TC 2400 EQ repair - oscillation at high level

[pvision]

I've been rebuilding a TC 2400 equaliser. In testing I found that it appears to oscillate when overdriven. I have an unmodified 2400 which, when tested, does exactly the same thing

Even with the EQ in bypass the unit will apparently oscillate before it clips. By this point one of the overload LEDs is lit, which should suggest to users to back off a bit, but I wondered if there was design change I should try to stop it oscillating?

The attachment shows a 5 KHz sine wave at the point oscillation starts

Schematic
https://music-electronics-forum.com/attachment.php?s=3544e170ff05afb5b706c05762dbf99b&attachmentid=24109&d=1372981589

Nick Froome

 

[Bo Deadly]

It should definitely not oscillate. I consider that a bug. I would want to be able to put as much signal through it as possible as long as it sounds good.

You have a scope so trace it. Where is it clipping? Trace backward and look for the oscillation.

Does the output gain affect oscillation? If yes, then it's not upstream from there.

It could be those fet gates.

Check to make sure grounding is correct (mains earth to chassis to PS filter cap ground to ground plane of PCB).

 

[pvision]

Thanks for the reply. I'll do a little digging when I have a moment

Some people consider these EQs to be lacking headroom - maybe this is part of the problem?

I am surprised this fault made it out the door

Nick Froome

 

[PRR]

Rail bypass caps.

 

[JohnRoberts]

Rail bypass caps.

+1

JR

 

[gyraf]

Could also be in the CMOS audio switching..?

Jakob E.

 

[CJ]

what are you driving it with?  is the source plugged into the same wall socket as the EQ? 

 

[pvision]

OK, a bit more background. I recapped one EQ and, testing afterwards, found the oscillation. My test was a 5 K sine wave and I pushed the boost on a 5k filter to get the oscillation. I naturally ascribed this to changes I had made (TLE2074CNs rather than the old TL074s)

I did my homework, redrawing the filter section till I almost understood it, and checking the stability of the stages using the calculator at http://sim.okawa-denshi.jp/en/optool.php. This suggested that each stage was stable. Working out poles & stability is a bit above my pay grade but I am trying to learn...

Next I decided to test the unmodified EQ and found exactly the same results, suggesting that the problem is a design fault / oversight (or, of course, a testing error...)

Test methodology was a 5 k sine wave from the DAW, output of the TC was not connected to anything apart from the scope on one channel. Scope & EQ were plugged into the same bank of sockets

This round of testing showed the oscillation occurred at any frequency, EQ in or out, and was gain dependent. The bypass switch is just after the  input gain and before the output gain. IIRC oscillation is triggered by input or output pots

Thanks for your help

Nick Froome

 

[Bo Deadly]

I recapped one EQ and, testing afterwards, found the oscillation.

What caps? Power supply too?

So did you fix it? Not clear.

 

[CJ]

got an iso transformer for the scope?  otherwise the clip lead is gnd,

try some resistance across the output when testing, like 10 K,

 

[JohnRoberts]

I looked at the schematic you linked to...  (TC?)  It is a SVF... Depending on the Q setting the HP output could saturate before the bandpass stage , used for the EQ output.

If you back off level a few dB does the oscillation clean up?

JR

 

[pvision]

Backing off the gain stops the oscillation. It takes quite a lot more gain, after oscillation starts, to clip the sine wave

On the modified EQ I replaced all caps inc PSU & put a decoupling cap between + & - rails at each IC socket

Modified & unmodified EQs show exactly the same symptoms. Oscillation occurs with the EQ in bypass just by pushing the input & output gain pots so it doesn't seem to be related to the EQ section

PSU uses zener diodes, there are no regulators

I don't have an iso transformer for the scope. My test leads were between pins 2 & 3 and the outputs are electronically balanced. I'll try a load on the output next time I test

I will have a look round the pcb with the scope to see where the oscillation is & isn't. Might have a chance to do that tomorrow

Nick Froome

 

[JohnRoberts]

With a SVF and variable Q the gain of the SVF HP stage could be several dB more than the BP output you are looking at... usually we see this as a THD increase but maybe if it needs better bypassing you get instability.... or not

JR

 

[PRR]

...PSU uses zener diodes, there are no regulators...

Zener are regulators.

But back to a point which has been hit twice without impact-- Rail caps!!

There are two 100uFd at the power supply, NO others noted. If this thing has fast opamp (it does), and is bigger than a fuzz-box, that's hardly adequate. It may have been "OK" when those 100u caps were fresh, but they go stale. In any case, common wisdom suggests a couple 0.1uFd every few chips or every few inches.

 

[pvision]

Zener are regulators.

But back to a point which has been hit twice without impact-- Rail caps!!

Messages received & understood. Will report back

Nick Froome