 -18 0 0 0 14 *DC 6.65 153 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 4 SAVE 4597 0 0 2 5.89289e-315 5.37752e-315 0 7 Ground~ 168 239 141 0 1 3 0 2 0 0 0 53360 0 0 0 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 3835 0 0 2 5.89289e-315 5.38788e-315 0 9 V Source~ 197 130 40 0 2 5 0 6 8 0 0 0 16752 0 2 0V 15 -2 29 6 2 V1 15 -12 29 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 3670 0 0 2 5.89289e-315 5.39306e-315 0 9 V Source~ 197 239 95 0 2 5 0 6 2 0 0 0 17008 0 4 300V 9 -2 37 6 3 VDS 12 -12 33 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 5616 0 0 2 5.89289e-315 5.39824e-315 0 9 V Source~ 197 64 72 0 2 5 0 3 2 0 0 0 17008 0 3 10V 12 -2 33 6 3 VGS 12 -12 33 -4 0 0 11 %D %1 %2 %V 0 0 0 5 0 1 2 1 2 0 86 0 0 0 1 0 0 0 2 Vs 9323 0 0 2 5.89289e-315 5.40342e-315 0 7 Ground~ 168 128 167 0 1 3 0 2 0 0 0 53360 0 0 4 GND1 -14 -26 14 -18 0 4 GND; 0 0 0 0 3 0 1 1 0 0 0 0 0 1 0 0 0 3 GND 317 0 0 2 5.89289e-315 5.4086e-315 0 13 4 1 2 0 0 8192 0 2 1 0 0 3 96 255 95 255 95 263 0 3 3 0 0 16384 0 0 2 8 0 6 102 122 86 122 86 177 74 177 74 235 81 235 1 0 4 0 0 8192 0 2 0 0 4 3 107 206 107 191 200 191 2 2 4 0 0 4224 0 3 2 0 0 3 200 54 200 229 133 229 2 2 5 0 0 12416 0 6 4 0 0 5 166 116 184 116 184 198 351 198 351 123 1 0 6 0 0 0 0 3 0 0 13 2 200 12 200 12 1 1 7 0 0 8320 0 4 5 0 0 4 351 81 418 81 418 186 447 186 3 1 3 0 0 8320 0 6 12 0 0 4 114 122 102 122 102 51 64 51 4 1 2 0 0 8192 0 6 13 0 0 3 129 142 128 142 128 161 2 1 8 0 0 4224 0 10 6 0 0 4 130 61 130 79 140 79 140 93 2 1 2 0 0 8320 0 12 7 0 0 3 64 93 62 93 62 120 2 1 2 0 0 0 0 11 9 0 0 2 239 116 239 135 1 1 6 0 0 8320 0 10 11 0 0 4 130 19 130 12 239 12 239 74 0 0 4 0 0 0 0 3 VDS 0 300 1 3 VGS 0 -10 -1 3 0 1 4 0 5e-05 2.5e-08 2.5e-08 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 -1 10 10 10 0 10 10 0 2992 2259520 100 229 0 0 77 66 617 396 0 97 640 568 617 66 77 66 617 66 617 396 0 0 0 0 0 0 0 0 12401 0 4 50 200 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 |
#26
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Looks like it works but I have no diode models and I've never been able to figure out how to add a model when circuitmaker grays out the Model Data button. Long time ago I faked it by editing a solid state diode model to become a tube diode. I've since have a tube diode model, but I forgot how I did that. |
#27
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
I managed to make a new sub class "tube diode" but, in keeping with all the other screwball things in circuitmaker, it won't let me name it vacdiode, because that exists (was hoping it would overwrite, substitute, or default to adding a model since none are in it), won't let me delete vacdiode, which it swears I must do in order to name mine that, and it won't let me add anything to the existing vacdiode nor will it macro copy anything to it either. So I don't know how you got yours into vacdiode.sub. Rename your old vacdiode.sub file to something else, and try using the below to create a new vacdiode.sub file *tube rectifier ..SUBCKT X38HK7R A K b1 A K i=5.3e-3*uramp(V(A,K))^(1.51) Cpk A K 11P ..ends X38HK7R *tube rectifier (heater not done) ..SUBCKT X35W4 A K b1 A K i=3.4e-3*uramp(V(A,K))^(1.44) Cpk A K 2.4P ..ends X35W4 *tube rectifier ..SUBCKT X35Z5 A K b1 A K i=3.4e-3*uramp(V(A,K))^(1.44) Cpk A K 2.4P ..ends X35Z5 *tube diode detector ..SUBCKT X6AL5 A K b1 A K i=2.2e-3*uramp(V(A,K))^(1.44) Cpk A K 2.46P ..ends X6AL5 *tube rectifier (heater not done) ..SUBCKT X5641 A K b1 A K i=1.04e-3*uramp(V(A,K))^(1.44) Cpk A K 2.4P ..ends X5641 *tube diode detector ..SUBCKT X6AL5-4vh A K *reduced heater voltage, crude limited plate current above 8V b1 A K i=2.4e-3*(uramp(V(A,K))-uramp(V(A,K)-8))^(1.24) Cpk A K 2.46P ..ends X6AL5-4vh * 3DG4 - DH full-wave power rectifier (350mA) one section ..subckt X3DG4 P K * Child-Langmuir law: mean fit error 3.52662 mA * Bp P K I=(2.698962402m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 2.22284 mA Bp P K I=(2.733421052e-3)*uramp(V(P,K)+(-0.3709046687))^(1.5) * Perugini model: mean fit error 2.21657 mA * Bp P K I=((2.604459133m)+(-0.0005396471359m)*V(P,K))*uramp(V(P,K)+(-0.2400294538))^(1.514047938) ..ends X3DG4 * 5AR4 - IDH full-wave power rectifier (250mA) one section ..subckt X5AR4 P K * Child-Langmuir law: mean fit error 3.57654 mA * Bp P K I=(3.241174316m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 3.48688 mA * Bp P K I=(3.257047182e-3)*uramp(V(P,K)+(-0.1046690061))^(1.5) * Perugini model: mean fit error 0.685736 mA Bp P K I=((1.801056955e-3)+(-0.009848868441e-3)*V(P,K))*uramp(V(P,K)+(0.5667044198))^(1.7184223 81) ..ends X5AR4 * GZ34 - IDH full-wave power rectifier (250mA) ..subckt XGZ34 P K * Child-Langmuir law: mean fit error 5.32372 mA * Bp P K I=(3.968200684m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 4.64243 mA Bp P K I=(4.050621191e-3)*uramp(V(P,K)+(-0.3522660485))^(1.5) * Perugini model: mean fit error 1.81874 mA * Bp P K I=((1.770910894m)+(-0.002635208504m)*V(P,K))*uramp(V(P,K)+(1.90184266) )^(1.721344738) ..ends XGZ34 * 5AU4 - DH full-wave power rectifier (325mA) one section ..SUBCKT X5AU4 P K * Child-Langmuir law: mean fit error 2.91988 mA * Bp P K I=(0.9904121399m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 2.03865 mA Bp P K I=(1.001318345e-3)*uramp(V(P,K)+(-0.5914399553))^(1.5) * Perugini model: mean fit error 1.39243 mA * Bp P K I=((1.243495938m)+(0.001247697537m)*V(P,K))*uramp( V(P,K)+(-0.8353102945))^(1.43333432) ..ends X5AU4 * 5V3 - DH full-wave power rectifier (350mA) one section ..subckt X5V3 P K * Child-Langmuir law: mean fit error 4.98363 mA * Bp P K I=(1.085098267m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 4.57773 mA Bp P K I=(1.093537823e-3)*uramp(V(P,K)+(-0.4353512857))^(1.5) * Perugini model: mean fit error 2.33562 mA * Bp P K I=((0.9707914136m)+(-0.0009299376857m)*V(P,K))*uramp(V(P,K)+(-1.077144669))^(1.549894427) ..ends X5V3 * 5U4 - DH full-wave power rectifier (275mA) one section ..subckt X5U4 P K * Child-Langmuir law: mean fit error 2.37547 mA * Bp P K I=(0.7784729004m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 0.949764 mA Bp P K I=(0.787802982e-3)*uramp(V(P,K)+(-0.6644060839))^(1.5) * Perugini model: mean fit error 0.844211 mA * Bp P K I=((0.6623814366m)+(-0.0002767670064m)*V(P,K))*uramp(V(P,K)+(-0.07623613033))^(1.545073678) ..ends X5U4 * 5V4 - IDH full-wave power rectifier (175mA) * Data from GE 5V4GA datasheet - traced on 09-09-01 by AF ..SUBCKT X5V4 P K * Child-Langmuir law: mean fit error 1.39637 mA * Bp P K I=(1.393901825e-3)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 1.20084 mA Bp P K I=(1.404180087e-3)*uramp(V(P,K)+(-0.1877759095))^(1.5) * Perugini model: mean fit error 0.948085 mA * Bp P K I=((1.324187528e-3)+(0.0007522446486e-3)*V(P,K))*uramp(V(P,K)+(0.1960102559))^(1.5054931 63) ..ends X5V4 * 5Y3 - DH full-wave power rectifier (125mA) * Data from GE 5Y3GT datasheet - traced on 09-09-01 by AF ..SUBCKT X5Y3GT P K * Child-Langmuir law: mean fit error 1.15597 mA * Bp P K I=(0.3533050537m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 0.958542 mA Bp P K I=(0.3560945008e-3)*uramp(V(P,K)+(-0.4420163009))^(1.5) * Perugini model: mean fit error 0.656689 mA * Bp P K I=((0.3169037014m)+(5.675621737e-05m)*V(P,K))*uramp(V(P,K)+(0.6363978284))^(1.51803 242) ..ends X5Y3GT * 274B - DH full-wave power rectifier (225mA) ..subckt X274B P K * Child-Langmuir law: mean fit error 1.17132 mA * Bp P K I=(0.3503318812m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 0.983373 mA Bp P K I=(0.3458749385e-3)*uramp(V(P,K)+(0.539276053))^(1.5) * Perugini model: mean fit error 0.393065 mA * Bp P K I=((0.368828156m)+(-0.0003153165025m)*V(P,K))*uramp(V(P,K)+(-0.6317505224))^(1.505633947) ..ends X274B * 6X5 - IDH full-wave power rectifier (70mA) ..subckt X6X5 P K * Child-Langmuir law: mean fit error 0.829063 mA * Bp P K I=(0.6876983643m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 0.666578 mA Bp P K I=(0.6965588691e-3)*uramp(V(P,K)+(-0.2713646197))^(1.5) * Perugini model: mean fit error 0.350721 mA * Bp P K I=((0.3658726552m)+(-0.0008687929442m)*V(P,K))*uramp(V(P,K)+(1.38423595 7))^(1.683868589) ..ends X6X5 * 6X4 - IDH full-wave miniature power rectifier (70mA) (this is essentially 6X5 in 7-pin base) ..subckt X6X4 P K * Child-Langmuir law: mean fit error 1.10646 mA * Bp P K I=(0.6787719727m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 0.603786 mA Bp P K I=(0.6955865558e-3)*uramp(V(P,K)+(-0.5138012527))^1.5 * Perugini model: mean fit error 0.372965 mA * Bp P K I=((0.4819803969m)+(-0.001701085314m)*V(P,K))*uramp(V(P,K)+(-0.06370839623))^(1.636154216) ..ends X6X4 * 6CA4 - IDH full-wave miniature power rectifier (150mA) ..subckt X6CA4 P K * Child-Langmuir law: mean fit error 0.585402 mA * Bp P K I=(1.672480011m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 0.352346 mA Bp P K I=(1.68085724e-3)*uramp(V(P,K)+(-0.1132675647))^(1.5) * Perugini model: mean fit error 0.349954 mA * Bp P K I=((1.673471733m)+(-0.0001489801373m)*V(P,K))*uramp(V(P,K)+(-0.1167055168))^(1.502190772) ..ends X6CA4 * EZ81 - IDH full-wave power rectifier (150mA) ..subckt XEZ81 P K * Child-Langmuir law: mean fit error 3.98557 mA * Bp P K I=(1.703564453m)*uramp(V(P,K))^(1.5) * Child-Langmuir law with contact potential: mean fit error 3.06551 mA Bp P K I=(1.616914135e-3)*uramp(V(P,K)+(0.8508860213))^(1.5) * Perugini model: mean fit error 1.36344 mA * Bp P K I=((3.069306816m)+(0.1206292787m)*V(P,K))*uramp(V( P,K)+(1.116202179))^(1.087711051) ..ends XEZ81 * EZ80 - IDH full-wave power rectifier (90mA) ..subckt XEZ80 P K * Child-Langmuir law: mean fit error 1.07583 mA * Bp P K I=(0.7265075684m)*uramp(V(P,K))^1.5 * Child-Langmuir law with contact potential: mean fit error 0.805197 mA Bp P K I=(0.7527537983e-3)*uramp(V(P,K)+(-0.553916594))^(1.5) * Perugini model: mean fit error 0.279904 mA * Bp P K I=((0.08550573077m)+(-0.000848630625m)*V(P,K))*uramp(V(P,K)+(3.586198141 ))^(2.164324675) ..ends XEZ80 * 5896 4v on 6.3 heater not valid over 30ma; with contact potential ..SUBCKT X5896-4hv P K Bp P K I=(1.259102866e-3)*uramp(V(P,K)+(0.6788214279))^(1.5) ..ends X5896-4hv * 6B10 diodes (common cathode) one diode section ..SUBCKT X6B10diode P K Bp P K I=(1.437120118e-3)*uramp(V(P,K)+(0.6846418964))^(1.5 Cpk P K 1.85P ..ends X6B10diode * 5903-13vh half heater voltage, crude plate current upper limit 8ma ..SUBCKT X5903-13vh P K Bp P K I=0.008-(uramp(0.008-((1.298177535/1.0e3)*uramp(V(P,K)+0.07-(0.09430728624))^(1.11)))) ..ends X5903-13vh * 5903-13vh CircuitMaker model ..subckt x5903vh13 P K Bp P K I=(0.538177535/1.0e3)*uramp(V(P,K)+(0.9430728624))^(1.5) ..ends x5903vh13 |
#28
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Hi Robert,
why different models for 5AR4 and GZ34? So far I thought this just to be the same tube within American & European naming systems. Tom -- Anyone who considers protocol unimportant has never dealt with a cat. - R. Heinlein |
#29
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote: On Tue, 13 Jun 2006 02:33:42 GMT, robert casey wrote: I managed to make a new sub class "tube diode" but, in keeping with all the other screwball things in circuitmaker, it won't let me name it vacdiode, because that exists (was hoping it would overwrite, substitute, or default to adding a model since none are in it), won't let me delete vacdiode, which it swears I must do in order to name mine that, and it won't let me add anything to the existing vacdiode nor will it macro copy anything to it either. So I don't know how you got yours into vacdiode.sub. Rename your old vacdiode.sub file to something else, and try using the below to create a new vacdiode.sub file snip Now that I've got things sorted out I went back to the pentode curve tracer to check some in triode mode and, assuming your tracer is accurate and I see no reason why it wouldn't be, there's no WONDER I had biasing problems when I went to build the 6V6 Plain Jane amp. The curves are way off, so I looked at the source and all the models of the 6V6, 6BQ5 and 6AQ5 I've got are exactly the same, apparently copies of each other except for the name, of course,... but the published plate curves ain't. 6GK6 - 6BQ5... ok (sorta, it's off but not nearly as much) but the 6AQ5-6V6 are quite different. 6BQ5/EL84 to 6V6 shouldn't be close and their difference is like EL34 and 6L6GC. The old data manuals I have don't seem to show a 6V6 = EL84. Anyway, about the only curves I really need for any tube is the Ra line for where Eg1 = 0V, and for pentodes and tetrodes these Ra lines and knees in curves can vary widely for screen voltages..... I have never placed any serious value on simulated testing of tube amps in the PC because I'd be wasting a lot of time and the real world breadboard tests and carefuly intelligent tweaking and measuring/listening is to me a shirtload more value than running airy fairy spice tests. Garbage in = garbage out, and so there is no substitute for real world testing with a hot soldering iron and away from the keyboard. Your comments about someone's tube curve tracer are amusing because anyone that can make a tube tracer which is accurate to within 3% of the real curves that exist is a true wizard indeed. You should assume that unless you have proven the tube tracer to be correct, it probably isn't. Never ever assume anything. The testers used in 1957 were not always very accurate either; bloody liars they were, actually. And the curves shown in manuals are not always right; maybe drawn to look pretty and interpolated by visual reckoning between a series of dots on a page and made to look pretty. Often the tubes can measure better than the curves indicate especially when power triodes are involved which may give less distortion that indicated when using a ruler during load line analysis. So entering curves from manuals into spice simulation programs could give a dodgy outcome especially with regard to distortions, where only slight line changes could be interpreted by the PC algorithms to produce very different relative harmonic spectra % compared to what ya get when ya solda somink and measure the bugger. Get real I say. Patrick Turner. |
#30
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Now that I've got things sorted out I went back to the pentode curve tracer to check some in triode mode and, assuming your tracer is accurate and I see no reason why it wouldn't be, there's no WONDER I had biasing problems when I went to build the 6V6 Plain Jane amp. The curves are way off, so I looked at the source and all the models of the 6V6, 6BQ5 and 6AQ5 I've got are exactly the same, apparently copies of each other except for the name, of course,... but the published plate curves ain't. 6GK6 - 6BQ5... ok (sorta, it's off but not nearly as much) but the 6AQ5-6V6 are quite different. The RCA tube manual says that, within its limitations, the 6AQ5 is the same as the 6V6. 6BQ5/EL84 to 6V6 shouldn't be close and their difference is like EL34 and 6L6GC. The old data manuals I have don't seem to show a 6V6 = EL84. It's 6V6 = 6AQ5 Anyway, about the only curves I really need for any tube is the Ra line for where Eg1 = 0V, and for pentodes and tetrodes these Ra lines and knees in curves can vary widely for screen voltages..... Be sure to set the screen voltage in my circuit to match that quoted in the published curves. I've made that error embarrassingly often... :-) I have never placed any serious value on simulated testing of tube amps in the PC because I'd be wasting a lot of time and the real world breadboard tests and carefuly intelligent tweaking and measuring/listening is to me a shirtload more value than running airy fairy spice tests. Garbage in = garbage out, and so there is no substitute for real world testing with a hot soldering iron and away from the keyboard. That's the next step after the simulation work is done. I wouldn't go into production only on simulations. Even at the big name semiconductor makers test "first silicon" on breadboards before going into production. Your comments about someone's tube curve tracer are amusing because anyone that can make a tube tracer which is accurate to within 3% of the real curves that exist is a true wizard indeed. You should assume that unless you have proven the tube tracer to be correct, it probably isn't. Never ever assume anything. The testers used in 1957 were not always very accurate either; bloody liars they were, actually. And the curves shown in manuals are not always right; maybe drawn to look pretty and interpolated by visual reckoning between a series of dots on a page and made to look pretty. Often the tubes can measure better than the curves indicate especially when power triodes are involved which may give less distortion that indicated when using a ruler during load line analysis. So entering curves from manuals into spice simulation programs could give a dodgy outcome especially with regard to distortions, where only slight line changes could be interpreted by the PC algorithms to produce very different relative harmonic spectra % compared to what ya get when ya solda somink and measure the bugger. Flipper is doing what one needs to do with simulations. You see something that looks crazy, and then you start digging in to find why. Presumably the published curves were extracted from several tests of nominal tubes. I've done at work characterization of various semiconductor IC chips, and you always do lots of chips. |
#31
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
I hope you read this right after the other because I worked it out. I decoded the .dat files and found where the user defined symbols and standard ones are listed and edited them back. Takes more than one place too, which is confusing as hell. Change it here and nothing appears to change till you change it in the other 3 or 4 places. sometimes twice in the same dern file. I vaguely remember agony of this sort with trying to create my own symbols and getting the models to go with them. So now I have two identical copies in the menu: vacdiode and tubediode. But I'll delete one of those out. The built in "Vac Diode" is also back and just as useless as it was to begin with. I don't know if it's needed as some kind of 'class' reference to something else or what but I don't think I'll try removing it again. Your default ckt still pulls that one up when the macro utilities are invoked but it doesn't really matter. I haven't messed with macros. I'll have to look to see what they do. I know some pieces of this SW pretty well, and other pieces not at all... :-) |
#32
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
The next step would be to breadboard it and see if the improvement or, conversely, added distortion (in the common anode feedback comparator case) 'mattered'. For example, in the SRPP case, tube variation mismatch might negate a lion's share of the benefit, or maybe not. Unfortunately, I don't have the equipment to do a 'real world' test of that nature, so that will have to be left to others (or trust Broskie), but the spice model indicates it might be worth looking at. I've used my PC's soundcard and CoolEdit software to make some distortion measurements. I've also used Cooledit to create various test tones that I burned to a CDR to play from a decent CD player. That then goes into the device under test, and the output of that via attenuation feeds the soundcard. Cooledit captures this audio, and then I can then use Cooledit's frequency spectrum plot. But first I connected the CD player directly to the soundcard to verify that it doesn't add its own distortions. Noise floor is about 95dB. 50dB harmonic levels show up nicely. There are limitations, of course (my soundcard samples at 44.1KHz, newer ones can do 96KHz and thus would allow you to see any ultrasonic crud). |
#33
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
I wasn't sure my sound cards were low enough distortion themselves, in either output or capture, to be of much use but if yours works well enough maybe mine would too. That's why I did a test without the circuit under test first. Your soundcard is probably fine for this, nothing special about my soundcard. Just a 16 bit job. I seem to be a bit short in the software audio tools department but I need to look around and take stock of what I've got as I may have something lurking in the archives. I just made up a 6AQ5 spice model, it is very different from a 6BQ5. I used the 5AQ5A designation (it's an audio tube made for series string TV sets) so I can recognize it as one I made up. It seems to compare well with the pentode and triode connected curves, except the slight tetrode kink isn't really modeled. *Vacuum Tube Tetrode (Audio freq.) ..SUBCKT X5AQ5A A S G K * * Calculate contribution to cathode current * *the number at the right end determines sharpness of knee Bat at 0 V=0.636*ATAN(V(A,K)/6) *the URAMP(V(S,K)/# mostly determines peak plate current, grid line spacing nearly constant, the sqrt and ^ adds some nonlinearity *the number at the right end determines slope of grid lines Bgs gs 0 V=URAMP(V(S,K)*0.19-(sqrt(V(G,K)*2.2)^(1.86))+V(A,K)/130) *the exponent sets the linearity of grid line spacing, and big impact on peak plate currrent Bgs2 gs2 0 V=V(gs)^(1.55) Bcath cc 0 V=V(gs2)*V(at) * * Calculate anode current, grid line spacing adjust and peak plate current * Ba A K I=0.26E-3*V(cc) * * Calculate screen current * Bscrn sc 0 V=V(gs2)*(1.1-V(at)) Bs S K I=0.46E-3*V(sc) * * Grid current (approximation - does not model low va/vs) * Bg G K I=(URAMP(V(G,K)+1)^(1.5))*50E-6 * * Capacitances * Cg1 G K 8p Cak A K 8p Cg1a G A 0.4p ..ENDS X5AQ5A Also can be found at the bottom of http://home.netcom.com/~wb2jia/tubes/spice.txt |
#34
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Now that I've got things sorted out I went back to the pentode curve tracer to check some in triode mode and, assuming your tracer is accurate and I see no reason why it wouldn't be, there's no WONDER I had biasing problems when I went to build the 6V6 Plain Jane amp. The curves are way off, so I looked at the source and all the models of the 6V6, 6BQ5 and 6AQ5 I've got are exactly the same, apparently copies of each other except for the name, of course,... but the published plate curves ain't. 6GK6 - 6BQ5... ok (sorta, it's off but not nearly as much) but the 6AQ5-6V6 are quite different. As noted elsewhere in this thread, I just created a 6AQ5 circuitmaker model. No magic software here, it was a lot of trial and error using that pentodecurv circuit and adjusting various parameters in the models to get it to look close to the published curves. I named it 5AQ5A, as a way to remind myself that I made it up, and a 5AQ5A is a series string TV tube identical to a 6AQ5 other than the heater. Find it at the bottom of my page: http://home.netcom.com/~wb2jia/tubes/spice.txt This model ought to simulate a lot closer to your 6V6 circuit than a 6BQ5 model would. A 6BQ5 model doesn't come anywhere close, no wonder it didn't work for you above. |
#35
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
With the problem finding circuitmaker models I thought I'd pull out Orcad again and see if I could make heads or tails of it. Well, I can made heads and tails but little else. I don't suppose you have any idea how to make that blasted thing run a simulation, do you? Sorry, never worked with Orcad. Though I remember it as just being a schematic capture tool. Though I've seen mention of it being a simulator. Maybe you need additional software for the simulations? |
#36
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Oh BROTHER!
One thing that we have to credit to Mr. McCoy is an active, shameless and limitless imagination. It reminds me in a perverse way of the attorney who died and went to the Pearly Gates to meet St. Peter. St. Peter went to his great book and, looked it over for a few moments and said: I see by the records that you are 143 years old. The attorney thought for a moment and said: Huh? I dropped dead at 47 with a massive heart-attack. Whereupon St. Peter replied: I only added up your billable hours. Mr. McCoy will face similar questions when his many, varied and oh-so-world-changing careers are added together. Peter Wieck Wyncote, PA |
#37
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
Follow up. Since I had some luck with the Orcad model I decided to try converting it to circuitmaker, because I ran across some 'tips'. Like most 'tips' they ended up being a bit short but, anyway, I do have something that at least runs. Maybe you'd like a copy and maybe you could double check the conversion. Duncan's, the author of the model, tips were... snip Thanks, I saved your post to a file. I'll try to convert some other PSpice models. Your conversion looks pretty close to the 5AQ5 model I made up. I can see someone using your 6V6 model and my 5AQ5A model in a P-P amp simulation to see what happens with a pair of slightly mismatched tubes. Harmonic distortions, bias mismatching if both share a cathode resistor or the grids are fed a minus bias voltage. |
#38
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
On Thu, 15 Jun 2006 20:16:43 -0500, flipper wrote:
On Thu, 15 Jun 2006 19:16:41 GMT, robert casey wrote: With the problem finding circuitmaker models I thought I'd pull out Orcad again and see if I could make heads or tails of it. Well, I can made heads and tails but little else. I don't suppose you have any idea how to make that blasted thing run a simulation, do you? Sorry, never worked with Orcad. Though I remember it as just being a schematic capture tool. Though I've seen mention of it being a simulator. Maybe you need additional software for the simulations? I've got the 'Unison Suit', albeit an old version. I was particularly frustrated when I posted as I *had* been able to make it at least do 'something' before and I still don't know what's so screwed up with the test I was trying at the time but I sorted it out in general later on and even managed to get a 6V6 model to work in it. Mainly by finding something that 'worked' and using that as a guide to expand from. Orcad makes Circuitmaker's obtuseness seem crystal clear by comparison Which is one reason why I ended up using circuitmaker so much. I was more interested in getting something done than in becoming an Orcad guru. Circuitmaker is simpler to use for small projects and has a much better simulation interface but it's harder to find models for and schematic capture has a nasty tendency to get confused and crash. I'll probably stick with Circuitmaker when I've got good models for it. Have you tried LTSpice? Get it off www.linear.com for free. It has a very big following and is continuously updated. - YD. -- Remove HAT if replying by mail. |
#39
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
flipper wrote:
Oh, wait, I just ran another curve on the 5aq5 with grid 15 to 0, step -5 and I get 4 curves. Which is the same as I get with 0 to -15, step -5, or 15 to -15, step 5. I'll bet it's reflecting around 0, rather than just 'stopping', because the curve is there.. Or in math terms, it appears there's something making it an absolute value. I found what I was doing to cause that error. Looks like when I took V(G,K) to the 1.8something power and then the square root of that it also did an absolute value as well... Took that out (I was trying to impose some nonlinearity but that's the wrong place for that).. Try the below, it gives reasonable positive grid curves except for less than about 100V on the plate in tridoe mode: *Vacuum Tube Tetrode (Audio freq.) ..SUBCKT X5AQ5A A S G K * * Calculate contribution to cathode current * *the number at the right end determines sharpness of knee Bat at 0 V=0.636*ATAN(V(A,K)/6) *the URAMP(V(S,K)/# mostly determines peak plate current, grid line spacing nearly constant *the number at the right end determines slope of grid lines Bgs gs 0 V=URAMP(V(S,K)*0.19+V(G,K)*1.73+V(A,K)/130) *the exponent sets the linearity of grid line spacing, and big impact on peak plate currrent Bgs2 gs2 0 V=V(gs)^(1.5) Bcath cc 0 V=V(gs2)*V(at) * * Calculate anode current, grid line spacing adjust and peak plate current * Ba A K I=0.32E-3*V(cc) * * Calculate screen current * Bscrn sc 0 V=V(gs2)*(1.1-V(at)) Bs S K I=0.55E-3*V(sc) * * Grid current (approximation - does not model low va/vs) * Bg G K I=(URAMP(V(G,K)+1)^(1.5))*50E-6 * * Capacitances * Cg1 G K 8p Cak A K 8p Cg1a G A 0.4p ..ENDS X5AQ5A |
#40
Posted to rec.audio.tubes
|
|||
|
|||
Tubes and Transistors
On Sat, 17 Jun 2006 06:28:36 -0500, flipper wrote:
On Sat, 17 Jun 2006 01:36:21 -0300, YD wrote: On Thu, 15 Jun 2006 20:16:43 -0500, flipper wrote: On Thu, 15 Jun 2006 19:16:41 GMT, robert casey wrote: With the problem finding circuitmaker models I thought I'd pull out Orcad again and see if I could make heads or tails of it. Well, I can made heads and tails but little else. I don't suppose you have any idea how to make that blasted thing run a simulation, do you? Sorry, never worked with Orcad. Though I remember it as just being a schematic capture tool. Though I've seen mention of it being a simulator. Maybe you need additional software for the simulations? I've got the 'Unison Suit', albeit an old version. I was particularly frustrated when I posted as I *had* been able to make it at least do 'something' before and I still don't know what's so screwed up with the test I was trying at the time but I sorted it out in general later on and even managed to get a 6V6 model to work in it. Mainly by finding something that 'worked' and using that as a guide to expand from. Orcad makes Circuitmaker's obtuseness seem crystal clear by comparison Which is one reason why I ended up using circuitmaker so much. I was more interested in getting something done than in becoming an Orcad guru. Circuitmaker is simpler to use for small projects and has a much better simulation interface but it's harder to find models for and schematic capture has a nasty tendency to get confused and crash. I'll probably stick with Circuitmaker when I've got good models for it. Have you tried LTSpice? Get it off www.linear.com for free. It has a very big following and is continuously updated. Hadn't tried it but I downloaded it and will take a look. Thanks. Judging by your gripes about Circuitmaker you might find it friendlier. And a friendly large user base at Yahoo Groups. - YD. -- Remove HAT if replying by mail. |
Reply |
|
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Vacuum tubes vs. transistors | Vacuum Tubes | |||
Tubes V's Semiconductors in Military. | Vacuum Tubes | |||
Some tube history about 6L6. | Vacuum Tubes | |||
Announcement: Building Homebrew Vacuum Tubes and Homebrew Transistors | Vacuum Tubes |