Effect of cross-coupled "local" FB on plate load of previous stage - Page 2 - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 25th February 2005, 04:32 AM   #11
diyAudio Member
 
Join Date: Oct 2004
Location: Adelaide South Oz
Default KYW

Ditto the above from Ian (Gingertube).
Since we are now thinking correctly - that is current not voltage from the 6SL7 stage is there anything else we can do to improve things.
I was thinking in terms of getting the rp of the 6SL7 up in value to enforce good current source behaviour, or is it more important to choose the operating point/ circuit values for a decent gm?
Cheers,
Ian
  Reply With Quote
Old 25th February 2005, 08:44 AM   #12
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Default Re: KYW

Konnichiwa,

Quote:
Originally posted by gingertube
I was thinking in terms of getting the rp of the 6SL7 up in value to enforce good current source behaviour,
You do that by applying cathode degeneration which is also what sets the gain.

Quote:
Originally posted by gingertube
or is it more important to choose the operating point/ circuit values for a decent gm?
Well, the obvious thing is that our 6SL7 wants to operate with lots (relativley speaking) of current. As the output signal of this valve is a current modulation the quiescent current determines whne this stage will clip and the lower the percentage of the quiescent current the current modulation becomes the more linear.

In other words, "low & hot".

Low & hot automatically also maximises gm and we can always bring down and linearise gm by adding cathode degeneration....

Sayonara
  Reply With Quote
Old 3rd March 2005, 07:23 AM   #13
diyAudio Moderator Emeritus
 
ray_moth's Avatar
 
Join Date: Jan 2004
Location: Jakarta
BTW, I was just looking at Norman Crowhurst's 1956 "Amplifier Design" article, part 5 - Feedback. He sets out to design a push-pull amplifier with feedback. In his design, he uses both an outer global FB loop and an inner FB loop.

The inner loop goes from the 6L6 output tube plates to the cathodes of the preceding 12AU7 paraphase splitter/driver. However, the inner loop appears to be positive feedback, because it isn't cross-coupled like I would expect. Can anyone explain what Mr. Crowhurst was doing here?
  Reply With Quote
Old 3rd March 2005, 08:36 AM   #14
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
konnichiwa,

Quote:
[i]However, the inner loop appears to be positive feedback, because it isn't cross-coupled like I would expect. Can anyone explain what Mr. Crowhurst was doing here? [/B]
Remember, the grid is the inverting input of a valve, the cathode is non-inverting input.

Therefor, in our circuit we need to cross the wires when returning inverse feedback to the driver grids (which IMHO has many advantages) but we must NOT cross our wires when returning inverse feedback to the driver cathode....

Sayonara
  Reply With Quote
Old 3rd March 2005, 07:26 PM   #15
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Default Hummmm

What about hum? Using local feedback from the plates makes the plate to ground voltage linear, but the speaker output is taken from the plate to B+. I suppose running class A will eliminate some of this to first order, but some 2nd order effects will still be there. Class AB will be humming away. Seems like it would be a lot safer to do local feedback from a grounded CT partial cathode feedback winding.

Don
__________________
Without MS, how would we live without all those hackers?
  Reply With Quote
Old 3rd March 2005, 08:21 PM   #16
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Default Re: Hummmm

Konnichiwa,

Quote:
Originally posted by smoking-amp
What about hum?
On the Amp in which I use CCLFB I have between 0.2mV unweighted noise depending on Valve (it is an Amp with between 25-50W Output - depending on output mode) which translates into around -83dbW Noise or -97...100db below full Power.

Quote:
Originally posted by smoking-amp
Using local feedback from the plates makes the plate to ground voltage linear,
That depends upon your view of operation of push-pull amplifiers.

Quote:
Originally posted by smoking-amp
but the speaker output is taken from the plate to B+.
Are you quite certain? Thick carefully before answering.

Quote:
Originally posted by smoking-amp
Seems like it would be a lot safer to do local feedback from a grounded CT partial cathode feedback winding.
Safer in what way?

I see no safety problems with taking the NFB via high value resistors from the Anodes. Do you?

Further, by using a feedback winding you bring the transformer back into the feedback loop, which is a rather terrible idea I might add...

Sayonara
  Reply With Quote
Old 4th March 2005, 12:29 AM   #17
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Well, first of all, are we talking strictly class A operation here?

For class A, one can picture the output xfmr. being driven by the difference between the plates, so clean plate signals would appear fine. But this would probably be better isolated from the power supply ripple if a common mode choke were put into the B+ center tap feed. Ripple on the B+ (with no common mode choke) will still modulate tube Gm and Rp, causing harmonic doubling of the ripple, although only as a low 2nd order effect.

For class AB, no question the output is between plate and B+, including the hum.

On the partial CFB winding, I was just referring to safety from hum. I agree on the pitfalls of feedback windings which include the xfmr. response in the feedback. But driven windings, like partial CFB, are OK.

The cross coupled feedback pathes, for plate feedback resistors, avoid thru xfmr responses only for class A drives in a similar way to partial CFB. But you are also depending on a perfect phase splitter for no thru xfmr effects, since the feedback correction comes from the other side of the splitter. This is just a problem for cross coupled topologies, not particularly plate feedback.

Don
__________________
Without MS, how would we live without all those hackers?
  Reply With Quote
Old 4th March 2005, 09:08 AM   #18
diyAudio Moderator Emeritus
 
ray_moth's Avatar
 
Join Date: Jan 2004
Location: Jakarta
Hey, thanks for sorting me out on that Crowhurst question, KYW. I guess I didn't have my head on straight yeterday.

This weekend I'm going to try out DC coupling again (I did it once before but I got some unbalance in one of the 6SN7 drivers). I want to get my x-coupled FB amp working properly, with sensible sensitivity and a decent bass response.
  Reply With Quote
Old 4th March 2005, 10:40 AM   #19
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Konnichiwa,

Quote:
Originally posted by smoking-amp
Well, first of all, are we talking strictly class A operation here?
Consider a EL34 Amp with a quiescent current of 30mA per Valve and 4K A-A output Transformer. That will give Class A up to 2W, which is for virtually all the music listening except very short and high level peaks. So any hum becomes an issue at High levels and high SPL where the ear is much less sensitive.

BUT, how much hum breakthrough are we talking. Let us assume a competently designed PSU, we should find only a few mV ripple (say 10-20mV) anyway, which compared to full power is still > -80db and the main noise component is a steady 100Hz, not that problematic.

Quote:
Originally posted by smoking-amp
For class A, one can picture the output xfmr. being driven by the difference between the plates, so clean plate signals would appear fine. But this would probably be better isolated from the power supply ripple if a common mode choke were put into the B+ center tap feed. Ripple on the B+ (with no common mode choke) will still modulate tube Gm and Rp, causing harmonic doubling of the ripple, although only as a low 2nd order effect.
Yup. The main question will be, "How much ripple remains on your supply and/or how much makes it through the Amplifier?".

Quote:
Originally posted by smoking-amp
For class AB, no question the output is between plate and B+, including the hum.
WRONG. For Class AB the output is between the anodes up to the point where the Amp transtitions into class B (one valve shuts off. In other words, for small signals the signal output remains between the anodes.

Quote:
Originally posted by smoking-amp
On the partial CFB winding, I was just referring to safety from hum. I agree on the pitfalls of feedback windings which include the xfmr. response in the feedback. But driven windings, like partial CFB, are OK.
I would say what is okay and what not depends on opinion and what one considers "okay". I care to differ, sorry.

Quote:
Originally posted by smoking-amp
The cross coupled feedback pathes, for plate feedback resistors, avoid thru xfmr responses only for class A drives in a similar way to partial CFB. But you are also depending on a perfect phase splitter for no thru xfmr effects, since the feedback correction comes from the other side of the splitter. This is just a problem for cross coupled topologies, not particularly plate feedback.
One may see this the other way around.

I actually can drive (this referrs to "on the bench") the balanced section with it's crosscoupled feedback from a single ended source, referenced to ground and get in both class A and B operation perfect balance at the Output transformer input. Different of viewing the topology I guess.

In fact, both HD spectrum and waveform show minimal differences between SE drive and balanced drive with the kind of circuit I employ (attached below).

Click the image to open in full size.

Sayonara
  Reply With Quote
Old 4th March 2005, 10:46 AM   #20
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Konnichiwa,

Quote:
Originally posted by ray_moth
This weekend I'm going to try out DC coupling again (I did it once before but I got some unbalance in one of the 6SN7 drivers).
You will get this. I originally played with a balance control, but using my own circuit with suitable values things fall well into place and I found that the effect of the drivers DC imbalance on the signal seemed minimal. If I use good quality NOS 6SL7's & 6SN7's there is very little difference (example Ken Rad VT-229 & Philips 6188), with chinese Valves around 50% pass the test for good balance.

To balance in my circuit replace one of the resistors to ground in the 6SL7 stage with a pot & resistor combo giving adjustment from 0.5 - 1.5 nominal resistance. Alternatively, changing the resistor between the cathodes of the 6SL7 to a RC series combo will nail down the offset but adds a largeish value capacitor where I really don't like it.

Quote:
Originally posted by ray_moth
I want to get my x-coupled FB amp working properly, with sensible sensitivity and a decent bass response.
Sensitivity can be quite widely set in my circuit using the resistor between the 6SL7 Cathodes.

Sayonara
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Seeking advice on "Circlotron" cross-coupled feedback OTL design arteitle Tubes / Valves 8 5th March 2010 10:55 PM
suspending a small tweeter in front of woofer for "coaxial" effect? bikehorn Multi-Way 3 2nd January 2006 03:04 PM
Plate load for 6SN7 gain stage? G Tubes / Valves 53 16th July 2003 11:07 AM


New To Site? Need Help?

All times are GMT. The time now is 02:31 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2