Please educate me on how to get this VOX V1154 amp up and running

[ausman]

I have acquired a VOX Viscount V1154 guitar amplifier that I am restoring but the power amp section has me stumped. I changed out the original germanium power transistors (because they were both shorted) with MJ15016's and have increased the value of the original Rbe resistors from 2.7ohms to 11ohms and the amp now works, however I an getting some type of distortion at the bottom of the sine wave (see pics). Input signal is about 300mvpp and the output is about 6vpp. Funny thing is if I place my finger on the case of the driver transistor Q301 (see schematic) the distortion reduces a little and the amplitude increases a little. I can provide all the transistor ecb voltages under these conditions if that would help?

[ausman]

Trying to get the other pics posted, bare with me, can you attach multiple pics? because that's what I tried but it didn't work.

[ausman]

The above pic is the input at about 200mv

[ausman]

This is the output at about 6v

[Gyro]

Welcome,

It could be negative feedback instability, the Silicon replacement transistors are probably much faster than the original Germanium ones.

There's a (more attachments) Just under the Browse button. Note the size limits and permitted file types.

[ausman]

This is the output with my finger on the case of Q301

[ausman]

...and here is the schematic

[ausman]

...and here is an overview of the pcb with the 11ohm resistors installed

[ausman]

I will be honest I don't know much about how transistor circuits work let alone the feedback section but I have come here to get an education so let me have it!

[Gyro]

I'm pretty weak on transformer driven transistor output stages too - you could try increasing the value of C306 (across the feedback resistor) though. I suspect that it might require playing with a few other component values to get it unconditionally stable though - Lucky it's a guitar amp, not HiFi!

[ausman]

Forgot to mention the above pics are with an 8 ohm load connected across the outputs. If I remove the load the distortion disappears but I still have a small amount of switching noise in the middle.

[ausman]

So C306 is part of the feedback circuit I assume since it's connected to the output and is feeding back to the input? What value would you start with for C306?

[Gyro]

Best to measure loaded, as long as your ears can stand it!

Yes, R314 and C306 are the global negative feedback path.

Thinking about it, probably play with C306 both ways (try double / 50%). Actually The first thing you could try disconnecting C306 and R314, that will disable global negative feedback. If the signal is then stable (will be bigger too) then that's the problem and you can start playing with values of both parts (increasing R314).

[Gyro]

Forgot to mention the above pics are with an 8 ohm load connected across the outputs. If I remove the load the distortion disappears but I still have a small amount of switching noise in the middle.

The switching noise in the middle will be crossover distortion due to non-ideal transistor biasing (Vbe is higher for Silicon than Germanium). Keep an eye on the temperature of R312 and R313 they will indicate how much output stage current is flowing. You might like the sound of the crossover distortion on an instrument amp. 

[ausman]

lol, my son plays guitar and we tested it with this distortion and he said it sounds pretty cool!

[Gyro]

Job done then! 

[Gyro]

Just a couple of safety checks - for the sake of the speakers.

With zero input, check the DC voltage across the speaker outputs. Also check the voltages across R312 and R313. You don't want to fry the speakers with DC offset or kill the transistors.

[ausman]

offset is practically zero

[ausman]

Yes, disconnecting the global feedback has removed that oscillation type distortion and yes the output increased.

[Ewald1963]

Hi,
I’m no expert, but might be some parasitic oscilllation.
Setting you scope to a faster time base may give a better indication of what’s going on in the distorted portion.

I recommend watching this video bij Allan w2aew: https://youtu.be/81C4IfONt3o

Actually I recommend watching all his other video’s as well :-)
From 7:02 he discusses parasitic ocillation in a bipolar transistor circuit and how to get rid of it using a ferrite bead (or basically by adding some resistance in at least one of the connections of the bipolar transistor).

Where are you measuring the input signal, before or after the driver tansistor? If it is parasitic oscillation it might be in the driver transistor as well.

Looking at the picture of your test setup I also wonder if your test wiring could play a role in what you are seeing on your scope as well. The long leads between the probe and the circuit might pickup noise or create trouble as well.

Hope this is helpfull.

Ewald

[Ewald1963]

Strange, when I posted my comment the last comment I saw in this thread was #10 after posting mine there suddenly appeared 9 more so mine is #20.

[Gyro]

Yes, disconnecting the global feedback has removed that oscillation type distortion and yes the output increased.

Excellent! 

It is feedback instability then. Try playing with a higher value for R314, maybe doublei'sh or more in combination with C306.

Before that, you might want to optimize the biasing of the transistors to reduce crossover distortion a bit (negative feedback will reduce it further). Increasing the values of R307 and R309 (as you already have) a little further, while watching the crossover distortion. If you do then the following cautions apply:

Don't run the amp with R307 and R309 disconnected as the transistors will turn hard on and blow something. Don't get to the point where you see any voltage drop across R312 or R313 - There is no temperature compensation for the output transistors so as they heat up, their Vbe goes down and they could go into thermal runaway. Also, you will start to see offset on the speaker output. Do any tweaking with the speakers disconnected until you're sure there's no offset.

It's up to you how far you want to go in search of lower distortion - Having written the above, it's probably wisest just to get the global negative feedback stable by playing with R314 and C306 and leave it at that - especially if it sounds nice anyway.

(Sorry, bed time).


@Ewald1963:  Good idea - the frequency of oscillation should give some idea of which way the stability is going, particularly in respect to the value of C306.

[ausman]

Wait before you go to bed! I changed out C306 for a .047uf and the oscillation has disappeared and even the switching distortion has cleaned up a lot. I am running 200mv input and getting a pretty clean 10v out. Transistors and resistors are luke warm after about 15 minutes.

You the man Gyro!

[Ewald1963]

Hats of to Gyro indeed! Happy I learned some things from this thread as well.  Bed time for me too.

[ausman]

Of course you didn't think that was the end of this did you, lol.

So in this pic I have a clean output with a 200mV input. However I decided to connect to an 8 ohm speaker (I was using an 8 ohm resistive load previously) and connected my iphone to the input and played some music starting at a very low level and the sound was very distorted. At some point, close to full output volume on the iphone, the sound become a lot louder (of course) but became a lot cleaner. On my scope the iphone output at full volume is about 1V p-p. I think I need to increase the resistance of Rbe to get the outputs switching on with a lower input signal, does this sound right? I don't want to get distortion as a guitar note decays down to a low amplitude.