Please Help with a difficult Ampeg SVT PRO 4 repair

[Jazz P Bass]

Good reply, Hodgy.

Yeah, when they blow, they blow.

Enzo has often pointed out that it helps to print out the circuit in question & highlight the failed components.
Then you simply connect the 'current' dots.

[rickybocker]

Hi Rickybocker. Working for Ampeg, I have repaired a dozen or more of these. Some folks on here may disagree with me, but here's what I do when fixing these. For starters, all of the Mosfets. It looks like you have shorted parts on both the + and - side, so you might as well do all 10. When the mosfets go, in more cases than not these are destroyed as well:
10 x .47ohm (the big guys) on the source
10 x 1.5k ohm on the gate
10 x 47 ohm on the gate
Start there, in addition to the stuff in the biasing circuit that you've already seen. I've seen Q106 and Q107 be all wonky even though it bugs out OK. My record was 76 components replaced on one of these bringing it back to life.
And get a variac and ammeter!

Thanks for all the great replies. Please excuse me if it takes a little bit to sort through all of your suggestions. I am ordering a Variac tonite. They dont train us with them at school so I haven't had one available and I have been slowly building up my home bench equipment when I have leftover scholarship money. The first repair I attempted, I replaced all 10 of the MOSFETs, but I only replaced the source and gate resistors I found to be open. I will go ahead and replace all next time. Im going to have to order so more FETs since surely they burned again When the gate resistors fried... again.

[rickybocker]

Good reply, Hodgy.

Yeah, when they blow, they blow.

Enzo has often pointed out that it helps to print out the circuit in question & highlight the failed components.
Then you simply connect the 'current' dots.

I have attached two scans of the schematics power amp section where I highlighted the defective parts with color pencils when I was doing my first round of troubleshooting. Red is defective parts, green shows parts that I replaced even if they tested OK but I still pulled, and blue shows parts I tested as good. This will probably demonstrate the damage better than my original post. I guess I am still not seeing where the short is coming from. I checked components several components down stream and back up from the defective area. I havent checked the opamp IC1 itself...but the resitors and caps in the feedback loop.

I went back through my notes and saw that the P channel FETs werent defective eventhough they are red... but I replaced the entire set of 10 FETs on the A side with matched sets.

[rickybocker]

If you have access to one, a variac with an amp meter is much preferable to a lightbulb limiter. Depending on how it the circuit is set up, you might want to jumper the protection relays if you are using a variac. Do not do your initial test with a load. Since you have a stereo amp with a working channel, you can check for discrepancies as you bring the voltage up SLOWLY and watch the current draw. A lightbulb limiter is better than nothing, but it is not what most bench techs would use in a shop environment. Since you are a serious student with a degree, you might as well do it right if you can.

I am ordering a variac tonite. I dont have access to one at school since we do mainly low current AC/DC there. Ive had to try to take the small amount of theory they have given me to apply it to my quest for becoming a Amplifier resurrectionist..haha. I've havent connected a load to the amp during the tests so far. When you say "jumper the relays" do you mean short them so the common is connected to the Normally open contact? Or keep the relay from closing? As soon as I hear the relay click the smoke comes out.

[rickybocker]

Agree, never hook up a load on the output until you know that the output section is stable.

The main voltages have got to be stable.
That includes the high voltage V+ & V- and also the +15 & -15 volt supplies.

There are a number of very good test point voltages listed (rare indeed).

Try to keep in mind that the output section itself is a balancing act between the two high voltage supplies.
And being a solid state amplifier, with massive feedback, it will not take much in the way of one blown component to upset the section.
If you have one half of the section slamming on (Vdc on the output) the polarity of the voltage can help in deciding what is wrong.

Both the high side & the low side turning on at the same time is a recipe for smoke.

By the massive amount of damage this amp has incurred, I would suspect that it was turned on more than once in the failed state.
You have your hands full on this one.
Dot your I's and cross your T's.
All it takes is one missing copper trace, open resistor or failed semiconductor.

I havent tested with a load on the speaker output if thats what you mean by load. I think me saying "no volume" was misleading. I meant I had all the knobs turned to 0 when I turned the amp on. i wasnt testing with any input signal either... just turned the amp on and watched two of the gate resistors start to glow red and R118 and R119 burn out again. The first time it failed there were several copper traces/pads burned out of the PCB....holding the Q102 and the collectors. I scratched the board to get more trace area and checked continuity after soldering in the new parts, but your right its probably one thing Im missing.

I guess I dont understand how to check if there is VDC on the output if the amp is burning up when powered on. Will it still show VDC on the output if the FETs are failed?

[olddawg]

I am ordering a variac tonite. I dont have access to one at school since we do mainly low current AC/DC there. Ive had to try to take the small amount of theory they have given me to apply it to my quest for becoming a Amplifier resurrectionist..haha. I've havent connected a load to the amp during the tests so far. When you say "jumper the relays" do you mean short them so the common is connected to the Normally open contact? Or keep the relay from closing? As soon as I hear the relay click the smoke comes out.

Many amplifiers have relays for speaker protection, a delay until the amplifier stablizes, etc. if you limit the power supply voltage with a variac, the relay may not energize. Sometimes it is necessary to jumper the relay contacts. More often you can monitor the unswitched signal side with your scope. Depends on the layout.

[rickybocker]

I agree with JM Fahey's approach, making a functional high voltage amp out of the front end, AFTER replacing Q102-thru Q105, and perhaps Q106 & Q107 if at all questionable. olddawg's suggestion of using a variac with ammeter highly recommended, as it allows watching the circuit come on as you bring up the mains, hopefully preventing further failures.

As Jazz P-Bass suggests, any copper trace damage and you've got real problems. I've had trace breakage on the solder pads of Q012 & Q105 before, very mechanically unstable, and you already indicated burnt traces in that voltage gain stage.

Defeating the output stage on the SVT4-Pro isn't so simple. Lifting up the source end of the gate resistors, and lifting up either end of the 0.47 ohm/5W source resistors of Q110 thru Q119 effectively takes the MosFET stage out of circuit. Little room in between the two heat sinks to get at them though. Maybe be easier to just remove Q110-Q119. Yup....not a service-friendly amp.

I've made myself a set of AC Mains extender jumpers (male/female 1/4" insulated fast-on 12" cables) so you can operate the main power amp PCB out of the chassis...with the chassis standing on it's power xfmr side, allowing it to connect the secondary wires to the PCB assy. and, of course, a large insulated surface underneath the PCB assy!!

If the high voltage 'line amp' does power up without failing, you may also want to lift the two zener's D114, D117 cathodes, so they're not in circuit. The current limiter is out of circuit by the 100 ohm connection from Q102/Q105's collectors. I was going to suggest tying those collectors together, and using around a 1k resistor to tie it to the feedback node at R109, R110 & RT112.

Don't load that circuit...run it open load and see that it's working on it's own. Then, you've got something to compare to Ch B, as far as driving open circuit goes.

If you don't have the PCB layout drawing, I have it here.

[ATTACH]32369[/ATTACH]

I am definitely getting a variac to do further tests. I didn't test it with a load yet. Just no input signal and volume and gain knobs at 0. I replaced all the general transistors in the push pull and the all the FETs. There were several pads melted out of the pcb. I scratched and repaired them and checked continuity. But I may have missed something. I really like the idea of pulling the FETs out for further tests. I havent checked IC1's output or testpoint TP4. I think I will start there again when my variac gets here.

Thanks for the PCB layout.. that will be handy tracing the circuit. I dont know why I didnt think of just extending those transformer leads too.. i have just been putting it back in the chassis. and only mounting the ground screw.

So if i just pull the FETs I should be able to check the test points before that stage? Do I need to stop the relay from closing? For some reason even though the amp is failing the relay is still clicking on. The amp was idling fine, the fan spun up for bit, then relay clicks on and then... fireworks.. I thought the comparator IC3 would prevent this.

Thank you for your reply...I've read some of your other post before when I was just trying to wrap my head around what some of the circuitry was for. There is still some aspects of the circuit I dont understand. I may need some more help from you. I was planning to get a job/apprentice at a shop this last summer but the guy retired before I could. Now there is no one in town that repairs amps. And a lot of my teachers don't know any practical knowledge.. just academic circuit theory stuff and lots of digital logic control.

[nevetslab]

beware if you're just buying the MosFET's from distribution, as those will NOT be matched. Ampeg uses matched parts, so they share current properly. Even with the 0.47 ohm Source Resistor, unmatched parts can leave you with uneven current sharing under load, and possibly stressing the ones carrying the bulk of the load. I had posted a thread on that current sharing last fall, I believe.

Oh, having jusr re-read your opening thread...having used matched pairs. Get those from Ampeg, or who, how?

Best of luck....keep us informed how it comes out, eh?

[rocky3chords]

Working for Ampeg, I have repaired a dozen or more of these. Some folks on here may disagree with me, but here's what I do when fixing these. For starters, all of the Mosfets. It looks like you have shorted parts on both the + and - side, so you might as well do all 10. When the mosfets go, in more cases than not these are destroyed as well:
10 x .47ohm (the big guys) on the source
10 x 1.5k ohm on the gate
10 x 47 ohm on the gate
Start there, in addition to the stuff in the biasing circuit that you've already seen. I've seen Q106 and Q107 be all wonky even though it bugs out OK. My record was 76 components replaced on one of these bringing it back to life.
And get a variac and ammeter!

Sorry for jumping in, if you don't mind.

Hodgy, I'm currently repairing one that keeps dropping up the mains switch everytime I flip the power on. Does this shows the same problem of damaged power FETs as well? I disconnected the 2x 50V supply leads, flip the power back to on and no more mains switch dropping. The low power circuits and the relays seem to work fine, as they clicked in time after power is applied.

If yes, is it safe to operate the power stage with only 2 pairs of power FETs at first - and then if nothing's wrong - progressing up to the rest of the pairs. Just a precaution, since installing all of 10 pairs at once would be a really expensive gamble if something else is wrong other than the power FETs .

And yeah, this one is not servicemen-friendly amp, that's a +1 from me!

[diydidi]

Check all zeners in the area of the Fets, as well as all flame proof resistors. Check all Source resistors for open circuit.. Then check for dry joints of the Bridge rect and all jack sockets as mentioned before.