|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
150.148.218.189
In Reply to: RE: 6C33 OTL posted by Banat on February 18, 2014 at 05:33:52
Dear Banat,
I was very interested to read that you recommend a capacitor in parallel with the recommended cathode resistors on the 6C33Cs. I also see that it is large enough in value (1000uF) to bypass all audio AC around the cathode resistor. That idea is good in one way; it prevents an increase in overall output impedance at audio frequencies that would otherwise be caused by the insertion of the 10R cathode resistors. But on the other hand, does it not introduce a new issue? Specifically, all the audio has to go thru that capacitor, which is in series with the output signal. In practice, have you perceived any audible problems with that? (In a way, it cancels out one of the great advantages of the circlotron vs the Futterman, direct coupling.) Otherwise, I like the idea, because I have been reluctant to use the larger value resistors, e.g., 1 ohm vs 10 ohms, for fear of raising output Z by an objectionable amount. Or is there something amiss with my analysis? Thanks.
Follow Ups:
Hi Lew !10R cathode power res. bypased with 1000uF Elko alltogether with separate DC coupled driver results on so called `combined ` power tube(s) bias , where minor negative bias voltage is achieved automatic on cathode R/C member(s) , but major power tubes(s) negative bias voltage come as usual from separate DC coupled drivers cathode(s).
This solution is basically one good Compromise , since on the one side result in improved stability operation of output power tubes , but on the other side result in slightly to insignificantly reduced sonic performance of the OTL Circlotron Amp .
This mean that order of magnitude of sonic performance decrease is 1000 times or even less than in case when we want to use any type of the best feromagnetic transimpedanse power device ( read OPT ) , by use of this `combined` bias method highest possible sonic resolution Amp performance is still obtained , which is typical sonic characteristic for any good designed OTL Amp .
BTW , my formula to chose optimum value of cathode W.W. res. is R=1/10 of B+ , for example , if B+ is 140VDC than Rk have to be 14R , but since 15R is common available res. value than this sligtly higher value is good too .
__
Enlightened Evolution-Astral Projection
Edits: 02/19/14 02/19/14
If we posit that the cathode resistor is used to promote current sharing and that the steady-state DC bias current will remain essentially constant during music reproduction, then we have AC current variation during music reproduction, only. So if the bypass capacitor shunts all audio frequencies, i.e., all AC, around the cathode resistor, does it not also negate the capacity of the cathode resistor to promote current sharing among the output tubes, when we listen to music?
This would be an especially important consideration for me, using 7241 output tubes, because of the need for cathode resistors to help the 3 internal cathodes to share current as equally as possible.
Thanks.
Lew
DC fixed negative grid bias is not steady during music reproduction , it is modulated with AC signal from driver(s) cathodes , hi constantly change the value depending from form & level of AC signal present on the Amp input , hi have steady state value only in - no input signal- condition . ,
but in the same time across the R/C member connected in catode(s) circuit that small DC voltage (automatic bias or self bias ) value ideally have to be constant .
Only what you need is to determine R/C time constant for lowest freq. that you want to amplify , and usually that is 20 Hz .
My point is actually next : for example , if accidetally input signal is become in subsonic range ( 5Hz or so ) , and is it with relative high amplitude than output power tubes un-necesery sufer relative high internal disipation without any final audible gain , further R/C member durring that sub-sonic ` fly back ` automatically raise up the self bias value causing efect of power tubes self protection regardless to grid subsonic high level input signal .
__
Enlightened Evolution-Astral Projection
what you say here about protecting the output tubes from subsonic input. And both you and Chris have corrected my assumption that the DC current remains relatively stable whereas there is lots of variation in AC current under dynamic conditions. Apparently that is not the case. So I've learned something from you guys.
If you aim for a cutoff of 20Hz, you will introduce phase shift up to about 200Hz.
That is why we set our cutoff frequency at about 2Hz, so there is no phase shift at 20Hz.
If you have subsonic input (just IMO mind you) its not the job of the amp to process that out, rather that of the source to the amplifier to see to it that its not there.
(Of course we go for wide bandwidth in our preamps too, so that puts the onus on the actual sources, analog or digital.)
I mentioned that in a post below, and I calculated that for a 30-ohm cathode resistor (per cathode of a 7241), I would need about 3000uF to bypass audio with -3db at 2Hz. (With 3 cathodes in a 7241, that would give me a net of 10 ohms to ground.) "3000uF" was calculated "in my head". When I later did it with an actual calculator, I get ~2600uF. But since the voltage is very low across the resistor, one can get away with very high quality electrolytics as used in filament supplies (16V ought to do it at practicable values of R). Then bypass the lytic with a good film cap, which can also be small in size because low in voltage.
Having said all that, I am not at all certain that I will try it.
In my Diy MA1s....i well need 56 caps to by-pass my 5 ohm cathode resistor ................some tell me why this is a good thing??
An if theres any one out there that cant hear the Bass an sound get better as i add more mf to the PS....you dont have the amps i have!
Thanks for any info on OTLs
Why it's a good thing: Banat told us. It puts a little local bias on the output tube and it eliminates the increase in output impedance associated with adding the cathode resistors, especially if you use the higher value cathode resistors that Banat favors.
The negatives: Banat admits that the audio goes through the shunt capacitor, inevitably, but with very minimal negative effects. Chris points out that the audio goes through the PS caps, anyway. My new question is whether the bypass cap in fact negates the capacity of the cathode resistor to promote current sharing under dynamic conditions, because AC current changes are shunted around the cathode resistor.
FWIW, I already said that I have the same finding as you: more PS capacitance gives better bass response. Chris's (and Ralph's) conception of the operation of the amp predicts that too. I don't think anyone is arguing about that at all.
"But on the other hand, does it not introduce a new issue? Specifically, all the audio has to go thru that capacitor, which is in series with the output signal. In practice, have you perceived any audible problems with that? (In a way, it cancels out one of the great advantages of the circlotron vs the Futterman, direct coupling.)"
Well, the audio signal in any case has to go through the power-supply capacitors, so it is really just "a bit more of the same."
Chris
Circlotron often baffles me, I admit, but I see the PS capacitors as being in parallel with the output, not in series. You may say that this is a distinction without a difference. I wouldn't know until I try it.
This is the gift of the Circlotrons output setup....is there any other amp or preamp that lets you put the PS caps on the Speakers??
THIS setup has to give the best bass of any thing that got tubes init.............right ....people say this all the time About the Circlotron......an any cap you put in well chang the sound....some add the V-cap an WOW....i have some 5600mf caps pr leg on the outputs of my diyed MA1s....for me this is were my old sprague730p 1.mf are sweet as honey....an my Apogee Stages are full of these sprague caps allso...
to my old ears there not much can beet these caps....thanks for saying it one more time...it fun to just think about the output setup....how can it not sound good??
"This is the gift of the Circlotrons output setup....is there any other amp or preamp that lets you put the PS caps on the Speakers??"
There's no big difference, in this particular respect, between the circlotron and a totem-pole or Futterman OTL. In each case, the audio signal that passes through the output tubes passes also through the power supply.
Chris
A Circlotron amplifier will function without filter caps in the power supplies. I've seen people talk about how the filter caps act as coupling caps, but if the amp works without them, seems to me maybe that isn't quite the case. Just say'n :)
"A Circlotron amplifier will function without filter caps in the power supplies. I've seen people talk about how the filter caps act as coupling caps, but if the amp works without them, seems to me maybe that isn't quite the case. Just say'n :)"
I suppose it depends what "function" means. Without power supply capacitors, the two supplies for the output stage are producing full-wave rectified sinewaves. This means the supply voltages are varying between (say) 150V and zero volts 120 times per second. That means there will be significant intervals of time during every 1/120 th second period when the amplifier is incapable of outputting any audio signal. This will inject a huge 120 Hz "modulation" onto the audio output.
In any case, the logical point is that whatever current passes through an output tube necessarily also passes through the power supply that connects to it. So the audio signal passing through the tube will necessarily pass also through the power supply. There is literally nowhere else that it can go. (As in the schematic at the top of the page; the wires from the output tube anodes connect to the power supply and nothing else.)
The discussion of what might happen if the power supply had no capacitors in it is an amusing one in its own right, but not really relevant in practice, I think. When one does use a capacitor, the audio signal passes almost exclusively through it as it passes through the power supply. Apart from anything else, for the majority of the mains cycle the sinewave voltage from transformer secondary is less than the voltage stored on the capacitor, and so the rectifier diodes are non-conducting. That leaves only the capacitor as the route by which the audio signal can pass.
Chris
The Futtermans I have seen employ an *actual* coupling cap in addition to the 'coupling caps' in the power supply. So we can look at the Circlotron has having 'one less'.
Another method of sorting out if a coupling cap exists is to examine the DC capabilities of the circuit. If a coupling cap is present the circuit will have a timing constant that will prevent DC response. We don't see that in the Circlotron- the output is capable of holding DC levels indefinitely, in fact requires a DC offset control system due to the lack of a coupling cap. So I've always seen this as a 'cake and eat it too' sort of thing; if you have a coupling cap you don't have DC response. This is of course a pragmatic viewpoint, I am sure stricter definitions exist!
"Another method of sorting out if a coupling cap exists is to examine the DC capabilities of the circuit. If a coupling cap is present the circuit will have a timing constant that will prevent DC response. We don't see that in the Circlotron- the output is capable of holding DC levels indefinitely, in fact requires a DC offset control system due to the lack of a coupling cap. So I've always seen this as a 'cake and eat it too' sort of thing; if you have a coupling cap you don't have DC response. This is of course a pragmatic viewpoint, I am sure stricter definitions exist!"I think an important point in the discussion is the relative timescales of the period of the audio signal, on the one hand, versus the timescale of the "pumping up" of the smoothing capacitor every 1/120 of a second. Or, in other words, the relative frequency of the audio signal under discussion versus the mains frequency. But first, it should be emphasised that regardless of the relative frequencies, the audio signal *will* go through the power supply, since there is no other route from the anodes of the output tubes than through the power supplies. This is just basic conservation of charge. The only matter for debate could be how much passes through the smoothing capacitor and how much through the rectifiers and power-transformer secondary.
If we consider an audio signal with a frequency that is large compared to 60Hz, then there will be many oscillations of the audio signal between the periodic replenishment of the charge on the smoothing capacitors, which happens every 1/120 of a second. During much of the time between the replenishments, the instantaneous power-transformer secondary voltage will be less than the voltage across the capacitor, and so the rectifier diodes are non-conducting and the *only* route for the audio signal to follow is through the capacitor. It is in series with the loudspeaker. In this regime, the audio path is clearly through the smoothing capacitor, and it is playing a role much like any other coupling capacitor.
At the other extreme, if we consider an "audio" signal that is very low in frequency compared to the 60Hz mains (and the DC response you mentioned is the limiting case of this), then there will be many replenishments of the charge on the capacitor during one cycle of the very low "audio" frequency. In this regime, the power supply with its smoothing capacitor is behaving rather differently from a pure coupling capacitor, since the recharging process from the power transformer is now playing a big role during the course of a single very low frequency "audio" cycle. That "replenishment" process doesn't occur with an ordinary coupling capacitor.
In practice, the various parameters like the current-supplying capability of the transformer, and the value of the smoothing capacitor, are chosen so that there is very little "sag" of the power supply DC voltage under the load demanded by the amplifier, and there is very little ripple on the DC output level from the supply. Under these conditions, it is therefore the case that the rectifier diodes are non-conducting for the majority of the mains cycle, and that therefore the power transformer is completely "out of circuit" for the majority of the time. Since the fluctuating current (the "audio signal") passing through the output tubes must necessarily pass through the power supply, it follows that it is necessarily passing overwhelmingly though the smoothing capacitors, and not through the rectifier diodes and power transformer secondary winding. (This is especially clear cut for the higher audio frequencies. One could debate a bit more about the "near to DC" extreme case.)
Chris
Edits: 02/20/14
So, if we used batteries instead, would there be coupling caps?
The circuit would work the same way.
"So, if we used batteries instead, would there be coupling caps?"
Maybe the term "coupling" is in any case not the best one to describe the role of a PS capacitor, since "coupling capacitor" carries the connotation of a capacitor between one stage of amplification and the next. The PS capacitor is in the audio path, and to that extent it has features in common with a coupling capacitor.
If batteries are used instead, then we now have the audio signal passing through them. Presumably if the batteries are essentially "ideal" (maybe like a set of car batteries in series?), they will have extremely low impedance and there will be no significant interference with the audio signal. However, given the way that passions are aroused by discussions of the pros and cons of different capacitor types, I could easily imagine analogous debates over batteries, whether Sears is better than Pep-Boys, etc., etc.!!!
I see Lew's point about whether the cathode bypass capacitors would negate the benefits of putting in cathode resistors for current-balancing purposes. My suspicion would be that any resulting unbalance for the AC audio currents would be a relatively minor issue, and the main concern is to ensure some degree of stability and balance of the quiescent DC currents through the tubes, which can be quite large and dissipate a significant amount of power. I know from my own experience that if one isn't careful, then with paralleled output tubes it can be possible for one of them to hog most of the current, and hence dissipate a lot more power than the others.
Chris
ChrisMr.Karsten says to you : "So, if we used batteries instead, would there be coupling caps?"
And you responded to him that you have responded , but I ask you now : What`s happen with car ( lead acid ) batteries when they during exploatation become discharged ? .
I think than they need some battery Charger to again become Charged !,( and usually simple battery charger consist from one main trans. and SS rectifier ).Any way I want to tell you next , only by using just my simple as possible sentence & words without Any intention to discredite you or anything simmilar !!! :
Batteries are just DC E.energy power Source Storage Units , same as Elko`s in convetional Amp PSU , further stored E.energy from batteries are partialy wasted (discharged) in power tubes , generating un-usefull heat, and finally some of that stored DC E.enegry is discharged on the load(speaker) , converting that stored E.energy in sound , also in un-usefull heat too , that generated sound is only Usefull Work that we actally need .
And yes ! all of this 3 device (charged battery , power tube ,load(speaker )are connected in series !
Finally I want to say that any serious E.Engineer look in this way how that E.machine (Amp)work , audophiles and amateurs don`t , they use term`s something like : audio-path , audio that or this , and so on and so forth , bla , bla , bla .....Best Regards !
__
Enlightened Evolution-Astral Projection
Edits: 02/20/14
"And you responded to him that you have responded , but I ask you now : What`s happen with car ( lead acid ) batteries when they during exploatation become discharged ? ...."
I'm not sure I'm catching the point you are making, but it seems anyway we are in agreement that the power supply (battery in this case) is in series with the speaker and tubes. The AC audio current that passes through the loudspeaker passes also through the battery. If you want to call that the "audio path" or not is just a matter of preference for words, but the facts are presumably not in dispute.
Chris
Chris
If we connect that same Amp ( basically one black box who amplify something ) which we usualy use in home for reproducing of music signal on the load of table shaker , which is usual tool in airplane industry , and excite that Amp with different input signal in range from 20Hz - 20Khz , same as is in audible freq. range , can we say that inside of PSU`s Elko`s pass some table shaker signal ? , I think no !
__
Enlightened Evolution-Astral Projection
"If we connect that same Amp ( basically one black box who amplify something ) which we usualy use in home for reproducing of music signal on the load of table shaker , which is usual tool in airplane industry , and excite that Amp with different input signal in range from 20Hz - 20Khz , same as is in audible freq. range , can we say that inside of PSU`s Elko`s pass some table shaker signal ? , I think no !"Banat, now I'm really puzzled where this discussion is going! I can't see that there is any new logical issue that arises for shaker tables that is in any way dfferent from the situation with loudspeakers and audio signals. They are exactly equivalent, as far as the point under discussion is concerned. It makes me suspect that the argument is more about peoples' preferred set of words to describe the same unambiguous and undisputed set of facts.
But just in case, let me pose three questions that have yes/no answers to you:
1) Would you say the shaker table signal is passing through the wires from the amplifier to the shaker table?
2) Would you say the shaker table signal is passing through the output tubes in the amplifier?
3) Would you say the shaker table signal is passing through the power supplies in the amplifier?
Chris
Edits: 02/20/14
ChrisI see no point to respond you on that 3 question !, only by giving yes or no answer , since I have allready to much clear&loud responses here , I don`t want to open new marathon of questions & answers , and it is little beat to late hours by me here on this side of planet Earth .
Best Regards !
__
Enlightened Evolution-Astral Projection
Edits: 02/20/14
nt
Maybe the term "coupling" is in any case not the best one to describe the role of a PS capacitor, since "coupling capacitor" carries the connotation of a capacitor between one stage of amplification and the next. The PS capacitor is in the audio path, and to that extent it has features in common with a coupling capacitor.
That's always been my opinion.
With regards to Banat's circuit (seems I saw that proposed on DIYAudio a few years back) it might be a good idea to use resistors that are more like 100 ohms. That would make the caps smaller and the bias would not be self bias as it would still need some bias voltage, but not nearly so much. :) Each approach has its own set of problems...
because, as you say, that would reduce the required size of the bypass capacitor. I calculated in my head that to achieve a -3 db point at 2Hz with a 30-ohm cathode resistor (so response would be flat at 20Hz), I'd need ~3000uF bypass caps. I'm sure one can get by well with a higher -3db point, however.
With my amps, that's a lot of new parts to buy and install (3 cathodes times 4 tubes per monoblock), if I were to do it. 24 new resistors and bypass capacitors, in total.
If you install such cathode resistors, you will want to increase the power supply voltage accordingly.
just after posting. 100R per cathode would drop around 13V per cathode, if I bias the whole 7241 at around 400mA, which is typical. Thankfully there is no real need to go that high in resistance.
Chris thanks............i can see with the 4ea 6C33 OTL..an the need for them to be fused....that 4 more caps may be the way to go...ezey to pull
if one feels there a down side....sound wise..i would think these should also have a smaller .1mf cap ...Ah...yes the never ending story of the OTL...an the ESL...
So it gos........
Yes but there not fully balanced differential output Amps....An some use caps on there outputs...............right??
An none i have heard sound like or as good as the Circlotron....Well too me..........
"Circlotron often baffles me, I admit, but I see the PS capacitors as being in parallel with the output, not in series. You may say that this is a distinction without a difference. I wouldn't know until I try it."
I would say that, roughly speaking, the upper half of the audio waveform passes through one of the power supplies, and the lower half passes through the other power supply. (With a relatively minor amendment to that statement, since there will be some lesser amount of quiescent current through each set of tubes at the cross-over point in the waveform.)
In any case, if you trace the current through the loudspeaker, tubes and power supplies, it is evident that whatever audio current passes through each set of output tubes passes also through the power supply connected to their anodes (since it can go nowhere else). So if it is agreed that the audio signal is passing through the output tubes, then it must necessarily also pass through the corresponding power supply.
Chris
That seems correct and logical.
But the issue you raised is whether that 1000uF capacitor bypassing the cathode resistor has a topology identical to or different from that of the PS capacitors, with respect to the signal. I suggested that the PS capacitors "look" to me to be in parallel with the signal, whereas the proposed cathode resistor bypass cap is in series. I am not in a position to argue, I just want to know whether I am right or wrong in thinking that way. That is, I want to learn something.
(By the way, I hope we all agree that the term "cathode resistor" is not used in the conventional sense of the term; the signal is taken from the distal end of the resistor, at the ground potential of the PS that serves that side of the circuit, not from the junction of the resistor with the cathode of the power tube. The resistor is there just to promote current sharing among the output tubes.)
Lew
Yes !, by Circlotron OPS PSU C`s are definitive in Parallel with load(speaker)! , but cathode(s) res. bypass C`s are definitive in Series with load(speaker) !
( That`s can easy open new `playground` for probes & experimentation with different sorts, brands names and values for Rk bypass Elko`s :), but it is relative easy now since this cathode bypass C`s are relative available & cheap.)
__
Enlightened Evolution-Astral Projection
"Yes !, by Circlotron OPS PSU C`s are definitive in Parallel with load(speaker)! , but cathode(s) res. bypass C`s are definitive in Series with load(speaker) !"
I don't agree. Look at the schematic; the upper pair of tubes, the supply B1 and the loudspeaker are all connected in *series*. Likewise, there is a second series-connected loop comprising the lower pair of tubes, the supply B2 and the loudspeaker. One of those loops predominantly handles the upper half of the audio waveform, and the other loop predominantly handles the lower half of the waveform. In each case the corresponding power supply is in series.
(The discussion is slightly complicated by the fact that each set of tubes conducts for rather more than just its half-cycle of the waveform, but this is inessential to the key point.) If we consider the bulk of the audio waveform, away from the crossover region, either one set of tubes or the other is passing the overwhelming portion of the current. We then have a simple loop, comprising the speaker, the set of conducting tubes, and the associated power supply, all in series.
One wants, for example, that the two power supplies should have as small an impedance at audio frequencies as possible. If the capacitance of the filter capacitors was too small then the output impedance of the power supplies would rise significantly at the lower frequencies and there would be a loss of low-frequency audio output, exactly as would happen with any situation where a capacitor is in series with the speaker.
Chris
FWIW, what you say is consistent with my experience: the more PS capacitance, the better the bass response, subjectively.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: