Hybrid Partial Feedback SE Amplifier - 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 13th September 2011, 10:53 PM   #1
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
Default Hybrid Partial Feedback SE Amplifier

Out of curiosity, has anyone out there tried a hybrid partial feedback SE amplifier using one of the IXYS or Supertex depletion mode fets for the front end? Both a Schade-style feedback or an E-linear scheme would be possible. I'm already doing something similar (but all sand-state) in my "SiC Puppy" SE amplifier (see the Pass Labs section if you're curious), and it sounds pretty nice. I'm definitely going to try it some time, but I have projects stacked up to the moon right now, so I won't act on this idea until I finish at least a few of the "works in process".
  Reply With Quote
Old 14th September 2011, 05:53 AM   #2
Jaap is offline Jaap  Netherlands
diyAudio Member
 
Jaap's Avatar
 
Join Date: Dec 2002
Location: Netherlands
In one of the sound practices issues (1994), J.C. Morrison had a design (Dinosaur) with a 2SK30 fet at the front.
  Reply With Quote
Old 14th September 2011, 06:27 AM   #3
diyAudio Member
 
Join Date: May 2007
Why not use "SEARCH" before asking?

Michael Koster has done it the way it should be done! This one is seriously engineered without wasting components:

Most linear triode-strapped pentode
  Reply With Quote
Old 14th September 2011, 06:28 AM   #4
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
If you could dig up that schematic, I'd like to compare it with what I have in mind. This isn't the first tube/fet hybrid I've done. The "Shrine", my very first tube amp, married a triode- cascoded jfet with a ultralinear 1625 output stage. I have a few other sand/vacuum mutant hybrids still in the pipeline.

I haven't seen anyone marrying a naked depletion mode mosfet with a pentode in the classic "Schade" feedback scheme. I've seen jfet/tube hybids used for RIAA input stages before I ventured to do the "Shrine" amp. Given the age of the example you cite, I suspect that may be another example.
  Reply With Quote
Old 14th September 2011, 06:29 AM   #5
diyAudio Member
 
Join Date: May 2007
Check the link in the post above yours. Couldn´t be more "naked"!
  Reply With Quote
Old 14th September 2011, 06:37 AM   #6
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
Revintage - you always have the snarky way of expressing yourself... The Koster circuit uses a mosfet in front, but it wasn't what I had in mind at all. I'll be doing a classic 2-stage Schade feedback circuit using something like a DN2540N5 for the front end (pentode in disguise) or one of the smallish IXYS 1000V depletion mode fets with the beam power pentode of my choice for the output. I may look at cascode vs. conventional front end to see what difference it makes. I really don't care if you think that wastes components or not, after all, they're mine to waste if I see fit.
  Reply With Quote
Old 14th September 2011, 07:02 AM   #7
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
The Koster circuit looks similar to what I've been simulating with a Zen variant I'll get around to building one of these days, except that mine's a class A push-pull (Revintage wouldn't like the component count). I may try his circuit with my "Shrine" amp (I've been musing about converting it to partial feedback for some time) to see the difference between the classic Schade scheme and his. I would still use the jfet cascoded with a triode I used in the original amp as input stage, as that's already in place. I'll try the classic Schade circuit first, as it's the least violent rearrangement of what I already have in place, and I don't have a lot of spare time. Right now the "Shrine" circuit uses a jfet cascoded with a triode at the input, driving an ultralinear-connected 1625 at the output. If I keep that lineup, I'll probably settle for the direct-to plate connection between the input stage load and the output, but keep the AC coupling between stages (it's more parts, but I'm not rebuilding the whole amp). If I want to go all the way, I'll build another amp.

Jaap - I'm still interested in seeing that 1994 circuit if you can lay hands on it.

Last edited by wrenchone; 14th September 2011 at 07:04 AM.
  Reply With Quote
Old 14th September 2011, 07:38 AM   #8
diyAudio Member
 
bayermar's Avatar
 
Join Date: Oct 2007
Location: Tuebingen
No need to start personal disputes...
I actually found the start of the thread very interesting.
Since wrenchone asked for a hybrid with depletion Mosfet and maybe Schade-style feedback, I don't see why Micheal Koster's circuit that revintage suggested does not fit the bill (low component count or not).

Could you (wrenchone) elaborate on why this is not what you had in mind?
I am still on the lower slope of the learning curve, so I always appreciate comments/discussions on a particular circuit - good way to learn.
There are often issues with a particular design, I just did not have on the radar.
Thanks for the effort.

Martin
  Reply With Quote
Old 14th September 2011, 08:53 AM   #9
diyAudio Member
 
wrenchone's Avatar
 
Join Date: Feb 2004
Location: Silicon Valley
I was looking for a simple classic two stage AC coupled amp, pentode input and output, partial feedback, with mosfet substituted at the input.

Koster's amp is an extreme stripped-down version of the same, omitting the AC coupling between stages and feeding completely back into the input stage from the output instead of partially. It took a second look to see that.

Two of the more conventional schemes as compared to Koster's are the "Schade" partial feedback circuit (named after its inventor, the father of the beam power pentode), which has a feedback resistor connected between the plate of the input stage and the plate of the output tube, and the "E-linear" scheme, which connects the plate load resistor of the input stage to an ultralinear tap on the output transformer primary instead of to B+. The Schade scheme was what I originally had in mind when I posted.

Koster's scheme moves that input plate load resistor right down to the output tube plate (the ultimate "E-linear") for even more feedback. If I had waded through an bunch of Schade posts/threads, I might have eventually found Koster's circuit, as it was buried 1/3 of the way into "the most linear triode" thread (a thread I stopped following far enough back I'd forgotten about it). The other innovation of Koster's circuit is the direct coupling between input and output stages.

As usual with DC coupled circuits, it has some compromises. I don't necessarily like blowing 70V across a cathode feedback resistor. The cathode is elevated at 70V above ground as well, which is getting a bit close to the usual 100V cathode to filament voltage limit. In the usual tube amp, it's possible to elevate the filament a bit with a bias network to reduce the heather to cathode potential, so that can be worked around if you accept the cathode resistor voltage drop. The advantage of using the mosfet at the input with Koster's scheme is that it can operate quite comfortably at 30V drain voltage and still have plenty of room for signal swing. This limits the amount of extra voltage dumped on to the cathode resistor of the output stage in addition to what already needs to be there to properly bias the output tube. The mosfet looks pretty much like a pentode in overall V-I characteristics, so it lends itself well to a Schade/E-linear-style feedback scheme, partial or full-tilt. You could potentially try the same thing with a pentode front end, but with a pentode, you would need to run the plate at a much higher potential to prevent it from running out of gas, meaning that you'd have a even higher potential across the cathode bias resistor of the output stage (more voltage/power burned). It may be possible to "starve" a pentode to allow it to run at lower plate voltage, but this would most likely eat most of the gain/transconductace that makes this feedback scheme so attractive. Compromising and using AC coupling between stages would allow you to do a similar scheme, at the cost of a coupling cap and another resistor.

The one big difference between the mosfet and the pentode is that the mosfet is more directly affected by Miller capacitance between gate and drain, not having the shielding effect of the pentode's screen grid. Cascoding the mosfet would help fix that, and might be worth a try - the input would be a bit easier to drive. It might also have better THD at high frequency - remains to be seen.

There are elements of Koster's circuit I might end up using, DC coupled or not. A "hey, other people are trying this, look at this link", would have worked just as well as the "direct" approach.

As I mentioned previously, I'll probably try an intermediate version on an amp I already have built up. I have about 8-9 amps in various stages of construction, so I want to limit involvement just now so I can finish a few things. I want to keep the input stage I already have to avoid completely tearing the amp apart. It uses a jfet/triode cascode on the input stage, which needs some voltage at the triode plate to work properly. The guts are all inside a 6" cube with tubes on top, so DC coupling and an extra 10W or more dissipation inside would not be appreciated. The amp uses a custom SMPS that also fits inside the cube - I don't want to rewind the transformer to make up for the extra voltage drop across the cathode resistor. This means, I'll try the Koster feedback scheme with AC coupling. Not as simple as some would like, but as simple as I'm going to make it for right now.
  Reply With Quote
Old 14th September 2011, 12:10 PM   #10
diyAudio Member
 
bayermar's Avatar
 
Join Date: Oct 2007
Location: Tuebingen
Thanks wrenchone, for going step by step through the mentioned circuit and the one you have in mind and mentioning the problems.
I had noticed the problem of limited voltage swing if the Mosfet would be replaced by a Pentode but did not see the problem of the 70V burned in the cathode resistor (and the cathode to filament voltage limit), which might be a problem in some cases.
Thanks again for the detailed explanation.

Martin
  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
Question Regarding a Partial Feedback 807 SE JPeitzman Tubes / Valves 7 28th January 2011 12:54 PM
Need help on 811A partial feedback WT Tubes / Valves 16 14th July 2010 07:43 AM
Partial feedback confusion Jeb-D. Tubes / Valves 4 22nd April 2008 05:20 PM
partial feedback stuff dave dove Tubes / Valves 3 14th January 2006 11:02 PM
Partial feedback DHP SE amp Fuling Tubes / Valves 21 5th August 2005 09:20 AM


New To Site? Need Help?

All times are GMT. The time now is 12:24 PM.


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