Crate PA8 distortion

[mikeskory]

I'm working on a Crate PA8.  It has a slight distortion through the output section.
I almost didnt aprechiate it at first but as the signal get compicated, you can definitly hear.  Like a slightly blown speaker.  The post you found me at describes the symptom perfectly.

[Enzo]

Ah, that was a three-pager, and I am an old man.

Isolate the problem.  May I assume we have tried other speakers to be sure?

APply a signal to the PA FX return on the rear.  That will be power amp only, still do it?  And you can run the FX send back there to some other amp and speaker for a listen just to verify the mixer part clean.

HAve we scoped the output with a sine wave running through to get an idea what the distortion looks like?  Clipping?  Peaks modulated by ripple?  Crossover distortion?  And does it look the same with no speaker load connected?  (in other words, does having a load alter performance?)

There is a bias adjust on this amp, so we can set it right to the edge of crossover without over warming it.

How ripple free are the main rails?  This thing is like 25 years old, so it may have tired caps in the supplies.

Note there are TWO zener derived 16v pairs, +/-16 for the preamp (mixer), and +VP/-VP for the ICs on the power amp itself.  Are all four of those up to voltage and completely free of ripple?

[mikeskory]

#1 yes I tried different speakers.
#2Used a scope from the mixer stage on pin 5 of the 5 pin connector called JL.  With 300 htz into the phone jack the wave looks good.
#3Checked voltage at this same connector. Schemo says 16 v.  I'm getting 17.5 neg & pos.
#4 Checked voltage at IC1 16 + 15.8 -
#5 The sine wave out of the IC101 pin #1 (TP1) looks terrible. Flattens out on the top and bottom.  FYI The unit is plenty loud!
#6 I'm way older then you

[Enzo]

Yeah well... I am 70 years old, and I feel every year of it.

PS voltages are close enough.

Pin 1 of IC, flattens out, any chance it is clipping at the supply voltages?  The nature of the circuit is that that pin is trying to correct any errors in the amp.  If the output cannot drive things properly, the correction signal will be overcompensating.

Does the speaker out signal look like the pin 1 IC signal?

Q3,8 are your limiters.  You could probably lift them to see if one or both is coming on too soon.

[mikeskory]

Honestly I'm gun-shy about putting my scope on the audio outputs.  Maybe this is a good time to talk about that.  The bench has a new 'isolation transformer' for my signal generator and scope.  Should I test the audio output with my ground on the circuit board ground and the probe at the positive speaker term OR use the scope right across the speaker terminals.

[Enzo]

Is the attached schematic the one you are using too?

Grounding and isolation should not be an issue with this circuit.  The output to speakers is referenced to chassis ground anyway, so I'd just ground my scope to the chassis and probe tip to speaker hot at jack.  The speaker negative is wired directly to circuit ground anyway, so you can clip it there if it is more convenient. 

But one thing we like to do is to view the signal with and without load, it OUGHT to be the same, but if the circuit acts different, that is a clue.

By the way, many people use isolation as you are doing, but I use it the reverse.  All my bench gear is grounded, and the unit under test is on isolation.  That allows me to ground circuits inside SMPS and such with impunity.  If I iso my scope and "ground" it to the V- circuit in an SMPS, then the entire frame of my scope sits at -170vDC with respect to earth.

Isolation is critical for service to switching supplies like that, but in most amplifier circuits, there is a power tranformer, which effectively provides isolation from the mains by its nature.


You seem hesitant about the scope.  Are you concerned over attaching the probe?  Or is it about the grounding?  The probe and scope circuit itself is nothing special, think of it as just another volt meter with a fancier display.  Grounding? Well, just keep it in mind as any other related grounding of a circuit  Remember, everything occurs in a context.

[mikeskory]

With 300 htz going into the 1/4 inch I have a complicated signal at the speaker out put.  Basically its 2 sine waves crossing.  As I turn up the input signal (at the generator) it goes into 1 wave but with distortion.  I'll post up some pictures...

[mikeskory]

Here is the 1st of 3 scope wave shots at output of Crate

[mikeskory]

Shots 2 and 3 from the output of the Crate PA8.  as the input signal is increased

[mikeskory]

I see.  one picture at a time.  Here's the middle one

[Enzo]

OK, the complex multiple waves?  I think that is your scope not syncing.  Try a couple things, first turn up the vertical to make it taller, does that stabilize it?

MAke sure you are on the right channel for sync.  Somewhere in the sweep controls it will allow selection of the sunc.  If you are using channel 2 to display this, make sure sync source is also channel 2  (or channel 1 for both)

And sometimes the frequency you have chosen and your sweep rate are just not compatible, so try turning your sweep rate up a notch or two, it will result in fewer complete cycles across teh screen but might sync them in.  Even just turning the FINE instead of COARSE sweep knob might stabilize this.

But I will wager my lunch money that the multiple waveform you see is not what is coming out the amp, it is a distorted display.

But look at the middle one, the one with the stable sine wave display.  I see fairly severe crossover distortion there.  I suspect that might be what you are hearing as distortion.  Near Q1 on the schematic is P1, the bias trimmer.  Set up the amp for that display, then carefully adjust P1 to reduce that notch in the middle of those waves.  You want the crossover notch to just disappear, turn no further.  In fact I usually back off the tiniest amount.

It is called crossover distortion because it occurs as the output waveform voltage crosses from positive to negative or vice versa.

[mikeskory]

The adjustment worked exactly like you said.  Not a lot of leeway but I found the exact spot.  The customer wanted new power caps so as soon as they are in I'll check the alignment again before we call the customer.  I'll let you know how it goes.

[Enzo]

I want to know if that solved the little distortion you were hearing.

[mikeskory]

Yes it appeared to solve the problem.  I installed the power caps and rechecked the sine wave for.m.  It looked good so I delevered it.  Cusromer picked it up last week.  So I think we got got.  THANK YOU