CCS, feedback and grid leak questions. - 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 1st October 2005, 12:25 PM   #1
diyAudio Member
 
bobhayes's Avatar
 
Join Date: Sep 2004
Location: Bungawalbyn, NSW
Default CCS, feedback and grid leak questions.

Hi all,

I've been working through a PP design using 6SN7 splitter (diff. pair) with 6BX7 Constant Current Sink, 6SN7 driver (diff. pair) with similar CCS and 6L6 output (Class A1), with 6CW5 CCS.

I have some questions!

*Do I need to tie the grids to any sort of reference, or does the effect of the CCS dictate the operating points?

*As the circuit is, there are only capacitors between the stages, with no grid leaks or stoppers. Is this OK, or do I need to place grid stopper resistors after each coupling cap?

*Can I put global feed back to the grid of the first valve, mixing with the input signal (it's capacitively coupled to the source) or will this have adverse effects on input resistance etc?

Thanks for advice and suggestions in advance... Bob
  Reply With Quote
Old 1st October 2005, 12:28 PM   #2
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
Grids always need to be returned to some sort of reference. A CCS will let you cheat a bit on the size of grid leaks, though.

Grid stoppers are usually a good idea.

Return back to the grid of the input stage is possible, but can have disadvantages in some applications (e.g., input impedance, noise). The summing point is a virtual ground, so the input impedance will roughly equal the series resistance beteen input and the summing point.
__________________
The more you pay for it, the less inclined you are to doubt it.- George Smiley
  Reply With Quote
Old 1st October 2005, 02:15 PM   #3
diyAudio Member
 
bobhayes's Avatar
 
Join Date: Sep 2004
Location: Bungawalbyn, NSW
Thanks SY,

Can I then derive a reference for the grids by a simple deviding network between B+ and gnd, on which the signal will be superimposed, or is there a better method without resorting to -ve supply rails?

Where can I inject some feedback on a differential pair CCS stage?
I assume that putting feedback to the cathodes (CCS anode) will unbalance the stage, defeating the object of the CCS, is this correct?

Is there any other method of reducing gain and distortion without NFB?
Perhaps local feedback somehow?

Thanks for the help... Bob
  Reply With Quote
Old 1st October 2005, 04:32 PM   #4
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
For me, at least, it would be helpful if you could post a schematic of what you're trying to do and we can start hacking away at it.
__________________
The more you pay for it, the less inclined you are to doubt it.- George Smiley
  Reply With Quote
Old 2nd October 2005, 01:54 AM   #5
diyAudio Member
 
bobhayes's Avatar
 
Join Date: Sep 2004
Location: Bungawalbyn, NSW
Ok, here's the schematic I drew. Sorry about the qulity, but I don't have a scanner or any drawing packeges on this mac!

Please tell me if I've made any really bad mistakes, or and mods that would improve performance.

You said I need to tie the grids to a reference, so where would that be?

Is it worth having the CCS in the output stage? I thought it would keep a perfect balance and eliminate transformer DC flux?

Where might feedback be applied?

Thanks again...Bob
Attached Images
File Type: jpg dsc00321.jpg (57.2 KB, 522 views)
  Reply With Quote
Old 2nd October 2005, 02:26 AM   #6
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
This is going to take some work. Let me hit two things first: the output stage CCS and the grid reference for the 6SN7s and the outputs.

Output stage CCS is something of a fashion at the moment. It's OK if you want to stay in class A (big sacrifice in power), but you have to think through carefully what happens with overload. And in a power amp, you WILL overload the output stage from time to time. Nonetheless, if you want the output stage to run as a CC diff amp, the current sink has to handle a lot of power, and the EL86 you have won't. You'll need a power tube roughly as hefty as your output tubes.

The first 6SN7 would have a grid reference to ground via the feedback resistor and the output transformer secondary, except you don't have the secondary grounded. You've also returned the feedback resistor to the input grid, which is curious. And it will be a big problem. Since you're running the first stage as a diff amp, why not take the input to one grid, then the feedback to the other?

The second 6SN7 has no grid reference whatever. With some juggling, you should be able to direct couple the first and second stages which takes care of that problem. If you go that route, you'll have to run the second 6SN7's heaters separately, biased up sufficiently so as not to exceed the maximum heater-cathode voltage limit. If you have a copy of Morgan Jones's "Valve Amplifiers" (you DO have a copy, don't you?), take a look at how he does this in the Crystal Palace driver stage.

You're also going to run into trouble trying to get the second stage to swing enough voltage cleanly to drive the outputs because of the undersized plate resistors. That's the least of your problems- the biggest is that the output stage grids are also not referenced to anything. If you're going to run an output stage CCS, you want each grid referenced to ground through a moderately large grid-leak (sat 470K or so).
__________________
The more you pay for it, the less inclined you are to doubt it.- George Smiley
  Reply With Quote
Old 2nd October 2005, 01:00 PM   #7
diyAudio Member
 
bobhayes's Avatar
 
Join Date: Sep 2004
Location: Bungawalbyn, NSW
Ok, I made some changes.

I've lost the output CCS altogether in favour of an adjustable bias/ balance configuration, tied the grids down, altered the operating point of the driver stage CCS to use larger anode loads, grounded the output secondary, moved the feedback point as suggested and finally added some grid stoppers!

How's it looking now?

Will it be ok to run 6L6's at 375V in triode mode? The data sheet suggests 250V so I'll probably need more cathode bias resistance (another 100R in series).
I had planned the output CCS to loose 125v and was aware I would need to dissipate more power than the EL86 could manage, but hadn't yet got to it.

Yes, I have Morgan Jones's Valve Amplifiers, but it's the second edition with no mention of 'Crystal palace', however, there is a circuit showing a splitter using LM334Z CCS and EF184 CCS for the driver stage. There is no global feedback used in the circuit.

I built a variation of the 'bevois valley' amp a couple of years back and it's performed flawlessly ever since. Very nice.

Thanks a lot for taking the time to correct my numerous mistakes!... Bob
Attached Images
File Type: jpg dsc00311_1.jpg (61.7 KB, 422 views)
  Reply With Quote
Old 2nd October 2005, 02:07 PM   #8
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
I'll try to give you a more comprehensive answer later when I have some more time, but quickly:

1. Get the Third Edition of Valve Amplifiers. Trust me on that one.

2. You're going to have to think about how you want to get sufficient voltage across the CCSs, along with getting their elements biased properly.

3. You've got essentially 100% feedback. Is that what you want?

More later. Gotta go.
__________________
The more you pay for it, the less inclined you are to doubt it.- George Smiley
  Reply With Quote
Old 3rd October 2005, 12:33 AM   #9
diyAudio Member
 
Join Date: May 2005
Location: Pretoria, South Africa
Hi Bobhayes,

Not to steal SY's thunder (he is doing admirably) but perhaps a few points from me while he is resting:

1. Do not apologise for your drawings; they are quite neat and legible. You should see some of the "back-of-cigarette-box" stuff that otherwise neat sites post - yuch! Then such a man want me to respect his design! But enough.

2.
  Reply With Quote
Old 3rd October 2005, 12:58 AM   #10
diyAudio Member
 
Join Date: May 2005
Location: Pretoria, South Africa
Sorry! I have pushed the wrong key and things went off before I had finished. To continue:

2. I agree with SY that the use of CCSs is sometimes overrated. They certainly have their use, but in some applications the advantage in audio is minimal at quite some extra expense and complexity. For the driver 6SN7 I see no advantage over a 390 - 470 ohm cathode resistor. CCS will only guarantee equal current in both triodes, not equal amplification factor if the units are not equal; for the latter you will still have to balance with a pot between the 47K loads if you want it precise. Also with a valve the about 5V bias needed is too low for proper pentode CCS operation. I have never used it myself, but it seems that even for the quite low anode capability of a 6AU6A at least about 25V will be needed on the pentode anode. (Only transistors will work at 5V).

3. With the input stage you can of course have about 60 - 80V on the cathodes - but then not with the second grid tied to common (through NFB). You will need a separate cathode resistor up top with 2 grid returns to set up proper conditions. In general I would suggest setting up with about 100V - 120V across the tubes and another about 120 - 150V across the anode load resistors to get into the linear region. As SY suggested, I think that unity feedback with this circuit will be too much; you will need further caps and a resistor divider for this. (Here I apologise for not posting a circuit right now. I have the same problem as you.)

3. You ask about triodes: The 6L6GC can work with up to 500V of h.t. (the actual voltage will still be some 40V less because of cathode bias). You will need to keep within the 35W plate dissipation rating for triode mode. But I will not ignore the use of ultra-linear or distributed load mode (screens tapped onto output transformer). To me there is a nett advantage; you have almost triode operation but at tetrode output capability. Data shows that up to the 15W max that the pp triodes can give, UL will give lower distortion.

Then you can peacefully use 470K grid resistors here instead of 100K; it will ease ac loading on the driver 6SN7.

4.
  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
When to use a grid leak resistor? bequerel Tubes / Valves 8 25th April 2009 01:09 AM
Value of grid-leak resistor hkoetz Tubes / Valves 4 11th January 2009 04:15 PM
Power-tube grid-leak bias Jeb-D. Tubes / Valves 7 6th September 2008 01:20 AM
Value of grid leak resistor Klimon Tubes / Valves 4 10th November 2006 04:40 PM
How to determine grid leak current? arnoldc Tubes / Valves 4 10th March 2005 10:34 PM


New To Site? Need Help?

All times are GMT. The time now is 05:54 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