Quad KT88 oscillation problem - 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 20th January 2009, 04:06 AM   #1
diyAudio Member
 
chris.busch's Avatar
 
Join Date: Sep 2006
Post Quad KT88 oscillation problem

Gentlemen,

Your help and experience would be greatly appreciated. I've just completed the first of 2 quad KT88 based amplifiers, and I'm unable to surpress a very persistant self-oscillation (with no input to the grids). The circuit utilizes 4 KT88s (Electro Harmonix) in a push-pull parallel, ultra linear configuration (biased for class AB1, but adjustable) with 6SN7s drivers, preamp and phase inverters. The output transformer is a Hammond 1650W (1.9K CT pri.). The power supplies are all choke-input. The B+ supply for the finals is 570 VDC with 2 LC sections (10H chokes and 220 UFD caps) in the filter. The supply is very"stiff" and will easily supply 600 ma with less than 15 V droop. Ripple is <.01% RMS with a full load. B+ for the drivers, preamp and phase inverters is 220 VDC with a single LC section (7H + 440 UFD) which will supply 100 ma with <.01% RMS ripple. The bias supply provides -90VDC again with a single LC filter (25H + 440 UFD) and is adjustable for each pair of tubes. The filaments are run on clean DC as well.

Here's the problem: with no drive (coupling caps from the drivers completely disconnected) and using just 2 tubes in the circuit, it is completely stable at any idling current (typically ~ 30ma / tube, but values from 0 to 100ma are completely solid). If you add the second pair of parallel tubes and increase the idle current to anything over ~20 ma/ tube, the circuit immediately goes into oscillation at approximately 60 Khz. All the tubes appear to be good (any 2 of the 4 will work fine and exhibit similar transfer characteristics). Lead dress is not a problem, the output is driving a solid 4 ohm, 400W, non-inductive load, all the supplies are well-bypassed, and the oscillation is not originating in the drivers. Does anyone have any ideas?

All help would be greatly appreciated,
Chris
Attached Images
File Type: jpg kt88 amp schematic reve.jpg (79.6 KB, 497 views)
  Reply With Quote
Old 20th January 2009, 04:44 AM   #2
diyAudio Member
 
HollowState's Avatar
 
Join Date: Aug 2006
Location: Taxland, New Jersey
Hi Chris,

The first thing I would do is to carefully verify the color coding on the output transformer primary windings by checking the DC resistance and their relationship to each other. If this pans out OK, you may need some stopper resistors for the KT88's. You may or may not need all of these typical values mounted right at the sockets. or 1 watt carbon comps or film.

1000 ohms for the control grids.
470 ohms for the screens.
10 or 20 ohms in the plates.
__________________
"I regard the brain as a computer which will stop working when it's components fail. There is no heaven or afterlife for broken down computers. That is a fairy story for people afraid of the dark." ~Stephen Hawking
  Reply With Quote
Old 20th January 2009, 05:19 AM   #3
diyAudio Member
 
Join Date: Feb 2004
Location: Hamilton, ON
Hi Chris,
You will probably need grid stoppers at pin 5. You may also need to put a resistor in series with the UL tap and each screen. 100 to 470R should do it. You might also try a 1k resistor in series with about a 100pf cap between the screen and the plate taps on the transformer - both phases.
__________________
Dan Santoni
www.dtsaudioelectronics.com
  Reply With Quote
Old 21st January 2009, 12:57 AM   #4
diyAudio Member
 
chris.busch's Avatar
 
Join Date: Sep 2006
Thank you so much for the assistance. I'll implement the changes that you suggested and let you know the results. I really appreciate the expertise.

Cheers,
Chris
  Reply With Quote
Old 21st January 2009, 04:15 AM   #5
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
Stoppers could help, but you might want to go through your circuit and do a stability analysis. It's possible that the additional capacitance of the second pair of output tubes puts a pole right where you don't need it.
__________________
And while they may not be as strong as apes, don't lock eyes with 'em, don't do it. Puts 'em on edge. They might go into berzerker mode; come at you like a whirling dervish, all fists and elbows.
  Reply With Quote
Old 21st January 2009, 04:29 AM   #6
diyAudio Moderator Emeritus
 
ray_moth's Avatar
 
Join Date: Jan 2004
Location: Jakarta
Anothr vote for stoppers, esp. with beam tetrodes - very prone to oscillate.
  Reply With Quote
Old 21st January 2009, 04:35 AM   #7
kevinkr is offline kevinkr  United States
diyAudio Moderator
 
kevinkr's Avatar
 
Join Date: Sep 2004
Location: Boston, Massachusetts
Blog Entries: 6
You might find the screen grids are the culprit here, and I recommend around 100 ohms directly at each screen grid. (pin4)

The only other concern would be if there was enough phase shift in the primary at 60kHz to allow the amplifier to oscillate.

I have had similar problems in the past and I usually found that there was a vhf oscillation at the root of it all, and it was due to feedback at the screens, if this is the case the resistors should take care of it.

Stoppers on the grids are highly recommended and something in the range of 1K - 10K is typical. I tend to favor values around 1K. Again must be installed right at pin 5 as previously mentioned.
__________________
"Extraordinary claims require extraordinary evidence." - Carl Sagan
  Reply With Quote
Old 21st January 2009, 06:21 AM   #8
diyAudio Member
 
richwalters's Avatar
 
Join Date: Sep 2003
Location: Alps:Tube amp designs over 150W, SMPS guru.
Chris; that circuit is a very poor design, I suspect the previous stages are so slew-rate slugged that the audio response is poor to avoid the output stage oscillating. There are no grid stoppers in the screen grid circuits, and the design with KT88's (high gm tubes) will guarantee to oscillate at the leakage resonant freq of the o/p tranny. I've enclosed a sim design with tamed output stage with sim power. Note; 570V B+ and an wild oscillating primary is mighty dangerous stuff, and inadequately rated components will simply puff up. One is up against the leakage/reactive components of the o/p tranny, and snubbing sim to that in the pic enclosed sorted my problem out. The screen to screen is a drastic measure, this highlights the inherent leakage inductance between these taps when compared to primary as the transformer size increases.
I presume you are using 40% UL taps. In reality I used 5-7W series snubber resistors (not wire wound) and min 1kV rated polypropy caps. Considerable amount of power is dissipated in the snubbers so be careful doing 5-10Khz square wave tests at power.This is to be taken seriously.

KT88's are becoming expensive and every method should be used to protect them, a low wattage screen resistor will do alot. Note 630mA fuse used in CT in pic. You haven't given a layout ? This will also play a big part.
Attempting to tame a wild rat isn't easy. Good luck

richy
Attached Images
File Type: jpg snubbers.jpg (83.1 KB, 342 views)
  Reply With Quote
Old 21st January 2009, 07:24 AM   #9
diyAudio Member
 
richwalters's Avatar
 
Join Date: Sep 2003
Location: Alps:Tube amp designs over 150W, SMPS guru.
A pic of some output stage waveform differences. Poorish bild quality but they illustrate before and after.
I wonder how many tube amps look like this ? The poor quality lower waveform goes into oscillation with a LS crossover .. The output stage with the upper snubbered square wave can take another 20dB global nfb before instability occurs. This won't be in every case and depends on tap and quality of o/p tranny.

richy
Attached Images
File Type: jpg dsc00359.jpg (69.3 KB, 303 views)
  Reply With Quote
Old 24th January 2009, 08:02 PM   #10
diyAudio Member
 
chris.busch's Avatar
 
Join Date: Sep 2006
Thank you all for the help.

Success (and very cautious optimism). Added the 1K stops at each of the control grids and the 470s at the screens as suggested. Not only is it now stable, but a cursery look at the output waveform (with a sine wave input) is surprisingly good with zero NF (other than the inherent NF from the untralinear configuration). I'm sure I'm in for lots more tuning once I look at the square wave response and the THD/IM numbers, but for now, it's stable enough to develop further which is great.

One last question: what have you found to be the optimum idle current for a KT88 in fixed bias, ultralinear P-P configuration running AB1 with 550-570V on the plates? The data sheets are all over the place from 30 ma. to 90 ma.

Cheers,
Chris
  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
Quad Golden Lion Genalex kt88 reissues Ang Swap Meet 1 3rd August 2009 11:11 PM
Se KT88 with Quad ESL Drewaudio Tubes / Valves 5 18th February 2007 02:40 AM
CDM-12 Problem - Whistling / Oscillation poynton Digital Source 4 8th December 2005 08:50 PM
oscillation problem ABO Solid State 26 17th June 2005 10:26 AM


New To Site? Need Help?

All times are GMT. The time now is 08:18 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