OTL Asylum

OTL, Output Transformerless Amplifier User Group.

Return to OTL Asylum

Message Sort: Post Order or Asylum Reverse Threaded

CCS in a circlotron???

71.178.181.244
Posted on December 17, 2014 at 13:20:54
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
 Someone pointed out to me that in many respects a circlotron stage "looks" like a cathode follower. This started me thinking about the utility of installing a CCS in a circlotron, specifically between the negative nodes of each of the two power supplies and the resistor that "floats" the supplies. (Perhaps if a CCS were installed, then one would not need that resistor; that's a separate question.) I admit I am shaky in my understanding of exactly how a circlotron works or better to say "why" it works. Because of the high currents in a circlotron amplifier, perhaps the use of a CCS would be more practical in the context of a preamplifier output stage, like that of an MP1 or MP3.

I searched Jon Broskie's site to find any mention of this idea, since he has written extensively about circlotrons. I find that he did propose a 5687-based circlotron stage that uses CCSs where I imagine using them, but with only one PS serving the whole stage. Thus the CCS in that case contributes separation between the two halves of the circuit, even with only one voltage source (PS). Any comments appreciated.

 

Hide full thread outline!
    ...
If you have only one power supply, posted on December 18, 2014 at 09:56:58
Ralph
Manufacturer

Posts: 4769
Location: Minnesota
Joined: April 24, 2002
 Its pretty hard to keep that one supply from being part of the load of each half of the circuit. That's why having 4 supplies in the output of the MP-1 allows it to drive stereo headphones- the power supplies don't represent a shunt to the circuit.

If you employ a CCS, it has to go in the tube bank it serves- IOW you would need two CCS circuits, one for each half. There is very little current through the two reference resistors (which in most of our amps are a pair of 600 ohm resistors) that reference the circuit to ground. A CCS would have no effect in that part of the circuit, as grid current is all the flows through them. They complete the circuit that allows a conventional driver to drive the power tubes.

The latter is the thing (innovation??) that we brought to the table when we created our OTLs. Prior to this, the Circlotron employed an integrated driver circuit that ran off of the Circlotron power supplies. Its a simple thing, but its what made the circuit practical in a number of ways.

Putting a CCS in the circuit might be a bit of a trick; due to the desired impedances I think a 2-stage solid state CCS is what is needed. It would have to handle the power of the output section without failure should a tube fail.

 

What is your conclusion?, posted on December 18, 2014 at 10:55:58
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
 It sounded as if you were giving me a good reason why a CCS (two CCSs, really) will not do much to "help" the circuit. Then at the end it seems like there is a rationale for the modification, provided it's done right. I am thinking ONLY in terms of the MP1, not any amplifier, where currents in the output stage are comparatively very high.

I brought this up for theoretical discussion, not necessarily because I am about to try it. Sounds like you would not place a CCS where I theorized one could be used, in the negative rail(s), connected to the common junction of all the cathodes of the tubes in the output stage, above the 600 to 1K ohm resistors. Because, as you wrote, very little current flows through those resistors. So, maybe if one takes the output of the circlotron from between the float resistors and the "bottom" of each of the two CCSs (where the output or "top" of the CCSs are in turn connected to the cathodes) that would be the way to go???

 

RE: What is your conclusion?, posted on December 18, 2014 at 12:26:15
Ralph
Manufacturer

Posts: 4769
Location: Minnesota
Joined: April 24, 2002
 The CCS would be in series with each tube in the circuit. In the MP-1, it would be in series with a pair of 6SN7 tube sections on one half of the circuit.

The output would be taken at the junction of the CCS and the cathodes of the 6SN7s. You would have a few hoops to jump through to set up the bias on the 6SN7s, but it seems very doable.

 

Follow-up question, posted on December 19, 2014 at 17:06:57
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
 You wrote, "The output would be taken at the junction of the CCS and the cathodes of the 6SN7s. You would have a few hoops to jump through to set up the bias on the 6SN7s, but it seems very doable."

Where is the float resistor in this topology, below the "bottom" of the CCS (in series with the CCS to ground, as I proposed above) or in parallel with the CCS, such that only the float resistor is connected to audio ground? (That actually seems like the only way to go, now I think about it more.) Sounds like you might be saying that the latter is correct, to connect the float resistor to the cathode(s), AND the CCS to the cathode. Thus the CCS is in series in the circuit, as you say. Do I have that right?

Need a discrete CCS for every cathode?

 

RE: Follow-up question, posted on December 22, 2014 at 09:18:08
Ralph
Manufacturer

Posts: 4769
Location: Minnesota
Joined: April 24, 2002
 Sounds like you got it.

I don't think you need a CCS for each cathode, but it wouldn't hurt.

 

In that scheme,..., posted on December 23, 2014 at 15:53:04
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
 only the float resistor is connected to circuit ground. Which makes sense.

 

RE: CCS in a circlotron???, posted on January 2, 2015 at 13:57:56
Banat



Lew

Basically circlotron power bridge consist from two cathode follower ,
but I would say that in the same time circlotron bridge consist from two dynamically modulated CCS working in opposite phase to each other , where only load(speaker)is common element for both modulated CCS .
I think that implementation of another two fixed biased CCS units inside of circlotron bridge will not bring to you any positive improvements , but only complication with very possible electric failure .
----------------------------------------------------------
BTW , one of this days I want to try with implementation of passive bootstrap network for driver tubes instead of active tube based CCS in tails of drivers tubes ,with major idea to improve driver stage linearity ,what do you think about this ? , and yes sorry for this my off topic !

 

RE: CCS in a circlotron???, posted on January 3, 2015 at 09:47:58
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
 Thanks for your input re the CCS. I had some misgivings about the idea of putting such a high impedance into the circuit. Where John Broskie used a CCS in circlotron, he was imagining a circlotron with one power supply, and I guess the CCS takes the place of the other half of the circuit. (I don't know if he ever builds his schematics; he just puts them out there for others to build, mostly.)

Can you explain the rationale for bootstrapping? Why is it superior to using a CCS in the cathode of each driver section? I am running "Super Linear Cathode Follower" (a la Allen Wright), one each to drive each output tube of my 4-tube 7241-based output stage. The sound of the amplifiers was really improved with just a CCS on each driver section, but it got even better with the SLCFs.

 

RE: CCS in a circlotron???, posted on January 3, 2015 at 11:21:00
Banat
 Lew

Your conclusion is correct , there is no advantage by putting two CCS in circlotron circuit , it is same effect as we put there two fixed high values resistors inside of circlotron circuit .
I not claim that bootstrapping of DC coupled driver stage is superior solution but before anything else I`m must say next :
By repairing and by serious upgrading for one of my customer his one very old but very good sounding SS amp finally I try to replace simple passive bootstrap circuit around voltage gain stage with active one(CCS) which was consist from one BJT , diode and few resistor , result was significant improvement by measuring THD level , but that original smooth and well defined final sound presentation have gone , special on bass spectrum ,after that I was put back original passive bootstrap circuit and everything was fine again.
Guided with this experience I want try to bootstrap DC coupled drivers stage in the same manner , only with passive components .
Main idea is to keep voltage across R1 & R4 Constant ,since if voltage across those two cathode res.is constant than the current is constant to, drivers tubes will work with constant current to across the whole audio spectrum and drivers stages linearity will be improved .
Original Rk with value of 47K want to divide in two power resistor connected in series , R1&R4 will be now 39K and R2&3 around 4K7 ,optimal value of bootstrap capacitors(Cb & Cb`)I need to found by experiment and with listening test to, I will try with many different values for Cb ,anything between 2u2/450VDC and 100uF/450VDC .
Finally I think that by this type of bootstrapping of driver stages some influence exist on the output impedance of circlotron power stage to , I expect some lower output impedance to .
And yes I have no doubt that both CCS and SLCF work just fine by your OTL amps .

 

RE: CCS in a circlotron???, posted on January 3, 2015 at 16:58:42
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
 Your depth of understanding is far superior to mine, of course. Can you explain what's happening in the bootstrap schematic? Funnily enough, I recently modified an old Silvaweld phono stage which incorporated a bootstrap by taking a lead from between two resistors that in series provided Rk for the CF output stage back to the cathode of the previous gain stage. I removed the "top" one of the two CF cathode resistors and the resistor that went back to the previous gain stage cathode and replaced the latter two with a CCS, such that between B+ and ground I have CF-CCS-22K resistor. It sounds good, but is it better than the original design? I do not really know. In fact, I do not really understand what's going on with a bootstrap topology.

 

RE: CCS in a circlotron???, posted on January 4, 2015 at 04:36:26
Banat



Lew

First of all I don`t think that I have any superior understanding over yours how this tube electronic circuit works :) , since many time by exploring some electronics circuits I feel like absolute beginner .

--------------------------------------------------------------------
Second I`m not familiar with that specific Silvaweld phono circuit,but I almost can imagine that circuit and modification that you have done .
Any way here is the close look on simplified schematic with only one half of circlotron power bridge and with only one associated driver tube .
I will try to explain in simple as possible way how that bootstrap technique Ideally have to work .
-If the both values of delta VAC1 and delta VAC2 applied across R1 cathode resistor have same values of AC voltage and the same phase status than VAC voltage across R1 resistor is ideally 0 VAC ,
If the AC voltage across R1 is now 0 VAC that further mean that no AC current can flow(leak) via R1 res. (from upper cathode point viaR1& R2 to ground ), so AC current passing across R1 is now (ideally) O , (no looses.)
This further mean that R1 is now virtually transformed on much higher value than 39K for ACV (ideally on infinite high value), so gain of cathode follower is now improved ,linearity to .
That also mean that current via driver tube is now ideally constant too .

--------------------------------------------------------------------
Of course that this solution need to be proved practically by me , think that first I have to found ideal ratio of values between R1,R2 and Cb by experiment,measurements,and listening test to achieve constant positive effect across full frequency band of interest , and yes all this explanation from above is pretty simplified (just principle of bootstrapping), since in this circuit exist another variables that have to be take on account for the final design .

 

RE: CCS in a circlotron???, posted on January 4, 2015 at 07:33:14
Lew
Audiophile

Posts: 10911
Location: Bethesda, Maryland
Joined: December 11, 2000
 Thanks for reminding me of the concept of "bootstrapping". I now recall that I've read about it in principle in one of the basic books. So, what are the determinants of the value of R1? And what is happening with CBx? You show a sine wave of equal amplitude on either side of CBx, so I guess you size it to pass the audio signal(?) I guess I need to think more before asking more naive questions, because I have a bunch of them right now.

I like your Old Roman saying. It's quite true.

 

RE: CCS in a circlotron???, posted on January 4, 2015 at 09:10:58
Banat
 Lew

I don`t find that any of your questions is naive , but actually is very good questions .

I think that value of R1 + R2(cathode load res.) need to have usual value as for standard non-bootstrapped circuit ,R1+R2=43-47K , that value is mainly determined from actual DC voltage values on bipolar DC supply sources , and in the same time establish driver tube optimum static and dynamic operating point.
To found optimum value for Cb(x) is little more complicated , and need to be carefully chosen to allow equal AC voltage on both side of capacitor over full frequency spectrum of interest, but on the same time to allow near as possible same AC voltage across R1 too .
Some math can be done but there is some variables in this circuit too ,as output power tube(s) transfer characteristic , and load(speaker)resonance points which can influence VAC signal level and phase on the right side of Cb(x) .
So final math will be pretty complicated for this circuit , and I think that easy way for me is to made experiment , measurements and listening test and to empirically found that optimum ratio of values for Cb(x) on the one side and R1/R2 ratio on another side.

 

Page processed in 0.025 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]