| Home |
| Search |
| Today's Posts |
|
#1
Posted to rec.audio.tubes
|
|||
|
|||
Pentode gm wired as a triode
Is there a simple relationship between the gm of a pentode wired as a
pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Cheers Ian |
|
#2
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
|
|
#3
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Another Ian asked;
Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Given constant voltage(s) on screen and anode, gm should be the same. The pentode gm is related to screen voltage; the triode to anode voltage. As an obvious example, if the screen voltage is set to be the same as anode voltage, then it won't make any difference whether the screen is connected to the anode or not. You don't say what the screen voltage was for the pentode test. gm measurement assumes constant screen and anode voltages, so the fact that you have not observed this requirement between your two tests is likely to be significant, wouldn't you think? Your test voltages are quite low, and possibly quite close to the knee of the pentode characteristic curves. Screen and anode voltage differences between your two tests will account for the different values of gm, assuming the tester maintains constant voltages. AVO testers work in mysterious ways. It may be easier to think in terms of mu. The mu of a triode-connected pentode is the screen mu of the pentode, not the anode mu. To a reasonable approximation. cheers, Ian |
|
#4
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
...
Your test voltages are quite low, and possibly quite close to the knee of the pentode characteristic curves... I didn't say why this is significant. At low voltages, screen current may be a large part of total current, and may vary sharply with changes in screen or anode voltage. If you were to measure cathode current, or the sum of screen and anode current, then your readings may be more similar. Ian "Ian Iveson" wrote in message k... Another Ian asked; Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Given constant voltage(s) on screen and anode, gm should be the same. The pentode gm is related to screen voltage; the triode to anode voltage. As an obvious example, if the screen voltage is set to be the same as anode voltage, then it won't make any difference whether the screen is connected to the anode or not. You don't say what the screen voltage was for the pentode test. gm measurement assumes constant screen and anode voltages, so the fact that you have not observed this requirement between your two tests is likely to be significant, wouldn't you think? Screen and anode voltage differences between your two tests will account for the different values of gm, assuming the tester maintains constant voltages. AVO testers work in mysterious ways. It may be easier to think in terms of mu. The mu of a triode-connected pentode is the screen mu of the pentode, not the anode mu. To a reasonable approximation. cheers, Ian |
|
#6
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote:
Another Ian asked; Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Given constant voltage(s) on screen and anode, gm should be the same. The pentode gm is related to screen voltage; the triode to anode voltage. As an obvious example, if the screen voltage is set to be the same as anode voltage, then it won't make any difference whether the screen is connected to the anode or not. You don't say what the screen voltage was for the pentode test. The AVO data sheets tell me to set plate to 100V and screen to 100V. gm measurement assumes constant screen and anode voltages, so the fact that you have not observed this requirement between your two tests is likely to be significant, wouldn't you think? Probably. I did do it first at plate connected to screen connected to 100V and got even larger gm readings. I am beginning to think I am using my AVO two panel tester in ways it was never intended to. Ian Your test voltages are quite low, and possibly quite close to the knee of the pentode characteristic curves. Screen and anode voltage differences between your two tests will account for the different values of gm, assuming the tester maintains constant voltages. AVO testers work in mysterious ways. It may be easier to think in terms of mu. The mu of a triode-connected pentode is the screen mu of the pentode, not the anode mu. To a reasonable approximation. cheers, Ian |
|
#7
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote:
... Your test voltages are quite low, and possibly quite close to the knee of the pentode characteristic curves... I didn't say why this is significant. At low voltages, screen current may be a large part of total current, and may vary sharply with changes in screen or anode voltage. If you were to measure cathode current, or the sum of screen and anode current, then your readings may be more similar. Ian The odd thing is the AVO data sheets recommend 100V for both plate and screen to measure pentode gm for a 6AU6. Strangely for an EF94 (supposedly identical) it says to use 100V plate and 150V screen. Even with 100V plate and screen as for 6AU6, my EF94 tubes measure consistently higher gms than do 6AU6. IAn |
|
#8
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
wrote: Patrick Turner wrote: wrote: Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Cheers Ian Gm can be tested with anode and screen taken to say a fixed 100V, and the grid biased for a given Ia of whatever amount you wish up to say 10mA, (Pda = 1watt). Suppressor can be grounded for the test. If you have a 10 ohm R between the B+ supply and anode etc, and apply a small signal, say 0.1 Vrms to the grid at say 400Hz, then you can measure the current in the 10 ohm resistor. Where gm = 4mA/V, you should see that 1Vrms applied to the grid produces 4mA of I change in the 10 ohms, seen as 40mV across the 10 ohms. Lowering the anode current by changing the grid bias to more -ve will give less voltage at the 10 ohms and by careful measurement you can draw a graph for gm vs Ia for B+ = Ea = 100V. The test above is the same for triode and pentode. Does connecting the suppressor to the B+ make any difference? The other way is to measure the voltage gain with a given known Ia and say 20k RL and 40k RL in both pentode and triode connection. The change from pentode to triode and with both RL values must be done with Ea, Ia and Eg2 all that the same voltage value for each test. A = µ x RL / ( RL + Ra ) With two equations for gain with two known RL, you can work out the two unknowns, Ra and µ, then apply gm = µ / Ra. When you have done all this on your breadboard come back and tell us what you found. I Shall. I just wish the bloody parts I need would arrive. Maplin are out of mains transformers so I have had to order them from elsewhere and the B7G sockets and tag board have still not arrived. I have the piece of wood cut out though all ready to go ;-) Ian Patrick Turner. About 13 years ago when I started all my investigations to teach myself how things worked I scrounged all this junk nobody wanted, like partially wrecked tape recorders, radios, crummy amps and these became the test beds for many an idea. I bought a cheap second hand CRO, and build test gear with nothing but a book or two and Wireless World articles to go by. I acumulated so much junk.....and I was never short of stuff to build a rig to test something. I doubt I could repeat it now because the world has become cleaned of its useful junk. Patrick Turner. |
|
#9
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
|
|
#10
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick Turner wrote:
wrote: Patrick Turner wrote: wrote: Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Cheers Ian Gm can be tested with anode and screen taken to say a fixed 100V, and the grid biased for a given Ia of whatever amount you wish up to say 10mA, (Pda = 1watt). Suppressor can be grounded for the test. If you have a 10 ohm R between the B+ supply and anode etc, and apply a small signal, say 0.1 Vrms to the grid at say 400Hz, then you can measure the current in the 10 ohm resistor. Where gm = 4mA/V, you should see that 1Vrms applied to the grid produces 4mA of I change in the 10 ohms, seen as 40mV across the 10 ohms. Lowering the anode current by changing the grid bias to more -ve will give less voltage at the 10 ohms and by careful measurement you can draw a graph for gm vs Ia for B+ = Ea = 100V. The test above is the same for triode and pentode. Does connecting the suppressor to the B+ make any difference? The other way is to measure the voltage gain with a given known Ia and say 20k RL and 40k RL in both pentode and triode connection. The change from pentode to triode and with both RL values must be done with Ea, Ia and Eg2 all that the same voltage value for each test. A = µ x RL / ( RL + Ra ) With two equations for gain with two known RL, you can work out the two unknowns, Ra and µ, then apply gm = µ / Ra. When you have done all this on your breadboard come back and tell us what you found. I Shall. I just wish the bloody parts I need would arrive. Maplin are out of mains transformers so I have had to order them from elsewhere and the B7G sockets and tag board have still not arrived. I have the piece of wood cut out though all ready to go ;-) Ian Patrick Turner. About 13 years ago when I started all my investigations to teach myself how things worked I scrounged all this junk nobody wanted, like partially wrecked tape recorders, radios, crummy amps and these became the test beds for many an idea. I bought a cheap second hand CRO, and build test gear with nothing but a book or two and Wireless World articles to go by. I acumulated so much junk.....and I was never short of stuff to build a rig to test something. I doubt I could repeat it now because the world has become cleaned of its useful junk. Patrick Turner. When I started building valve gear over 40 years ago I did much the same thing. I joined the local radio club and the secretary, who had a TV repair business, said he had some old stuff to get me started. I went round his house and in his back garden there was a pile a good 10 feet tall of old radio and TV chassis. He said I could help myself. The only thing that limited me was how much I could carry on my push bike. Ian |
|
#11
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
The odd thing is the AVO data sheets recommend 100V for both plate
and screen to measure pentode gm for a 6AU6. Strangely for an EF94 (supposedly identical) it says to use 100V plate and 150V screen. Even with 100V plate and screen as for 6AU6, my EF94 tubes measure consistently higher gms than do 6AU6. http://www.mif.pg.gda.pl/homepages/f...93/6/6AU6A.pdf Compare characteristic curves at bottom of page 3, for pentode connection with Vs=150V, against the triode curves at bottom of page 4. Firstly, notice that the screen current for this pentode is generally quite a high proportion of total current. I wonder if this is typical for sharp cut-off pentodes. Second, notice that Va = 80V takes you close to the knee, where anode current is beginning to fall sharply and screen current is rising correspondingly quickly, with respect to falling anode voltage. Looking at the pentode curves, for Va = Vs = 150V, a change from -2V to -3V on the grid results in an anode current change of about 2.8mA, and a screen current change of about 1.2mA. Looking at the triode curves, for Va = Vs = 150V, a change from -2V to -3V on the grid gives a change in anode + screen current of about 4mA. A tester measuring gm for a valve with those characteristics and using those grid voltages would show 2.8mA/V for the pentode and 4mA/V for the triode. The total current change for both is the same, since 1.2 + 2.8 = 4. Your tests give much greater values, at lower voltages. Anyway, you can see how the difference between the pentode and triode arises, considering your change of voltages and the high screen current. AVO valve data are derived from a combination of calculation, fudging, guesswork, and feedback from outraged customers. AFAIK, valve datasheets were compiled using a similar method, with a little wishful thinking thrown in for good measure. cheers, Ian |
|
#12
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick said:
And BTW, when you triode connect the 6AU6, you shoudn't see a higher gm than pentode, ever. Flipper responded: Are you sure about that? Of course. Patrick is always sure. Seems to me that since screen current is lost in pentode mode but not in triode mode that gm would tend to be correspondingly higher in triode mode. Nearly. It is the loss of *variation* in screen current ( ie screen gm) that makes the difference though. See my last post. Ian |
|
#13
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote: Patrick said: And BTW, when you triode connect the 6AU6, you shoudn't see a higher gm than pentode, ever. Flipper responded: Are you sure about that? Of course. Patrick is always sure. I have never made any measurements to indicate gm for a beam/pentode tube in triode is hugely higher than gm in beam/pentode. Testing with screen and anode tied together to the same B+ would reveal equal gm as would testing anode current measured separately for triode. When you lazy guys do your measurements for µ, Ra and gm using the two load gain test, instead of relying on a tube tester then you can say something that shows you understand. Seems to me that since screen current is lost in pentode mode but not in triode mode that gm would tend to be correspondingly higher in triode mode. Yes indeed G2 current is lost in pentode, and is usually up to 25% of anode current. But getting 7mA/V in triode compared t 3.9mA/V is plain wrong. That's the point I was making. How could anyone accept that? If it is unacceptable, why? what to do to find out why? the mind should race into question mode and stay there till answered. It does mean late nights and sundays away from the PC.... Patrick Turner. Nearly. It is the loss of *variation* in screen current ( ie screen gm) that makes the difference though. See my last post. Ian |
|
#14
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick Turner wrote:
wrote: Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Cheers Ian And BTW, when you triode connect the 6AU6, you shoudn't see a higher gm than pentode, ever. Not only that, none of the data books which state the gm for triode and pentode connection of the same tube will show higher triode gm than pentode. I suggest there is a mistake in your tester settings, or some other problem. Patrick Turner. I thought the result was rather suspicious. The most likely explanation is the tester/settings. Ian |
|
#15
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote:
snip AVO valve data are derived from a combination of calculation, fudging, guesswork, and feedback from outraged customers. AFAIK, valve datasheets were compiled using a similar method, with a little wishful thinking thrown in for good measure. cheers, Ian LOL Ian |
|
#16
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick Turner wrote:
snip It does mean late nights and sundays away from the PC.... Yippee!! Ian |
|
#17
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick said:
And BTW, when you triode connect the 6AU6, you shoudn't see a higher gm than pentode, ever. Flipper responded: Are you sure about that? Of course. Patrick is always sure. I have never made any measurements to indicate gm for a beam/pentode tube in triode is hugely higher than gm in beam/pentode. You have never made any measurements with a 6AU6. You seem not to have made measurements with pentodes at relatively low voltages. And yet you are still so sure. Testing with screen and anode tied together to the same B+ would reveal equal gm as would testing anode current measured separately for triode. Meaningless obfuscation. When you lazy guys do your measurements for µ, Ra and gm using the two load gain test, instead of relying on a tube tester then you can say something that shows you understand. Cheeky boy. I'm not interested in showing I understand, and I can't see why Flipper would be either. You're the one with a desperate desire to flaunt your pathos. Seems to me that since screen current is lost in pentode mode but not in triode mode that gm would tend to be correspondingly higher in triode mode. Yes indeed G2 current is lost in pentode, and is usually up to 25% of anode current. The 6AU6 is what we are talking about. Strut and squirm all you like, you haven't measured one, or even looked at the datasheet. But getting 7mA/V in triode compared t 3.9mA/V is plain wrong. That's the point I was making. Liar. How could anyone accept that? If it is unacceptable, why? what to do to find out why? the mind should race into question mode and stay there till answered. What would you know about the mind. It does mean late nights and sundays away from the PC.... You will never learn as long as you think you already know. Ian |
|
#18
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
But getting 7mA/V in triode compared t 3.9mA/V is plain wrong. That's the point I was making. Liar. How could anyone accept that? If it is unacceptable, why? what to do to find out why? the mind should race into question mode and stay there till answered. What would you know about the mind. It does mean late nights and sundays away from the PC.... You will never learn as long as you think you already know. Ian OK, the text books and data tell us that triode gm for the same Ia is slightly greater than pentode and I was wrong. However, one rarely ever uses a trioded pentode at high Ia and gm measures low and NOT very much more than what we might use with pentode. To take advantage of the extra gain possible for the pentode we may have Ia higher, so gm then would equal the triode case of be greater than the triode case. In all my tests the gm is rarely much greater than pentode, and wherever possible I try to use triodes in signal stages with as little current change as possible, ie as pure voltage gain devices for the lowest THD, where gm does not matter much at all. Pentode gain as you should know roughly = gm x RL, and gm is more important to us. We need the pentode gain to be able to apply the NFB to correct the THD and reduce the Rout, and to nullify variations in gm between samples of pentodes. The trioded pentode has its NFB applied from anode to the screen, and IS thus a triode, but of course saying this won't be gladly agreed with by our resident dopey Iveson who has continually denied there could be FB in a triode, without ever proving why there isn't, despite the far brighter minds who say there is FB there. The greater the gm, the higher the gain of the pentode, but not for a triode, because of its NFB. Anyway, the gm varies somewhat a lot with Ia, and for a given Ia it can be taken to be roughly equal for triode and pentode as I said, and the difference of 7 to 3.9 isn't what you might measure with a 6AU6 in either triode or pentode respectively. I have never ever measured or calculated such a large difference difference in 6AU6 triode or pentode gm. If you think I need more correcting, then please carry out the gain tests I suggested, lest everyone here think you are a nit picking arsole. Patrick Turner. |
|
#19
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick Turner wrote:
The trioded pentode has its NFB applied from anode to the screen, and IS thus a triode, but of course saying this won't be gladly agreed with by our resident dopey Iveson who has continually denied there could be FB in a triode, without ever proving why there isn't, despite the far brighter minds who say there is FB there. Your "thus" is awkwardly placed on the brink of topple. I hoped we might avoid another "internal feedback" skirmish, considering that it would be unlikely to assist the OP. Easier to think like the triode connection simply moves the anode to the position of the screen. Something that has been pointed out to you many times is that it should not be necessary to prove that the non-existent does not exist. That just puts your argument on the level of "conspiracy theory", flying saucers, ghosts and magic. Some time ago I stated my position on the matter of "internal feedback" clearly and definitively and I don't think I've changed my view much since then. I repeat the post below. Incidentally, on the issue of real feedback, I made another statement that you scoffed at: "If a linear closed loop system has a transfer function in its feedback path, then the effect of that function is the same as if it were placed in the forward loop, and its inverse were placed in the forward path ahead of the summing point." Which you may like to consider more humbly, because it is useful in general, and because in particular it may help you, as you continue to savage that poor long-suffering VT100, not to through out the baby with the bathwater. Me, 08/01/06, he "Some say there is feedback in triodes. I believe this is not true. "Our craft of building valve amplifiers is supported by a body of knowledge contained is such works as RDH4. These works bring together several elements: definitions and parameters of components, specification of system requirements, etc., and the specific application of engineering analysis to the problem of connecting components together in such a way that the resulting system meets those requirements. "All these works use a common set of analytical methods. In general these methods are themselves not derived or explained in much detail: the emphasis is on the application. RDH4 may walk you through a particular calculation, but it will rarely explain why it uses one method rather than another. Terms such as "bandwidth", "transient response", "gain" and hundreds of others may be illustrated in terms of amplifier design, but you will rarely find an explanation of why these particular concepts are used and whether there may be alternatives, or how they fit in with the world in general. "That set of concepts implicit in our methods of analysis, and the terminology used for its expression, is in turn supported by the generic discipline of engineering. Those who have studied engineering will know that the general concepts are applicable to all systems, whether they be mechanical, electrical, or whatever. "The field of engineering underpinning works such as RDH4 is generally called "control systems engineering". This is the place to look if you wish, for example, to find the derivation and general significance of terms like "marginal stability", "bode analysis" and just about every other concept used in amplifier design. Hence I have previously urged those who wish for a deeper and more general knowledge of design, and in particular those who wish to transfer and apply existing knowledge of mechanical systems (such as car design), might find an introductory book on control systems engineering helpful. "Here are a few chapter headings from "Feedback and Control Systems" (Di Stefano III et al, McGraw Hill, 1967): Control systems terminology; Linear systems and differential equations; The Laplace transform; stability; block diagram algebra and transfer functions of systems; Signal Flow Graphs; Nyquist design, Bode design. "Any such work will contain, within the first three chapters, a definition of "feedback". Every one will contain the same definition and use the same basic system diagram to illustrate it. They will all make it clear that the definition is absolute and crucial to the meaning of all that follows. Get the definition or the diagram wrong, and you are lost without hope. "Engineers will know that once you have grasped the theory, you realise that is the simple part. The real challenge is to frame a particular problem in such a way that the theory can be applied. Once you have done this a few times, and you settle into a particular area of engineering, you get a book such as RDH4 which saves you the trouble. "The "canonical" diagram for an open-loop system (ie one without feedback) comprises an input signal, a transfer function, and an output signal. "The "canonical" diagram for a closed-loop system adds to the open-loop diagram a feedback path and a summing node. The feedback path *must* be from the output to the summing node. The input must also be directly to the summing node. If the summing is positive it is a positive feedback system, otherwise it is a negative feedback system. There may or may not be a transfer function in the feedback path. "And that is that. There is absolutely no room for manoevre. All of those concepts, and all of those methods of analysis, depend on that definition. That is the sense in which the word "feedback" is used in RDH4, and every other sensible exposition or discussion of amplifier design. "In its usual mode, the control input to a triode is a voltage applied between the grid and cathode, Vgk. The output is a voltage between anode and cathode, Vak. There is a possible feedback path via the interelectrode capacitance, Cag, often called the Miller capacitance. However, there is no summing node, and therefore no feedback. A change in Vak will not be summed with Vgk in the absence of an external circuit. "That external circuit must at least provide a summing node to enable the Miller effect. A resistor in series with the grid is usually used to achieve this. Commonly other feedback paths are added for the purpose of adding bandwidth, reducing distortion, and adjusting input and output impedance. In each case a summing node is required, although a single node may serve several paths. "If a feedback path is connected to a current sensor at the system output, it is "series derived". If from a voltage sensor it is "parallel derived". If the summation is of voltage, it is "parallel applied". If of current it is "series applied". Any of the four possible permutations are allowable, and are likely to have different effects. "There is no such thing as "internal feedback", and there is no feedback in triodes. "There are other more vague meanings of "feedback" (e.g. "audience feedback"), but they need not concern us. Certainly they should not be confused with ours. "Finally, it is possible by contrivance and notional reconfiguration of inputs and outputs to get *anything* to fit the canonical diagram. This is true not only of triodes but also of resistors, ducks, the colour blue and the man in the moon. In these cases the feedback is a function of the contrivance, and not a property of the duck or the triode. No real property applies equally to all things, obviously. All cows are black at night. "I am serious about looking at a book, honestly. It would be a revelation for anyone able to follow the maths." Ian |
|
#20
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote: Patrick Turner wrote: The trioded pentode has its NFB applied from anode to the screen, and IS thus a triode, but of course saying this won't be gladly agreed with by our resident dopey Iveson who has continually denied there could be FB in a triode, without ever proving why there isn't, despite the far brighter minds who say there is FB there. Your "thus" is awkwardly placed on the brink of topple. I hoped we might avoid another "internal feedback" skirmish, considering that it would be unlikely to assist the OP. Easier to think like the triode connection simply moves the anode to the position of the screen. Something that has been pointed out to you many times is that it should not be necessary to prove that the non-existent does not exist. That just puts your argument on the level of "conspiracy theory", flying saucers, ghosts and magic. Some time ago I stated my position on the matter of "internal feedback" clearly and definitively and I don't think I've changed my view much since then. I repeat the post below. Incidentally, on the issue of real feedback, I made another statement that you scoffed at: "If a linear closed loop system has a transfer function in its feedback path, then the effect of that function is the same as if it were placed in the forward loop, and its inverse were placed in the forward path ahead of the summing point." Which you may like to consider more humbly, because it is useful in general, and because in particular it may help you, as you continue to savage that poor long-suffering VT100, not to through out the baby with the bathwater. Is there anyone else in the group who understands what Ian just said?????? BTW, the ARC VT100 will sing beautifully when I have finished with my surgery on its vocal chords. It will also have 8 adjust pots for each output tube to be biased separatly to enable the full benefits of the fixed bias to be realised. The bias adjustments will be easy, with a screw driver and cheap DVM access on the top of the amp, as it should have been done by ARC in the first place. There will be two red LED facing forward on the front panel to show if one or more tubes conducts too much Ik, ie, Ik rises by 50% from 50mA to 80mA for longer than 4 seconds. There will be an auxilliary transfromer to mamange how the amp turns on and to shut it down if Ik on any or all of the output tubes becomes excessive. After my work, this amp will never make smoke again, have lower THD/IMD, lower Rout, be more sensitive, and yet have only 10dB of global NFB as originally intended. There will be NO dc adust pots for the CCS in the input stages, and no fragile j-fets for CCS; just ONE easy to buy reliable MJE340. 6922/6DJ8 will not be used in the balanced amp and buffer CF stage which fail badly. Rugged Oz made NOS 6CG7 will be used instead, because they won't fail, and are more linear in the simpler revised schematic. The trouble with many high end companies is that they tend to use 49 parts where 29 are fine, and then they try to tell us its all really good stuff, and charge a fortune for it, but really its hopelessly unreliable, and servicing is a pain. Ian wants to tell us I am a butcher, but for him to think the major names in tube audio could not improve their product is sheer stupidity. I am always on thre side of the music, the reliability, and the customer's peace of mind. Below is a repeated pile of hot air from Ian, and not a syllable about why there IS NFB in triodes. Ian could have spent the effort proving NO FB exists in a triode, and he'd never have had to resort to the BS below, which I can promise everyone is a very tiring read about many things but nothing on the kernel of the matter, NFB in triodes. Patrick Turner. Me, 08/01/06, he "Some say there is feedback in triodes. I believe this is not true. "Our craft of building valve amplifiers is supported by a body of knowledge contained is such works as RDH4. These works bring together several elements: definitions and parameters of components, specification of system requirements, etc., and the specific application of engineering analysis to the problem of connecting components together in such a way that the resulting system meets those requirements. "All these works use a common set of analytical methods. In general these methods are themselves not derived or explained in much detail: the emphasis is on the application. RDH4 may walk you through a particular calculation, but it will rarely explain why it uses one method rather than another. Terms such as "bandwidth", "transient response", "gain" and hundreds of others may be illustrated in terms of amplifier design, but you will rarely find an explanation of why these particular concepts are used and whether there may be alternatives, or how they fit in with the world in general. "That set of concepts implicit in our methods of analysis, and the terminology used for its expression, is in turn supported by the generic discipline of engineering. Those who have studied engineering will know that the general concepts are applicable to all systems, whether they be mechanical, electrical, or whatever. "The field of engineering underpinning works such as RDH4 is generally called "control systems engineering". This is the place to look if you wish, for example, to find the derivation and general significance of terms like "marginal stability", "bode analysis" and just about every other concept used in amplifier design. Hence I have previously urged those who wish for a deeper and more general knowledge of design, and in particular those who wish to transfer and apply existing knowledge of mechanical systems (such as car design), might find an introductory book on control systems engineering helpful. "Here are a few chapter headings from "Feedback and Control Systems" (Di Stefano III et al, McGraw Hill, 1967): Control systems terminology; Linear systems and differential equations; The Laplace transform; stability; block diagram algebra and transfer functions of systems; Signal Flow Graphs; Nyquist design, Bode design. "Any such work will contain, within the first three chapters, a definition of "feedback". Every one will contain the same definition and use the same basic system diagram to illustrate it. They will all make it clear that the definition is absolute and crucial to the meaning of all that follows. Get the definition or the diagram wrong, and you are lost without hope. "Engineers will know that once you have grasped the theory, you realise that is the simple part. The real challenge is to frame a particular problem in such a way that the theory can be applied. Once you have done this a few times, and you settle into a particular area of engineering, you get a book such as RDH4 which saves you the trouble. "The "canonical" diagram for an open-loop system (ie one without feedback) comprises an input signal, a transfer function, and an output signal. "The "canonical" diagram for a closed-loop system adds to the open-loop diagram a feedback path and a summing node. The feedback path *must* be from the output to the summing node. The input must also be directly to the summing node. If the summing is positive it is a positive feedback system, otherwise it is a negative feedback system. There may or may not be a transfer function in the feedback path. "And that is that. There is absolutely no room for manoevre. All of those concepts, and all of those methods of analysis, depend on that definition. That is the sense in which the word "feedback" is used in RDH4, and every other sensible exposition or discussion of amplifier design. "In its usual mode, the control input to a triode is a voltage applied between the grid and cathode, Vgk. The output is a voltage between anode and cathode, Vak. There is a possible feedback path via the interelectrode capacitance, Cag, often called the Miller capacitance. However, there is no summing node, and therefore no feedback. A change in Vak will not be summed with Vgk in the absence of an external circuit. "That external circuit must at least provide a summing node to enable the Miller effect. A resistor in series with the grid is usually used to achieve this. Commonly other feedback paths are added for the purpose of adding bandwidth, reducing distortion, and adjusting input and output impedance. In each case a summing node is required, although a single node may serve several paths. "If a feedback path is connected to a current sensor at the system output, it is "series derived". If from a voltage sensor it is "parallel derived". If the summation is of voltage, it is "parallel applied". If of current it is "series applied". Any of the four possible permutations are allowable, and are likely to have different effects. "There is no such thing as "internal feedback", and there is no feedback in triodes. "There are other more vague meanings of "feedback" (e.g. "audience feedback"), but they need not concern us. Certainly they should not be confused with ours. "Finally, it is possible by contrivance and notional reconfiguration of inputs and outputs to get *anything* to fit the canonical diagram. This is true not only of triodes but also of resistors, ducks, the colour blue and the man in the moon. In these cases the feedback is a function of the contrivance, and not a property of the duck or the triode. No real property applies equally to all things, obviously. All cows are black at night. "I am serious about looking at a book, honestly. It would be a revelation for anyone able to follow the maths." Ian |
|
#21
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
The inputs to a tube are voltages and the output is current. This is why it's called a transconductance device. But there does not have to be any current change within a triode. So it is called a voltage device. A large voltage change occurs because a smaller one controls voltages. For a triode the first voltage input is the grid, the second voltage input is the plate, the summing junction is the cathode, I don't see the cathode as THE summing junction. But there is a resultant field action on the space charge between cathode and grid which results from anode voltage and grid voltage which do form a summing junction, and the sum of the fields controls the electron flow, or the the charges themselves at the anode if no I change occurs. and one can find the transfer characteristics for each input in the tube's datasheet with the first, and most popular, being a set of curves showing the effect of the plate input voltage on output current when the grid input voltage is held constant with the second being a set of curves showing the effect of grid input voltage on output current when the plate input voltage is held constant. The curves for plate characteristics are like the diode Ra curve. If we then run the current output through a resistor so that it generates a voltage signal in phase opposition to the grid, and if that phase opposite signal, representing a portion of the output, is then applied to the plate input, it sums at the cathode negatively with the grid signal and, so, is referred to as negative feedback. ipso facto, case closed. Hmm, I doubt you've said anywhere enough to convince Ian of anything. His problem is that he's set his mind against NFB in triodes, and he can't accept NFB now. To do that would be to admit defeat, and we could all laugh about why it took so long for him to "get it". He hates being the laughing stock. He hates being wrong. He's only happy when other ppl are wrong, so he can dance on their corpse. Whoopee, eh. He does not have a real sense of humour, and this defines to us that he is a nut because he takes everything so seriously. But there isn't anything unreal or unusual about news groups attracting nut cases. There are so many nut cases about. They wouldn't be tolerated anywhere else. This place is like a pub without a bouncer. Patrick Turner. |
|
#22
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Flipper said:
The inputs to a tube are voltages and the output is current. This is why it's called a transconductance device. For a triode the first voltage input is the grid, the second voltage input is the plate, the summing junction is the cathode, and one can find the transfer characteristics for each input in the tube's datasheet with the first, and most popular, being a set of curves showing the effect of the plate input voltage on output current when the grid input voltage is held constant with the second being a set of curves showing the effect of grid input voltage on output current when the plate input voltage is held constant. If we then run the current output through a resistor so that it generates a voltage signal in phase opposition to the grid, and if that phase opposite signal, representing a portion of the output, is then applied to the plate input, it sums at the cathode negatively with the grid signal and, so, is referred to as negative feedback. "Finally, it is possible by contrivance and notional reconfiguration of inputs and outputs to get *anything* to fit the canonical diagram. This is true not only of triodes but also of resistors, ducks, the colour blue and the man in the moon. In these cases the feedback is a function of the contrivance, and not a property of the duck or the triode. No real property applies equally to all things, obviously. All cows are black at night. "I am serious about looking at a book, honestly. It would be a revelation for anyone able to follow the maths." ipso facto, case closed. Hardly. Your argument equally applies to a resistor, or anything else in the universe. And you haven't closed a loop anyway...you need to reduce your logic such that your system has only one input (hence the "back" in "feedback"). Patrick has realised that, hence the blah-de-blah about stuff like field effects. What I have made very clear is that I am using the term feedback in the strict, engineering sense. Some know what I mean, and others can't be bothered to find out. There is no shortage of elementary texts on the subject of control systems engineering. Seriously, check one out. No reason for me to think you or Patrick ever will. Once again, this internal feedback mumbo jumbo confuses an issue that is otherwise quite straightforward. We got the answer right without it. Engineering theory is intended to be useful, not to mystify. Ian |
|
#23
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote: Flipper said: The inputs to a tube are voltages and the output is current. This is why it's called a transconductance device. For a triode the first voltage input is the grid, the second voltage input is the plate, the summing junction is the cathode, and one can find the transfer characteristics for each input in the tube's datasheet with the first, and most popular, being a set of curves showing the effect of the plate input voltage on output current when the grid input voltage is held constant with the second being a set of curves showing the effect of grid input voltage on output current when the plate input voltage is held constant. If we then run the current output through a resistor so that it generates a voltage signal in phase opposition to the grid, and if that phase opposite signal, representing a portion of the output, is then applied to the plate input, it sums at the cathode negatively with the grid signal and, so, is referred to as negative feedback. "Finally, it is possible by contrivance and notional reconfiguration of inputs and outputs to get *anything* to fit the canonical diagram. This is true not only of triodes but also of resistors, ducks, the colour blue and the man in the moon. In these cases the feedback is a function of the contrivance, and not a property of the duck or the triode. No real property applies equally to all things, obviously. All cows are black at night. "I am serious about looking at a book, honestly. It would be a revelation for anyone able to follow the maths." ipso facto, case closed. Hardly. Your argument equally applies to a resistor, or anything else in the universe. And you haven't closed a loop anyway...you need to reduce your logic such that your system has only one input (hence the "back" in "feedback"). Patrick has realised that, hence the blah-de-blah about stuff like field effects. What I have made very clear is that I am using the term feedback in the strict, engineering sense. Some know what I mean, and others can't be bothered to find out. There is no shortage of elementary texts on the subject of control systems engineering. Seriously, check one out. No reason for me to think you or Patrick ever will. Once again, this internal feedback mumbo jumbo confuses an issue that is otherwise quite straightforward. We got the answer right without it. Engineering theory is intended to be useful, not to mystify. Ian If ya can't see the difference between a resistor and a triode your'e a real ****in dumb **** for sure. Patrick Turner. |
|
#24
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick Turner wrote:
Ian Iveson wrote: Flipper said: The inputs to a tube are voltages and the output is current. This is why it's called a transconductance device. For a triode the first voltage input is the grid, the second voltage input is the plate, the summing junction is the cathode, and one can find the transfer characteristics for each input in the tube's datasheet with the first, and most popular, being a set of curves showing the effect of the plate input voltage on output current when the grid input voltage is held constant with the second being a set of curves showing the effect of grid input voltage on output current when the plate input voltage is held constant. If we then run the current output through a resistor so that it generates a voltage signal in phase opposition to the grid, and if that phase opposite signal, representing a portion of the output, is then applied to the plate input, it sums at the cathode negatively with the grid signal and, so, is referred to as negative feedback. "Finally, it is possible by contrivance and notional reconfiguration of inputs and outputs to get *anything* to fit the canonical diagram. This is true not only of triodes but also of resistors, ducks, the colour blue and the man in the moon. In these cases the feedback is a function of the contrivance, and not a property of the duck or the triode. No real property applies equally to all things, obviously. All cows are black at night. "I am serious about looking at a book, honestly. It would be a revelation for anyone able to follow the maths." ipso facto, case closed. Hardly. Your argument equally applies to a resistor, or anything else in the universe. And you haven't closed a loop anyway...you need to reduce your logic such that your system has only one input (hence the "back" in "feedback"). Patrick has realised that, hence the blah-de-blah about stuff like field effects. What I have made very clear is that I am using the term feedback in the strict, engineering sense. Some know what I mean, and others can't be bothered to find out. There is no shortage of elementary texts on the subject of control systems engineering. Seriously, check one out. No reason for me to think you or Patrick ever will. Once again, this internal feedback mumbo jumbo confuses an issue that is otherwise quite straightforward. We got the answer right without it. Engineering theory is intended to be useful, not to mystify. Ian If ya can't see the difference between a resistor and a triode your'e a real ****in dumb **** for sure. Patrick Turner. And if you cannot see the similarity between a variable resistor and a triode then you are just as dumb. Ian |
|
#25
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
flipper wrote
Flipper said: The inputs to a tube are voltages and the output is current. This is why it's called a transconductance device. For a triode the first voltage input is the grid, the second voltage input is the plate, the summing junction is the cathode, and one can find the transfer characteristics for each input in the tube's datasheet with the first, and most popular, being a set of curves showing the effect of the plate input voltage on output current when the grid input voltage is held constant with the second being a set of curves showing the effect of grid input voltage on output current when the plate input voltage is held constant. If we then run the current output through a resistor so that it generates a voltage signal in phase opposition to the grid, and if that phase opposite signal, representing a portion of the output, is then applied to the plate input, it sums at the cathode negatively with the grid signal and, so, is referred to as negative feedback. "Finally, it is possible by contrivance and notional reconfiguration of inputs and outputs to get *anything* to fit the canonical diagram. This is true not only of triodes but also of resistors, ducks, the colour blue and the man in the moon. In these cases the feedback is a function of the contrivance, and not a property of the duck or the triode. No real property applies equally to all things, obviously. All cows are black at night. "I am serious about looking at a book, honestly. It would be a revelation for anyone able to follow the maths." ipso facto, case closed. Hardly. Your argument equally applies to a resistor, or anything else in the universe. No, it doesn't and good luck on finding two independent variables in a one variable device. You would need to tell me what you are on about, naturally, and how you think it supports your argument, whatever it is. And you haven't closed a loop anyway... Yes, it does close the loop, as I clearly showed you need to reduce your logic such that your system has only one input (hence the "back" in "feedback"). Patrick has realised that, hence the blah-de-blah about stuff like field effects. All electronic devices operate on 'field effects'. Including resistors, which have internal feedback in exactly the same sloppy sense you have been using. What I have made very clear is that I am using the term feedback in the strict, engineering sense. Which is precisely what I showed. No you didn't. Some know what I mean, and others can't be bothered to find out. There is no shortage of elementary texts on the subject of control systems engineering. Seriously, check one out. Spent my life working with control systems. So has everyone. So what? Should I tug my forelock, or give you a prize? No reason for me to think you or Patrick ever will. Less reason now. Once again, this internal feedback mumbo jumbo confuses an issue that is otherwise quite straightforward. We got the answer right without it. Engineering theory is intended to be useful, not to mystify. There's no 'mystery' to it for anyone familiar with control theory and feedback. It doesn't make any contribution to the question from the OP, nor to any other practical problem. It fails to enlighten and so its presence is a red herring. Draw a canonical feedback diagram for a triode and I'll show you what is either unreal or illegal about it. If you manage to make it both real and legal, I'll show you how a resistor can be represented similarly. No use has ever been made of any theory of "internal feedback" in triodes. These theories (every protagonist has his own different one...you can't agree amongst yourselves, just like the theologians who spent centuries arguing about how many angels can stand on a pin) serve only to cause confusion, such as when SS apologists argue that all valve amps have feedback because it is inside the valves. You have all ****ed on our trump card. Anyway, we have been round these circles many times. I have read all the references suggested or supplied by everyone here who has challenged my argument, and all have been risible; but no-one such as yourself, still at sea, has read any of mine, or those cited by RDH4, or any other mainstream engineering text. You and the few others still clinging to straws are making no effort to learn. Bye. Ian |
|
#26
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
On Mon, 25 Jun 2007 12:23:48 +0100, wrote:
Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Cheers Ian RDH4, Chapter 2, section 5 (page 34) discusses "Triode operation of pentodes" There one finds the statement, "When the cathode current of a valve is shared by two collecting electrodes (e.g., plate and screen) the mutual conductance of the whole cathode stream (i.e., the 'triode gm') is shared in the same proportion as is the current." They then give 3 examples, the first with the plate and screen voltages the same when in pentode mode. They say that in that case, "...the method is exact". If the plate and screen voltages are not the same in pentode mode (as in the second example), then they say "...it is necessary to make an assumption which is only approximately correct...the error being within about 5% for most conditions." I notice that on this data sheet: http://www.mif.pg.gda.pl/homepages/f...93/6/6AU6A.pdf they use different cathode bias resistors for the various measurements, and they don't have a case where the screen and plate are at the same voltage to compare with a triode connection. Furthermore, the suppressor is connected to the cathode in pentode mode, but to the plate for triode mode. This makes it hard to do an apples to apples comparison. Perhaps one could supply the screen through a high value resistor, and have a large bypass capacitor to ground (with a cathode resistor or not, as the case may be). Different DC voltages could be applied to plate and screen in pentode mode, and a measurement of gm made with an audio frequency AC signal on grid 1 and an AC short on the plate; I think this is the usual method. Then the ground end of the screen bypass capacitor could be moved to the plate and gm could then be measured in triode mode, but still with different DC voltages on plate and screen. |
|
#27
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
|
|
#28
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
|
|
#29
Posted to rec.audio.tubes
|
|||
|
|||
|
NFB in triodes was Pentode gm wired as a triode
Once again, this internal feedback mumbo jumbo confuses an issue that is otherwise quite straightforward. We got the answer right without it. Engineering theory is intended to be useful, not to mystify. Ian If ya can't see the difference between a resistor and a triode your'e a real ****in dumb **** for sure. Patrick Turner. And if you cannot see the similarity between a variable resistor and a triode then you are just as dumb. Ian Somebody said a triode was no different to a resistor. I called the person who said that a ****in real dumb ****. Now the goal poats have been moved, and the resistor has suddenly become a variable one, and I am deemed to be dumb because I will never agree that a triode is like a variable resistor. But get back to the evidence of the device behaviour. Say you have a EL34 in triode set up at a fixed grid bias, and with 70mA of dc flowing with Ea = +350V. Pda = 24.5 watts and from the dc conditions, if the triode was a resistance, triode is an R = 350 / 0.07 = 5,000 ohms. Suppose we raise the EA to 400V. If it was an R, the R wouldn't change, and we'd see 80mA at 400V, as 400 / 0.08 = 5k. But we will see that Ia increases approximately 40mA to a total of 110mA, so from the dc conditions, R at 400V = 400 / 0.11 = 3,636 ohms. How come the resistance went low when we applied a higher Ea? What magical hand has been operating inside the tube to reduce its apprarent resistance looking into the anode circuit????? Variable resistances could be said to be like a triode, but only if we say that their adjustment in ohm value is automatic and internal; if there is an adjust knob someplace, then its a vraiable resistance, not a triode, right? But back at the EL34 triode, which would cook itself to death if left with Ea = 400V and Ia = 110mA, we can establish what the DYNAMIC resistance is if we divide the ccanhe of Ea by the change in Ia, and in this case 50 volts Ea rise gave us 40mA for this particular sample, so Ra, the dynamic resistance is 50 / 0.04 = 1,250 ohms. Ho come? The rise in anode voltage raises the force of attraction of electrons at the space charge around the cathode. The flow of electrons is controlled by the sum of anode voltage and grid voltage. grid is negative, trying to stop electrons flowing, anode is positive, encouraging electrons to flow, and the two fields of opposite polarity sum to form a resultant field, setting the steady current in a stae of equilibrium. If the anode voltage rises while grid voltage remains constant, then there is more encouragement given to electrons to flow, and off they go, and it is as if the anode has a transconductance effect on the current attracted to itself, ie, when V rises 50V, I rises 40mA, so there is 0.8mA/V of anode transconductance. This is all facilitated with an EL34 in triode, but if the screen is taken to a fixed voltage aof +350V, then things change drmatically, and for a rise in EA of 50V, we might find Ia rises only 4.167mA, so Ra would be calculated at 50 / 0..4167 = 12,000 ohms. How come when you have 5,000 ohms of static dc resistance the dynamic resistance is 12,000 ohms? We have removed nearly all the effect the anode has on its own current, and the anode gm is 0.0833mA/V, or about 10% of the effect when the tube is in triode. So the screen is what is injecting 90% of the effect of the anode. You can write up all sorts of equations about all this and I leave the brighter among you to do that. But the reason why the triode Ra is so low is that the anode voltage or a screen attatched to it has a large ability to oppose changes to current flow. Its NFB. Any resemblance of a triode to a variable resistance such as a potentiometer is simplistic and irrelevant to the issue of NFB in triodes. Patrick Turner |
|
#30
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
The Phantom wrote:
On Mon, 25 Jun 2007 12:23:48 +0100, wrote: Is there a simple relationship between the gm of a pentode wired as a pentode and its gm wired as a triode.? I have just done some tests on some 6AU6 pentodes. Wired as pentodes I get the expected gm of around 3.9 mA/V on my AVO two panel tester. If I alter the settings on the tester to wire them as triodes (suppressor and screen grids to plate and reducing plate volts from 100 V to 80V) I get gm readings of over 7mA/V. Is it typical for the triode wired gm to be nearly twice the pentode wired value? Cheers Ian RDH4, Chapter 2, section 5 (page 34) discusses "Triode operation of pentodes" There one finds the statement, "When the cathode current of a valve is shared by two collecting electrodes (e.g., plate and screen) the mutual conductance of the whole cathode stream (i.e., the 'triode gm') is shared in the same proportion as is the current." They then give 3 examples, the first with the plate and screen voltages the same when in pentode mode. They say that in that case, "...the method is exact". Good old RDH4. I have been looking all over the place in that book and wouldn't you know it there is a chapter on this very question. If the plate and screen voltages are not the same in pentode mode (as in the second example), then they say "...it is necessary to make an assumption which is only approximately correct...the error being within about 5% for most conditions." But in general it means that the triode connected gm is higher than the pentode gm. As a rough rule, screen current is typically 20% of cathode current so you would expect triode connected gm to be roughly 1/0.8 of the pentode gm or about 25% higher. Clearly a lot less than the values I was reading on my AVO tester. I notice that on this data sheet: http://www.mif.pg.gda.pl/homepages/f...93/6/6AU6A.pdf they use different cathode bias resistors for the various measurements, and they don't have a case where the screen and plate are at the same voltage to compare with a triode connection. Furthermore, the suppressor is connected to the cathode in pentode mode, but to the plate for triode mode. This makes it hard to do an apples to apples comparison. Perhaps one could supply the screen through a high value resistor, and have a large bypass capacitor to ground (with a cathode resistor or not, as the case may be). Different DC voltages could be applied to plate and screen in pentode mode, and a measurement of gm made with an audio frequency AC signal on grid 1 and an AC short on the plate; I think this is the usual method. Then the ground end of the screen bypass capacitor could be moved to the plate and gm could then be measured in triode mode, but still with different DC voltages on plate and screen. I think that is the way to go. As soon as the rest of the parts I have ordered turn up I shall be able to n=make some direct measurements. I also plan to try out the triode connection where suppressor and anode are grounded and the screen is used as a plate. RDH4 mentions this as a means of reducing hum and noise in preamps. Thanks for the input. Cheers Ian |
|
#31
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
flipper wrote: On Wed, 27 Jun 2007 11:40:22 GMT, Patrick Turner wrote: The inputs to a tube are voltages and the output is current. This is why it's called a transconductance device. But there does not have to be any current change within a triode. Yes, there does. You and I have been round on this before and while it may be convenient to think in terms of 'ideal' current sources and assume infinite impedances the real world fact of it is that infinite impedances do not exist and no current change = no voltage change. Or, to put mathematics to it, V=IR and if I=0 then V=0 whether you're talking DC or signal. The best you could hope for is to postulate the impossible infinite impedance exists, in which case you have 0 times infinity, which is undefined and every equation you use to design tube circuits also becomes undefined because they all depend on small signal current change per voltage change in their definition. Your being a silly nit picker on this issue. Everyone knows that a 20megohm load on a typical triode is a constant current source, and it makes no measurable difference to gain measurements or character parameters if the finite resistance of the load was an infinite value of resistance. Come down to earth. No need to dispense with wisdom ebacue tou get a silly concept like zero x infinity. So it is called a voltage device. It is not a 'voltage device', it's a transconductance device and it's listed right there in the datasheet as gm. And there is no transfer function nor equation that creates an 'output V per input V' without an external device creating the output V, because the output is I. But as I keep saying, and as you don't what to know, the tiode will operate as a gain device without any ia change. So triode gm with a CCS anode load cannot be measured; its as if its gm does not exist. The gm is fully measurable when the anode voltage is prevented from any change when grid voltage is changed. And while it was not common when there were only tubes around, because tubes generally need a voltage input to the grid, a tube can be used in it's native voltage in---current out transconductance modality. And I've done so using transistors as the next device because they operate with current input and, so, do not necessarily need an interposing conversion to voltage. A large voltage change occurs because a smaller one controls voltages. There is no voltage change unless one runs the output current through an external device such as, for example, a load resistor or the impedance of an inductor/transformer. And you can have a situation where the voltage change occurs without current change. For a triode the first voltage input is the grid, the second voltage input is the plate, the summing junction is the cathode, I don't see the cathode as THE summing junction. But there is a resultant field action on the space charge between cathode and grid which results from anode voltage and grid voltage which do form a summing junction, and the sum of the fields controls the electron flow, or the the charges themselves at the anode if no I change occurs. You're quibbling and the same quibble can be made about the 'junctions' in a transistor, which is not a 'bright line' either but a diffuse region where the field effects take place. We could also get into how the junction region shrinks and grows depending on field strength but that unnecessarily complicates first order principles. Well, the cathode IS NOT the actual place where the summing of anode and grid voltages occurs. Its a fact, and the summing point is away from thr actual electrode. The voltage at the cathode IS NOT the summed voltage that controls I flow. I just saw no need to confuse the basics of it with quantum effects, space charge, nor where the 'virtual' cathode is. and one can find the transfer characteristics for each input in the tube's datasheet with the first, and most popular, being a set of curves showing the effect of the plate input voltage on output current when the grid input voltage is held constant with the second being a set of curves showing the effect of grid input voltage on output current when the plate input voltage is held constant. The curves for plate characteristics are like the diode Ra curve. That's nice. And the characteristics of one PN junction in a transistor looks like a diode too because, taken alone, it is a diode. Why try to compare transistors to triodes when the discussion is about NFB in triodes, or pentodes wired as triodes? However, in both case, a triode and transistor, the 'one' thing is not alone. If we then run the current output through a resistor so that it generates a voltage signal in phase opposition to the grid, and if that phase opposite signal, representing a portion of the output, is then applied to the plate input, it sums at the cathode negatively with the grid signal and, so, is referred to as negative feedback. ipso facto, case closed. Hmm, I doubt you've said anywhere enough to convince Ian of anything. His problem is that he's set his mind against NFB in triodes, and he can't accept NFB now. To do that would be to admit defeat, and we could all laugh about why it took so long for him to "get it". He hates being the laughing stock. He hates being wrong. He's only happy when other ppl are wrong, so he can dance on their corpse. Whoopee, eh. Well, you've described how this news group tends to treat people who may discover something different than their first impression and it probably does lend to entrenchment because people don't generally like to be laughed at. I used to think there was no internal NFB in triodes until I saw the light. OK, it was easy to adjust my thinking. But not the Ian Ivesons of this world. But I, for one, don't mind discovering "by golly, I hadn't thought of that before" and don't consider it to be a 'flaw' in others either. Quite a few folks resent any change to their understanding of the truth. Just look at what happened to so many people in history when the truth arrived in their heads, such as the idea that the sun was the centre of the solar system, not the earth, and we go around the sun, and the sun does not go around us. Patrick Turner. He does not have a real sense of humour, and this defines to us that he is a nut because he takes everything so seriously. But there isn't anything unreal or unusual about news groups attracting nut cases. There are so many nut cases about. They wouldn't be tolerated anywhere else. This place is like a pub without a bouncer. Patrick Turner. |
|
#32
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
...Two voltage inputs, grid and plate, with a current output...
As I said, the canonical system has only one input. That's all you are getting from me until you demonstrate you have at least an inkling. You are too rude and ungrateful. Ian "flipper" wrote in message ... On Thu, 28 Jun 2007 00:47:23 GMT, "Ian Iveson" wrote: flipper wrote Flipper said: The inputs to a tube are voltages and the output is current. This is why it's called a transconductance device. For a triode the first voltage input is the grid, the second voltage input is the plate, the summing junction is the cathode, and one can find the transfer characteristics for each input in the tube's datasheet with the first, and most popular, being a set of curves showing the effect of the plate input voltage on output current when the grid input voltage is held constant with the second being a set of curves showing the effect of grid input voltage on output current when the plate input voltage is held constant. If we then run the current output through a resistor so that it generates a voltage signal in phase opposition to the grid, and if that phase opposite signal, representing a portion of the output, is then applied to the plate input, it sums at the cathode negatively with the grid signal and, so, is referred to as negative feedback. "Finally, it is possible by contrivance and notional reconfiguration of inputs and outputs to get *anything* to fit the canonical diagram. This is true not only of triodes but also of resistors, ducks, the colour blue and the man in the moon. In these cases the feedback is a function of the contrivance, and not a property of the duck or the triode. No real property applies equally to all things, obviously. All cows are black at night. "I am serious about looking at a book, honestly. It would be a revelation for anyone able to follow the maths." ipso facto, case closed. Hardly. Your argument equally applies to a resistor, or anything else in the universe. No, it doesn't and good luck on finding two independent variables in a one variable device. You would need to tell me what you are on about, naturally, and how you think it supports your argument, whatever it is. Your example of a 'resistor'. Can't have feedback because there aren't enough variables to do it. And you haven't closed a loop anyway... Yes, it does close the loop, as I clearly showed you need to reduce your logic such that your system has only one input (hence the "back" in "feedback"). Patrick has realised that, hence the blah-de-blah about stuff like field effects. All electronic devices operate on 'field effects'. Including resistors, Correct. So your field effects 'bla bla bla' is nonsensical. which have internal feedback in exactly the same sloppy sense you have been using. Nope, and it's impossible because there are not enough independent variables to make a loop with. What I have made very clear is that I am using the term feedback in the strict, engineering sense. Which is precisely what I showed. No you didn't. Sure did. All it takes is for you to actually look at it and put in 2 cents worth of thought. Some know what I mean, and others can't be bothered to find out. There is no shortage of elementary texts on the subject of control systems engineering. Seriously, check one out. Spent my life working with control systems. So has everyone. So what? Should I tug my forelock, or give you a prize? Come on. At least try to make sense. No reason for me to think you or Patrick ever will. Less reason now. Once again, this internal feedback mumbo jumbo confuses an issue that is otherwise quite straightforward. We got the answer right without it. Engineering theory is intended to be useful, not to mystify. There's no 'mystery' to it for anyone familiar with control theory and feedback. It doesn't make any contribution to the question from the OP, nor to any other practical problem. It fails to enlighten and so its presence is a red herring. Whether is 'enlightens' the OP is quite irrelevant. Draw a canonical feedback diagram for a triode and I'll show you what is either unreal or illegal about it. If you manage to make it both real and legal, I'll show you how a resistor can be represented similarly. Go ahead and give your resistor a shot. Can't be done because there aren't enough independent variables and you stomping your foot about it doesn't make it so. No use has ever been made of any theory of "internal feedback" in triodes. I've made use of it. You lose. Just to illustrate that inventing nonsensical absolutes you can't possibly know is nonsensical. These theories (every protagonist has his own different one...you can't agree amongst yourselves, just like the theologians who spent centuries arguing about how many angels can stand on a pin) serve only to cause confusion, such as when SS apologists argue that all valve amps have feedback because it is inside the valves. You have all ****ed on our trump card. Hehe. Well, sorry about your trump card but facts don't change just because they're inconvenient to your game. Anyway, we have been round these circles many times. I have read all the references suggested or supplied by everyone here who has challenged my argument, and all have been risible; but no-one such as yourself, still at sea, has read any of mine, or those cited by RDH4, or any other mainstream engineering text. You and the few others still clinging to straws are making no effort to learn. I've read all your arguments and they are mostly arm waving and foot stomping "I'm right, you're wrong" with a complete ignoring of anything that's presented to you. A portion of the output is sampled by the plate load and applied as a voltage signal to the plate. The two sum at the cathode and, viola, negative feedback. And I can use the thing without feedback by running the plate into a current mode device, like a transistor, and I do just that with a current mirror in my "PC Speaker Tube Hybrid" amp. And I did it because I wanted the gain that would otherwise get sucked up by the local negative feedback. Bye. Ian |
|
#33
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
But in general it means that the triode connected gm is higher than
the pentode gm. As a rough rule, screen current is typically 20% of cathode current so you would expect triode connected gm to be roughly 1/0.8 of the pentode gm or about 25% higher. Clearly a lot less than the values I was reading on my AVO tester. But the 6AU6 has much greater screen current that that. If you follow my link to the datasheet and the arithmetic, then the ratio given by your meter is about right, especially if you take into account the different voltages you used. For purpose of *comparison* there is not much wrong with the AVO, AFAIK, even if it is not calibrated accurately. Ian |
|
#34
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote:
But in general it means that the triode connected gm is higher than the pentode gm. As a rough rule, screen current is typically 20% of cathode current so you would expect triode connected gm to be roughly 1/0.8 of the pentode gm or about 25% higher. Clearly a lot less than the values I was reading on my AVO tester. But the 6AU6 has much greater screen current that that. If you follow my link to the datasheet and the arithmetic, then the ratio given by your meter is about right, especially if you take into account the different voltages you used. I don't see that but maybe I am looking at the wrong part of the data sheet. The part I see under 'Characteristics and Typical Operation' has a column with screen and plate volts both 100V and a gm of 3.9mA/V. These are the voltages my AVO tester is set up with for meauding the 6AU6 in pentode mode. The plate current is 5mA and the screen current is 2.1mA, about 30% of cathode current - OK that's higher than my 20% figure but not hugely. This gives a triode equivalent gm of 3.9 x 7.1/5 = 5.5mA/V, still rather lower than the 7 I was measuring. I can see in the right hand column where the plate is at 250V and the screen at 150 that the gm rises to 5.2. The triode equivalent under these conditions works out at about 7.2mA/V which is much closer to what I measured but with rather different voltages. Ian |
|
#35
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Other Ian said
I don't see that but maybe I am looking at the wrong part of the data sheet. The part I see under 'Characteristics and Typical Operation' has a column with screen and plate volts both 100V and a gm of 3.9mA/V. These are the voltages my AVO tester is set up with for meauding the 6AU6 in pentode mode. The plate current is 5mA and the screen current is 2.1mA, about 30% of cathode current - OK that's higher than my 20% figure but not hugely. This gives a triode equivalent gm of 3.9 x 7.1/5 = 5.5mA/V, still rather lower than the 7 I was measuring. I can see in the right hand column where the plate is at 250V and the screen at 150 that the gm rises to 5.2. The triode equivalent under these conditions works out at about 7.2mA/V which is much closer to what I measured but with rather different voltages. Firstly, 30% is much greater than 20% in my book. 50% greater, in fact. Second, the issue is gm not current, ie rate of change of current with voltage, dIa/dVgc. My inspection of the curves was made at 150V because the given pentode curves are with the screen at 150V Screen gm at that voltage is about 1.2mA/V, whereas anode gm is 2.8. That is over 40%. When does a difference become "huge"? As you get closer to the knee of the curves, the proportion increases further. All I am saying is that your meter may be showing the right ratio of values, although it may not be showing the correct absolute values. I have an AVO CT 160 that is fairly accurate in the middle of its ranges, but rather poor at the extremes. However, it has always given usably accurate comparisons, in that the error is reliable and favourably distributed. There is much greater provision for calibration on yours, but AFAIK the fundaments of its operation are equally bizarre, using AC on screen and anode. cheers, Ian1 |
|
#36
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
On Thu, 28 Jun 2007 21:41:35 -0500, flipper wrote:
On Thu, 28 Jun 2007 12:21:08 GMT, "Ian Iveson" wrote: ...Two voltage inputs, grid and plate, with a current output... As I said, the canonical system has only one input. Look at a tube datasheet and, in particular, Plate V and grid V. That's 2. Pay close attention, folks. Here is where the argument takes its big twist. Thanks, as always, to all, Chris Hornbeck "Money doesn't buy happiness. But happiness isn't everything." - Jean Seberg |
|
#37
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Chris said:
As I said, the canonical system has only one input. Look at a tube datasheet and, in particular, Plate V and grid V. That's 2. Pay close attention, folks. Here is where the argument takes its big twist. Engineers can't afford to entertain big twists, Chris, or to argue about how many angels can stand on a pin. That's what convention is for. Hence most things they make don't fall apart because one man's common sense is another's catastrophic blunder. DO NOT SCALE. ASK. The canonical system has only one input. This is not just about systems either. The whole principal of *analysis* depends on it, together with its partner, the principle of superposition. How would you arrive at a bode plot, for example, for a system with two inputs? Perhaps at two unrelated frequencies? Can't be done. Either the inputs are functionally related, in which case you would reconfigure and reduce to one input, or they are not, in which case you would need to posit two systems, which may be nested, or may be in series, or in parallel. For conventional systems theory to apply, and for its methods to be valid, each system considered must conform to the canonical model. The canonical model has an input that arrives at a summing point, from which a forward function leads to the output. From the output, a feedback function leads to the same single summing point. Notice there can be one input only. Not two, or three, or any other multiplicity. Just one. 1 cheers, Ian1 |
|
#38
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote: Other Ian said I don't see that but maybe I am looking at the wrong part of the data sheet. The part I see under 'Characteristics and Typical Operation' has a column with screen and plate volts both 100V and a gm of 3.9mA/V. These are the voltages my AVO tester is set up with for meauding the 6AU6 in pentode mode. The plate current is 5mA and the screen current is 2.1mA, about 30% of cathode current - OK that's higher than my 20% figure but not hugely. This gives a triode equivalent gm of 3.9 x 7.1/5 = 5.5mA/V, still rather lower than the 7 I was measuring. I can see in the right hand column where the plate is at 250V and the screen at 150 that the gm rises to 5.2. The triode equivalent under these conditions works out at about 7.2mA/V which is much closer to what I measured but with rather different voltages. Firstly, 30% is much greater than 20% in my book. 50% greater, in fact. My data book for 6AU6 has the following data, Ig2 = 4.3mA, Eg2 = 150V, when Ia = 10.6mA, Ea 250V, gm = 5.2mA/V, And Ig2 = 2.1mA, Eg2 = 100mA when Ia = 5.0mA, Ea = 100V, gm = 3.9mA/V So over a broad range of voltages, Ig2 = 40% of Ia. However, if you go to the RCA sheets with 6AU6 TRIODE curves, at Ia = 10mA and Ea = 250V the µ = 38.6 and Ra = 9.1k, and gm = µ / Ra = 38.6 / 9,100 = 4.24 mA/V Who was it who told me that triode gm was HIGHER than pentode gm???????? Who was it who mocked me like a howling hyena because I said triode gm would not be higher than pentode gm? Ian Iveson of course. This arch would be know all who really is a chief ****wit who only ever has got a prize for being long winded without actually saying anything anyone can understand. He refuses to do the hard yards to examine data sheets or perform breadboard experiments to measure what a 6AU6 will actually do. If ya wanna see what you get with a 6AU6 in triode or pentode, set one up in a gain circuit and ****in measure the *******! If yer don't look, yer won't know! Patrick Turner. Second, the issue is gm not current, ie rate of change of current with voltage, dIa/dVgc. My inspection of the curves was made at 150V because the given pentode curves are with the screen at 150V Screen gm at that voltage is about 1.2mA/V, whereas anode gm is 2.8. That is over 40%. When does a difference become "huge"? As you get closer to the knee of the curves, the proportion increases further. All I am saying is that your meter may be showing the right ratio of values, although it may not be showing the correct absolute values. I have an AVO CT 160 that is fairly accurate in the middle of its ranges, but rather poor at the extremes. However, it has always given usably accurate comparisons, in that the error is reliable and favourably distributed. There is much greater provision for calibration on yours, but AFAIK the fundaments of its operation are equally bizarre, using AC on screen and anode. cheers, Ian1 |
|
#39
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Ian Iveson wrote:
Other Ian said I don't see that but maybe I am looking at the wrong part of the data sheet. The part I see under 'Characteristics and Typical Operation' has a column with screen and plate volts both 100V and a gm of 3.9mA/V. These are the voltages my AVO tester is set up with for meauding the 6AU6 in pentode mode. The plate current is 5mA and the screen current is 2.1mA, about 30% of cathode current - OK that's higher than my 20% figure but not hugely. This gives a triode equivalent gm of 3.9 x 7.1/5 = 5.5mA/V, still rather lower than the 7 I was measuring. I can see in the right hand column where the plate is at 250V and the screen at 150 that the gm rises to 5.2. The triode equivalent under these conditions works out at about 7.2mA/V which is much closer to what I measured but with rather different voltages. Firstly, 30% is much greater than 20% in my book. 50% greater, in fact. Depends on its effect. Second, the issue is gm not current, ie rate of change of current with voltage, dIa/dVgc. Er, I thought it was about how gm is split in the ratio of the plate and screen currents - so its about current AND gm. My inspection of the curves was made at 150V because the given pentode curves are with the screen at 150V Screen gm at that voltage is about 1.2mA/V, whereas anode gm is 2.8. That is over 40%. When does a difference become "huge"? When it makes a significant change to the answer. And no, I am not going to define significant ;-) As you get closer to the knee of the curves, the proportion increases further. All I am saying is that your meter may be showing the right ratio of values, although it may not be showing the correct absolute values. I have an AVO CT 160 that is fairly accurate in the middle of its ranges, but rather poor at the extremes. However, it has always given usably accurate comparisons, in that the error is reliable and favourably distributed. OK There is much greater provision for calibration on yours, but AFAIK the fundaments of its operation are equally bizarre, using AC on screen and anode. Yes it is odd isn't it. They even have (had) a patent on the technique. Ian cheers, Ian1 |
|
#40
Posted to rec.audio.tubes
|
|||
|
|||
|
Pentode gm wired as a triode
Patrick Turner wrote:
Ian Iveson wrote: Other Ian said I don't see that but maybe I am looking at the wrong part of the data sheet. The part I see under 'Characteristics and Typical Operation' has a column with screen and plate volts both 100V and a gm of 3.9mA/V. These are the voltages my AVO tester is set up with for meauding the 6AU6 in pentode mode. The plate current is 5mA and the screen current is 2.1mA, about 30% of cathode current - OK that's higher than my 20% figure but not hugely. This gives a triode equivalent gm of 3.9 x 7.1/5 = 5.5mA/V, still rather lower than the 7 I was measuring. I can see in the right hand column where the plate is at 250V and the screen at 150 that the gm rises to 5.2. The triode equivalent under these conditions works out at about 7.2mA/V which is much closer to what I measured but with rather different voltages. Firstly, 30% is much greater than 20% in my book. 50% greater, in fact. My data book for 6AU6 has the following data, Ig2 = 4.3mA, Eg2 = 150V, when Ia = 10.6mA, Ea 250V, gm = 5.2mA/V, And Ig2 = 2.1mA, Eg2 = 100mA when Ia = 5.0mA, Ea = 100V, gm = 3.9mA/V So over a broad range of voltages, Ig2 = 40% of Ia. However, if you go to the RCA sheets with 6AU6 TRIODE curves, at Ia = 10mA and Ea = 250V the µ = 38.6 and Ra = 9.1k, and gm = µ / Ra = 38.6 / 9,100 = 4.24 mA/V Who was it who told me that triode gm was HIGHER than pentode gm???????? Actually is was page 34 of RDH4. Ian |
| Reply |
| Thread Tools | |
| Display Modes | |
|
|
Similar Threads
|
||||
| Thread | Forum | |||
| Pentode in Triode Mode | Vacuum Tubes | |||
| Using power triode or power pentode wired as a triode as a split-load phase splitter tube? | Vacuum Tubes | |||
| Using power triode/pentode wired as triode as a split load phase splitter tube? | Vacuum Tubes | |||
Linear Mode
