Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

March 28, 2024, 10:52:04 AM

Login with username, password and session length

Recent Posts

 

Fried Ampeg SVT-200T

Started by sewage666, July 03, 2012, 01:26:40 PM

Previous topic - Next topic

sewage666

Ha! I SWEAR I checked the continuity of that switch jack... in fact, checking it again out of circuit, the old jacks test fine. They just didn't work... can the metal corrode or something where it starts to act like a resistor? All I know is this is the second time I had this same mysterious problem with old Cliff jacks... test fine, but don't work in circuit.

Moving on...

Fired up the amp again today to update my test points, and +16v rail went down to around 6? +61 rail is okay (tested around +71)... but right after the 1.5kohm resistor (which tests okay), voltage drops with 66 volts measuring across resistor. If the resistor isn't bad, then where is that voltage going?
Measured D20: measured 1.7 one way, .3 the other. Bad diode?! Could I have shorted that out?

Tested
Q6: good
Q7: good
Q8: good
Q2: good
Q3: good

Moving on and waiting for some parts in the mail...

sewage666

Also just noticed... Q6 (bias transistor) is listed as 2N3402 on schematic, but the actual part in my amp is a 2N3403. Could be equivalent?

Roly

Sorry, I mis-read the TP table; the DC voltage at TP6 and TP7 should be +1.45V and -1.45V, not +/-40 as I previously said (that's the fully driven AC voltage at these points)

Quote from:
Ha! I SWEAR I checked the continuity of that switch jack... in fact, checking it again out of circuit, the old jacks test fine. They just didn't work... can the metal corrode or something where it starts to act like a resistor? All I know is this is the second time I had this same mysterious problem with old Cliff jacks... test fine, but don't work in circuit.

Yes, all jack contacts seem to be prone to this.  One trick is to give them a good scrub with the edge of a bit of newspaper to clean any crud off, then they work fine again, for a while...

+16V rail needs to be 16 volts or very close to it.  6V implies it is being heavily loaded by something. The output stage depends on the +/-16V rails being about right, so this won't help at all and you need to find out why it's so low.

D20:  presuming you didn't get your fingers into the act, and you measured it with one end disconnected, this is very wrong - should measure about 600mV one way and open circuit the other.  You can't get a reliable reading in circuit (because of C10).  But these measurements won't give you 6 volts on the supply either.

I'm inclined to think there is actually nothing wrong with D20 or R24.

Now previously you had about the correct voltages on TP4 and TP5, and more importantly TP10 and TP11.  If the +16V is now only +6V then something on the +16V rail is drawing too much current, and the most likely suspect is one of the op-amps powered off that rail.

Pull the one you have socketed (IC1?) and see if that restores the 16V supplies.  IC1 is not needed in place to measure the DC conditions of the output stage.

The only difference between the 2N3402 and 2N3403 is the gain range, 75 to 225 against 180 to 540, so they overlap.

I think we need a fresh set of DC voltage readings for all the Test Points.
If you say theory and practice don't agree you haven't applied enough theory.

sewage666

QuoteD20:  presuming you didn't get your fingers into the act, and you measured it with one end disconnected, this is very wrong - should measure about 600mV one way and open circuit the other.  You can't get a reliable reading in circuit (because of C10).  But these measurements won't give you 6 volts on the supply either.

I'm inclined to think there is actually nothing wrong with D20 or R24.

I did all the measurements out of circuit. R24 is fine. D20 tested funny, so I compared it D21, which tested good. I've got a batch of 1N4745As on their way to me now.

QuoteNow previously you had about the correct voltages on TP4 and TP5, and more importantly TP10 and TP11.  If the +16V is now only +6V then something on the +16V rail is drawing too much current, and the most likely suspect is one of the op-amps powered off that rail.

Pull the one you have socketed (IC1?) and see if that restores the 16V supplies.  IC1 is not needed in place to measure the DC conditions of the output stage.

Luckily I socketed both ICs, because I had the same thought... but unluckily, it didn't make a difference. I traded around ICs in both. My TL072s were recycled from some random circuit boards I found in a salvage store, so their reliability is suspect. I've ordered NE5532s.

When I get these parts replaced that I know are variables, I'll be taking new test point notes. Or, maybe I'll be playing loud bass through my now working amp. I'd prefer the latter!

Roly

I'd be checking that you have good +/-16V rails before you plug the op-amps in.
If you say theory and practice don't agree you haven't applied enough theory.

sewage666

And, I'm back.

I got my parts in the mail... I replaced both diodes on the 16V rail, D20 and D21. Both, just because I thought I should. I got my +16V back, I assume the diode was bleeding to ground.

I plugged in my two new NE5532 ICs... and nothing. Well, not nothing. I ran the sine wave back into the power amp line-in, and it definitely was coming through. It's coming through about the level before I replaced the shorted jack. Very quiet. I tried the same trick to short the jack connection, in case that was the problem, but no effect. I also tried to short the speaker jack, no result. Messed with the biasing resistor (and put it back to where it was), no result.

I measured across the .33Ohm 5W resistors and didn't get a reading. They're stone cold, too.

I re-measured all my test points, and there is definitely still some discrepancies from the table. Namely 3 (giving me a negative DC voltage reading) and 14 (way off from the table)

Again, I'm stumped. I'm gonna approach this again with daytime eyes, but maybe someone else has a thought? Thanks. New schematic and test point results attached!

Roly

{this drawing seems to double in size every time you upload it. (???)}

You need to be specific about what voltages you are seeing where - you are our eyes here.

You appear to have added two voltages at the ends of R30 which are most interesting; +5.2V and +5.4V.

Now if we start at the half-rail (point (E)), and we apparently have no current flowing in any of the output emitter resistors R33-R38, then we have no voltage drop across them.

The typical voltage drop across the base-emitter junction of a transistor is around 600mV or 0.6V, so this is what the bases of Q13-Q15 should be (similarly for the negative side).

Q11 adds another 0.6V so the voltage at the base of Q11 should be;

2*0.6=1.2V.
3*0.6=1.8V

TP6 and TP7 are given as 1.45V+/-20mV

1.45/3=0.48333333

So let's compromise and estimate around 0.5V across each of the three EB junctions in series.

If your voltages are right then there has to be an open Base-Emitter junction in both the upper and lower path, e.g. Q11 and Q12.

We have two choices; mistaken measurements, or some dead silicon between Q11 and Q18, and I'll bet on Q11 and Q12.

What is a serious concern here is that the TP table says you should have only +/-1.45V at TP6 and TP7 (i.e. across Q6), and until you determine why there is so much voltage across Q6, almost 12 volts when it should be closer to 3V, simply replacing dead drivers Q11 and Q12 will only fry the new ones.

Also, note that the voltage at TP3 cannot be corrected if the loop is open, and if there are transistors with open BE junctions in the loop then TP3 could be just about anything, so don't worry about that for the moment.

Important point; can the voltage between TP6 and TP7 be brought down to +/-1.45V by adjusting AR1, and if not we need to find out why and fix that before anything else.


BTW, you said "I also tried to short the speaker jack" - I hope I misunderstand because that is something absolutely NOT to do at any time.

So,
* post the table of DC voltages for all your Test Points.
* Does adjusting AV1 bring TP6 and TP7 to about 3V apart, or a bit less?
If you say theory and practice don't agree you haven't applied enough theory.

sewage666

Oh boy, that's my bad!! Those voltages aren't on either side of r30, those are TP6 and TP7. I realize now I wrote them at a different spot before... Sorry! I wiped clear my old readings and just re-added each point.

Roly

Well okay, but that doesn't really change anything much; you still have +/-5 -something volts where you should only have +/-3V (2x 1.45 = 2.90), and it just means that another couple of transistors are suspect.

You still need to establish if altering the bias control will get these TP's to a difference of 3 volts or a bit less, then you have to move down each line of B-E junctions getting lower voltages as you go.  The emitter of Q9 (base of Q11) should be about half a volt lower than TP6; the emitter of Q11 (bases of Q13/14/15) should be about half a volt lower again.

Similarly working along from TP7 on the negative side through Q10, Q12 and Q16/17/18.

At one of these transistors (on both the positive and negative sides) you will find a base-to-emitter voltage much larger than half a volt, and those devices are dead.

But as I said before, there is no point in just replacing them until you find out why TP6 and TP7 are so far apart, and getting the voltage between them back in spec - +/-1.45V.
If you say theory and practice don't agree you haven't applied enough theory.

sewage666

I hadn't seen your response until I'd gone mad and decided I needed to test every diode!!! Well, I tested them in place, and any that didn't read normal I tested out of circuit as well. Of course, that was just about every diode. Fruitless search?

Not necessarily... I found D10 and D13 were passing no voltage... these are connected to the bases of Q7 and Q8. Also suspiciously near the original resistor burn out zone. D12 also tested funny (was reading as passing 2x as much voltage as other 1N914s) so I'm just going to replace that whole array of diodes there.

My parts are already on there way via snail mail... When I get them put it in, we'll again 1. see what happens and then 2. I'll start tests down the rails as you've suggested.

Thanks a lot for your help and patience, Roly!

Roly

I'm confused by your "D10 and D13 were passing no voltage".  Components pass current and have voltage across them, or between points.

Specifically did you find D10 and D13 open or shorted?

Shorted I hope because open wouldn't explain the symptoms.  The diode test range on a DMM displays millivolts, thus a good diode will look like an open circuit one way, over-range in reverse; and somewhere between 500 and 700mV the other, forward, way.  A shorted diode will generally read something low or quite low, less than 100mV, both ways, while an open diode will read over-range both ways.

Patience is something you must have if you are going to solve these sorts of problems - they simply don't yield to macho gust and bust but to thorough and methodical.
If you say theory and practice don't agree you haven't applied enough theory.

sewage666

Quote from: Roly on July 12, 2012, 06:47:45 AM
I'm confused by your "D10 and D13 were passing no voltage".  Components pass current and have voltage across them, or between points.

Specifically did you find D10 and D13 open or shorted?

Sorry, nomenclature fail... There is no voltage across them either direction, so they're passing no current. They're shorted.

Quote from: Roly on July 12, 2012, 06:47:45 AM
Patience is something you must have if you are going to solve these sorts of problems - they simply don't yield to macho gust and bust but to thorough and methodical.

Do I get points for being pissed at every diode, but methodically testing them all in order? :)

Roly

Quote from: sewage666
There is no voltage across them either direction, so they're passing no current. They're shorted.

That's good because that makes sense with the voltage readings.  :tu:

Now you may think I'm being picky here, but unless you think about the properties of these components the right way in terms of resistance, voltage and current, you are never going to make sense of what is going on.

In this case they are shorted, like a length of wire, they have no voltage across them - but does a length of wire, a short, pass current?


D10 and D11 aren't in this particular part of the circuit to prevent current from flowing backwards, but rather because they have a fairly constant voltage drop independent of current; in other words they are acting here like very low voltage zeners to provide a fixed drop of 2x 0.6V (roughly).

Because one on each side is shorted the effect is to turn Q7 and Q8 on too much, giving too much current though Q6 and therefore excessive voltage between TP6 and TP7.

Hopefully when you get good diodes in those positions TP6 and TP7 will only be about 3 volts apart.

Yes, you do get +1 for methodically testing them all in order.  :dbtu:

This is one hell of an amplifier you are servicing, a bit like learning to drive in a tank, and you are gaining ground, but with an amp like this I would "hasten slowly" myself, so you are doing fine.  ;)
If you say theory and practice don't agree you haven't applied enough theory.

J M Fahey

Agree and add:  those diodes *are* passing current.
They show no voltage drop because being shorted they "have no resistance".
Please be careful with the use of words because we won't understand what you mean and you will not understand our instructions.
Good luck.

Enzo

I think part of the word problem is that for many people the words voltage and current both mean the same thing:  electricity.