SCA-35 rebuild

I obtained a SCA-35 chassis that was minus the tubes, transformers, faceplate, knobs and power cord with the intent of rebuilding with Dave's EFB PS board, PC11 preamp board and PC10A output boards as well as DynaKits power and output transformers. I finally cobbled it all together, fired it up on a DBT and variac using a dummy load and all seemed ok. Applied full input power and set bias with no indication of a problem. However, when hooked up to speakers I get a squeal at turn on that lasts approximately 5 seconds. It starts out low in volume and increases the whole time. It sounds as if a capacitor is charging with the noise going to both channels equally. Balance has no effect. After 5 seconds or so it stops and music plays as if nothing happened. Knob positions also have no effect and it occurs with or without an input connected. Turning volume to max level does not cause it to return. Any ideas before I start troubleshooting would be appreciated.

Hi --- Many years ago, I had the squeal/beating thing also on my SCA-35 during power up. Disappeared after about 5+ seconds.   In my case I think that I changed out my 7199 tubes and the problem went away?  (Either the 7199's or the 12AX7's.)  

Try the 7199's first and see if the problem goes away.

evoroadster wrote:I obtained a SCA-35 chassis that was minus the tubes, transformers, faceplate, knobs and power cord with the intent of rebuilding with Dave's EFB PS board, PC11 preamp board and PC10A output boards as well as DynaKits power and output transformers.  I finally cobbled it all together, fired it up on a DBT and variac using a dummy load and all seemed ok.  Applied full input power and set bias with no indication of a problem.  However, when hooked up to speakers I get a squeal at turn on that lasts approximately 5 seconds.  It starts out low in volume and increases the whole time.  It sounds as if a capacitor is charging with the noise going to both channels equally.  Balance has no effect.  After 5 seconds or so it stops and music plays as if nothing happened.  Knob positions also have no effect and it occurs with or without an input connected.  Turning volume to max level does not cause it to return.  Any ideas before I start troubleshooting would be appreciated.

This is how sovtek 7199 works. Change to real 7199 and it works fine.

First, thanks for your input.  The squeal occurs even with both 12AX7 tubes removed.  The 7199 tubes have been replaced with 6GH8A tubes as a result of using Dave's PC10A output boards.  Swapping the 6GH8A tubes into different combinations results in squeal in whatever channel is being driven.  I am leaning towards the EFB PS boards C4 or C24D.  The only thing common to both channels seems to be the EFB bias circuit.  I first need to check the EFB board voltages both during startup and after squeal goes away.

All EFB voltages well within spec.

evoroadster wrote:First, thanks for your input.  The squeal occurs even with both 12AX7 tubes removed.  The 7199 tubes have been replaced with 6GH8A tubes as a result of using Dave's PC10A output boards.  Swapping the 6GH8A tubes into different combinations results in squeal in whatever channel is being driven.  I am leaning towards the EFB PS boards C4 or C24D.  The only thing common to both channels seems to be the EFB bias circuit.  I first need to check the EFB board voltages both during startup and after squeal goes away.

All EFB voltages well within spec.

When using the sovtek 7199 it squeeled as yours. Reason was that it had higher amplification
and that the circuit has an element of positive feedback.

Look at R33 which channels an in-phase signal to the cathode (and g2) of the pentode.
The cure suggested was to eliminate the positive feedback by turning the C18's left connection,
the one that connects to the cathode pin6 to ground instead. Thus the voltage to g2 is
kept (it also has a feedback chain, but this chain is to give g2 a suitable voltage to keep
the cathodyne stage in balance.

Do you by any chance has an 6u8 to test with before moving the C18 ?
I'm curious if the squuel is there with 6u8 too.

Being this is my first foray into tube amps I have no other tubes to swap in.  I will order a pair to try and in the meantime deal with the squeal.  I will let you know what happens with a tube change but it will be a week or so.


A matched pair of low noise, low microphonic 6U8A tubes are on order.

An additional observation : Classic Valve Design sells among oher stuff new boards for sca35 that
is adapted to 6gh8. On page 2 is a revision note "6-17-2013"  that says
R37 is changed to 30k ( was 82k) and has a 150pF cap in parallel.  

( this increases NFB for the amp, and it _might_ prevent the squeel. I am out of adapters but will
dust off my sca35 and do some testing for myself when i get new adapters)

Dave gillespie also has boards for this purpose, he has also some mods ( not identical)
See http://www.tronola.com/PC-10A_Installation.pdf for schematics,
one diff is a lowpass filter C1/R38 another is 10k resistors between eyelet 8 and the
grid on V6.

I'm sorry not being able right now to reproduce your problem. But i will in due time.

peterh,

I am already using Dave's PC-10A output boards as well as his capacitor board with EFB and his PC-11 preamp board. I will have a set of 6UH8A tubes early next week.

Something else to consider is that you are using Dynakit OP transformers that have 8 ohm and 4 ohm taps while the original transformer was 16 ohms and 8 ohms with the feedback taken at the 16 ohm tap. It is possible that some alteration of the feedback related components on the PC-10A may be required with the Dynakit transformers.  The SCA-35 feedback arrangement is unusual, and quite finicky. The original circuit was really only stable with a "real" 7199. For a time, Sovtek made a "supposed" replacement labeled as such, but it was not actually the same. Although the Sovtek tube performed acceptably in many applications, this was often not the case with the SCA-35, and the squeal on power up was fairly common.

Note that Dave's PC-10A is basically the same board previously sold by Audioregenesis. It has some modifications allowing it to operate with either 6GH8 or 6U8 driver tubes while still being stable. A large number of each of these tube types, from various manufacturers, was employed during development of the modifications. Many of these boards were sold by AR over a number of years with zero reports of stability issues. However, development was done using only the original Dynaco transformers with the FB taken from the 16 ohms tap, for which the PC-10A was intended. The AR boards were only sold fully assembled and tested, while Dave sells only bare boards. I will assume you have double checked your assembly for errors?

This seems to be confined to the pentode/triode driver. if the amplification is large enough in the
pentode it will oscillate. A higher B+ could be the triggering factor , when B+ lowers ( as the
el84 heats up) amplification comes below the threshold for oscillation.
Thinking about the circuit, instead of grounding C18' left side, a resistor of , say 47k in series
would reduce the feedback enough to eliminate the oscillation.

A test for this theory is to remove the el84 and attach a scope to the el84 grid connections.
If the oscillation continues (instead of dying out) it confirms that it is the higher B+ that
triggers the oscillation.

From what I have been noticing I think you may be on the right track. I don't have the values in front of me at the moment but the values at the power supply board are initially above what is expected and they then decrease to the expected range. Power xformer input to the board is spot on and steady.

I will scope the el84 grid tonight.

Well no need to go any further. The 6U8A tubes showed up today. A big mahalo to tube depot...great prices, reasonable shipping, outstanding response.

Swapped in the 6U8A tubes and no squeal. Moved the amp into the house and will run it on and off for the next few days to make sure everything is OK.