Distortion puzzle?

[ITAM805]

This guitar amp has an odd fault that I can't fathom. In the attached images you see an undistorted signal from V1 anodes, but when it reaches the V2 grids (pins 2 & 7) it has it +V peaks clipped even though it's going through passive components?

It's has 500mV 1kHz sine into V1 inputs, then it is switched between either ch1 or ch2, the latter via a passive EQ stack.

The opto-couplers I believe are 'VTL5C1' types which are sometimes used in guitar amps as soft switches

Any idea guys?

Many thanks

[ms660]

Series impedance and +ve grid is probably why V1 anode is ok and V2 grid is clipping.

500mV input seems a bit high to me, might be causing the clipping.

Lawrence.

[ITAM805]

Thanks Lawrence.

Humbucker pick-ups can easily kick out a volt hence the test signal, certainly the issue disappears with a lesser voltage (100mV) but the amp sounds awful at 'normal' kerrang levels. Not sure I understand the +V on the grids though?

edit: I should say the 500mV signal is going into the low gain input

[ms660]

If the grid goes +ve the grid and cathode will act like a diode and the +ve peaks of the signal will get clipped.

Lawrence.

[ionburn]

Not something I have been involved with for some time, but high output pickups will easily overload many amps. That is perhaps how they get to distorted sustain. What is shown looks like a bias issue, but may be at the limit of clean input. I agree with Lawrence. What do the amp input specs say?

[ms660]

According to some calcs from the schematic info V2 looks like a ECC83/12AX7, the bias voltage is approx. 1.5 volts therefore a signal on the grid of over 3 volts peak to peak will suffer from clipping.

Lawrence.

[ITAM805]

Lawrence, that's interesting, not something I've seen or heard of before? Thanks for the info in #6, I'll have a check of the bias around V2 (yes 12AX7)

Ionburn, nothing in the specs about input sensitivity of this amp although I'm guessing it's much the same as most of the valve types out there. The distortion is nasty, certainly not the intended 'blues' sound of a £1k amp for sure

edit: both cathodes of V2 measure 1.34V, as your calculation suggest Lawrence I'm measuring 4.5V on pin 7 of V2 so there's your answer ! That said, to eliminate the clipping I have to reduce the input signal to something around 70mV

[ms660]

Originally Posted by ITAM805

Lawrence, that's interesting, not something I've seen or heard of before?

http://www.valvewizard.co.uk/Common_Gain_Stage.pdf

Read it all when you have time....

...for clipping look at 1.13 (cut off clipping)
and for the type of clipping being discussed so far...1.14 (grid current clipping)

Lawrence.

[ITAM805]

Thanks Lawrence, that's certainly a comprehensive article

So now all I need to discover why this amp sounds like

[Silicon]

I would check C104, C105 and C106 for leakage.

[Argus25]

Just a few points:

This type of clipping (grid current)will always occur to some extent if the drive levels are high enough. However, for a continuous test signal it often gets ameliorated to some extent by the grid coupling capacitors charging on peaks and the average grid voltage shifting negative. So that is why its worth checking the coupling caps for electrical leakage as suggested by Silicon because this would push the grid voltage positive. Likely if that was happening though the anode voltages would be low and the cathode voltages high too.

In most amplifiers though, what is supposed to happen is that all the drive circuits (stages) leading to the output stage have enough dynamic range that the output stage is driven into clipping (which is usually fairly symmetrical) long before any of the input stages are over-driven into clipping. Though, in the guitar amp world, clipping or overdrive at the input stages has been favored by some because of the sound of the distortion. But it is a general principle for an amplifier that the driver circuits should be able to push the output stage to clipping, before they clip themselves. If this is not the case, the full capability of the output stage is not utilized.

In any case, one thing to do is, feed the amplifier's output into a dummy speaker load with the scope across that. Check to see if the output stage clips first, before any of the driver circuits clip.

[ms660]

If it's both halves of V2 that are clipping, then C103 needs to be on the list as well if the capacitors are suspect, a no signal DC voltage check on the grids should confirm.

Lawrence.

[ITAM805]

Testing the caps mentioned (actually the pots attached to them) I get a spurious pulsing of 0-20mV, no sure if it's just stray pick up though? The grids of V2, p7 -20mV and p2 0V.

Going back to Lawrence's observations it would seem that the amp is being over driven from V1. With my Strat which has rather low output pickups the amp sounds quite raucous and rather 'buzzy' at full tilt. With an SG Standard fitted with somewhat hotter humbuckers it sound horrible and over driven

Argus25, I tested the emission of the valves, all good, the 2x EL34's being matched within 12% of each other, good enough for rock'n'roll. I ran the amp up to full power until just before the onset of clipping. The trace was clean and symmetric, 36W into 8R dummy load. This is supposedly a 40W model but the board states 30W so take your pick ! This was with the input at around 100mV, I wound up the generator until clipping while checking the signal at the various stages

So here's rub, it looks ok with a sine wave but sounds crap with a complex signal?

[ms660]

What's the Gibbo like with it's output backed off by it's own controls?

Lawrence.

[ITAM805]

It's a bit late in the evening to try it now Lawrence, neighbours and all , but I imagine that would clean things up. I need to speak to the owner and see about what guitar, pedals and so on, did the amp suddenly start doing this or was it always like it etc,?

[Argus25]

It is possible to have an amplifier design where all of the pre-amp and driver stages are running well inside their dynamic range as the power output stage goes into clipping, but just one of them is much closer to the edge. Then with a wide dynamic range signal like that from a high Z guitar pickup there can be peak clipping in that pre-amp or driver stage. If that is problematic and alters the sound in an undesirable way, that stage can be re-designed for a wider dynamic range before clipping.

For example, in this case, one thing you could do is to change the valve in that stage from an 12AX7(ECC83) to a 12AU7 (ECC82). The cathode & anode resistor would need altered values as this valve runs a higher plate current. Typically the bias or cathode voltage when set up for this valve is about 8 or 9 volts, not around 1.5 to 2 volts for the 12AX7. So it can tolerate a higher grid voltage signal swing before grid current occurs. But the gain would be lower, but it might not matter.

I would only consider a mod like this if I was 100% confident there was no actual fault going on and I was unhappy with just the behavior of the one stage.

[ITAM805]

Thanks Argus, I'm pretty sure there must be a fault so modification shouldn't be necessary.

I fitted a new 12AX7 in V2 and all the resistors are spot on. The cathode voltages have increased slightly to 1.42V but the HT is down to 178/176V, the sheet says is should be 205/200V, and the supply on the hot end of R112 should be 350V but is 310V? This will affect the headroom somewhat although in my experience the specs often dont tally with the reality?

[ColinTheAmpMan1]

While output-stage clipping gives "better-sounding" distortion, it isn't always true to say that distortion in the pre-amp stages isn't going to happen. Indeed, just feeding in a bigger signal (from a humbucker compared to a single-coil pickup, for instance) can get you into distortion in guitar amps. This is why some amps had master volume controls, so that a distorted pre-amp signal could be amplified, but at a relatively low volume, by the output stage. Another way of doing this could involve a signal booster pedal between the guitar and the amp input.

Could you tell us what make and model of amp this is? (If only so that we can avoid them ).

Colin.

[ITAM805]

Hi Colin, it's a Rivera Chubster 40. As chance would have it another Rivera has turned up, different model but it has the same front end but a burnt out o/p board. So I should be able to compare and contrast up to a point