Ampeg B-100R Melting Bridge Recto

[SYSDOA]

Hi Guys,

Wanna say thanks to everyone who's helped in the past and the forum in general. It's been something of a bible to me over the last couple years, mucho appreciated-o! I have a Ampeg B-100R that came in for intermittent volume drop. Everything looks normal, couple of questionable soldering points. The soldering on the bridge rectifier looked a little goofy. Cleaned it up. Powered on and showed no symptoms. Started checking voltages to confirm stability and shorted accidentally between R54 and R57. They're really close together and I kinda had a late night prior, lil shakey... Anyway, fuse blew and now won't stabilize. I replaced the output transistors (same batch but unknown regarding matched er not). and brought up slowly over the variac to 60VAC. The cements began to smoke. Tested and ok. After doing this a couple of times I noticed the bridge recto contorting different shapes and spitting out plastic. Pretty cool looking but not a good thing in my guestimation.

I've done stupid stuff in the past and will continue to do so, but I'd like to understand why this happened and how I can correct it. I work on quite a few crampeg circuits in our shop and from time to time a stinker of a situation pops up like this. I'd like to progress through this one gracefully. So what could I focus on to get the amp stable again? Why did shorting the two resistor leads cause such an result?

Any help, direction, guidance, stiff talkin' to, is much obliged.

[R.G.]

First, before anything else, get yourself a lightbulb limiter put together. Not least because USA law makes the incandescent light bulb phase out next year, but also because it will self-limit the current, indicate it's condition by the light, and stop you blowing fuses and melting things.

Next: figure out what's blown. This is not as hard as it sounds because you can start at the transformer working your way toward the amp. It's clear that something is causing too much current to flow. It's also clear that it's (at least) going through the rectifier. It's not clear where it goes after that. So divide and conquer. Open circuit the transformer secondary and carefully insulate the leads so you don't get an accidental short. If the fuse stays intact when you power just the transformer (or better, if the light bulb settles down to a very dim glow or dark) the power transformer is not killed, which is one possibility if it's had to be melting rectifier bridges. If fuses/light bulb say OK, then carefully measure the secondary voltages. If they're OK, you're in luck and the transformer is still with you.

Then replace the rectifier bridge and open the wires to the first filter caps. Power it on - gently again! - and see if DC appears right at the rectifier bridge. If so, power off, hook the rectifiers up to the first filter caps, and UNhook the filter caps from the rest of the amp. Bring it up gently and see if DC comes up nondestructively and non-over-currrently on the filter caps. They may be shorted. If they are, replace, retest, until you get DC coming up on the first filter caps. Keep walking your way, one component section at a time, towards the power amp, all the while leaving the fuse/bulb/variac there to keep from killing more stuff as you test.

[Jazz P Bass]

Mt money is on Q3 & Q4.
Check for shorts.
And run it up without an output load until the amp is stable.

[SYSDOA]

RG -

Thank you for the reply. On side note, ur site has been something of a 'new testament' for me along with this one. Much appreciated. Can I clarify? I do use a variac which I was under the impression was a step up from a bulb limiter. I do have one that I made many moons ago, but I have been using the variac. I'll follow your instructions and report back.

[SYSDOA]

Jazz -

I've been working with no load on this one. I learned the hard way awhile back and read here that I didn't need a load to work on solid state (well, most of the time). I'll check for shorts at Q3 and Q4 and report back. What draws you to this conclusion? I'm interested in understanding what I did (I've done it before and given up).

[Enzo]

The variac is a powerful tool, but if you turned it up to the point the bridge melted and the resistors started smoking, I have to think you were not watching the mains CURRENT the amp was drawing at that 60 volts. The point of the light bulb limiter is that IF the amp has something shorted so it draws a lot of current, the bulb will light up to prevent the amp having to take the excess current. The variac has no such limiting action if you just turn it up. Your eyes on the current meter take the place of the bulb.

And let us not blame AMpeg for a "stinker" situation, we can't blame anyone else when we blow something up. As to why this happened, you said it yourself, you shorted the circuit.

If the bridge looks melted, it likely is bad, but a simple check with an ohm meter will tell you if it is shorted.

If you shorted together the two resistors you metioned, then look at the schematic. You shorted the -40v rail to the output. Since the speaker was connected there (right?) its low impedance to ground means the -40v rail was grounded out. That does no favors for the bridge that supplies the -40v. Even without a load, like most all SS amps, it is heavily fed back, and when it sees -40v on its output, it will try REAL hard to "correct" that. And that means turning on the transistors to +40 to compensate. That will result in the plus side outputs connected from +40 to -40, with darn little to limit current.

That short stresses the plus side output transistors, so check Q5,Q6 for shorted.

Q3,4 are certainly worth checking, though looking at it, they each wind up with a 0.33 ohm resistor more or less in parallel, protecting them. Why suspect them? Your errant -40v was connected right to them via a couple resistors.

Make sure none of those 0.33 ohmers has opened or gone way off. And while you are at it, just to the right of them schematically is a 3.3 ohm 1w "OPT6". Make sure that has not opened. That is your stability network.

[SYSDOA]

Enzo/J Bass -

I went ahead and shotgunned Q3-6 as well as all of the cements. It's still wanting to chew up the bridge. The "OPT" thing, I'm confused about. Non of those symbols are on the board itself. Neither are the components. Does OPT stand for "option"? Thank you again for the direction.

RG -

I tested the filter caps. One read ok with my ohm meter and cap meter. The other read ok with the cap meter but not the ohm meter. Replaced the weird one but still couldn't get resolution. I'll keep testing individual components until I hear more. Is that normal though? I'd like to think I knew how to test a capacitor!

Thanks again

[R.G.]

I went ahead and shotgunned Q3-6 as well as all of the cements. It's still wanting to chew up the bridge.

Shotgunning is in general not too productive with shorts that are melting things. What works well for me is to open up the current path from the power supply, testing parts and "building" a working power supply behind me as I go. The melted bridge is probably toast; the transformer may be (might have been) toasted. I would go back like I said and start with the secondaries to the power transformer open. Got good voltage, not shorted? Good. Run the (current limited with a bulb limiter or watching the current meter on the variac) voltage up till the Xfmr secondaries measure good. Then reconnect them to a (good) bridge, and disconnect the bridge from the filter caps. Run up the (current limited) primary voltage again and see if you get DC on the + and - terminals of the bridge. When that works, connect the bridge to the first filter caps and disconnect the rest of the unit from the filter caps. Bring up the (current limited) primary voltage again and see if the filter caps will charge to correct voltages by themselves. Notice that if the filter caps can't get to correct operating voltages, it doesn't matter whether something past them is bad or not. Once you're getting good power voltages on the filter caps, reconnect the first thing after them and again bring up a current limited primary voltage. At some point you'll have good power before you connect something and bad power after you connect it. What you last connected is pulling the power down. Do NOT stop after you find the first one until it's all checked out. Failures in solid state stuff can let enough current flow to cause a chain of destruction. There is often not just one flaw.

The "OPT" thing, I'm confused about. Non of those symbols are on the board itself. Neither are the components. Does OPT stand for "option"? Thank you again for the direction.

Not sure what they meant. Ignore the "OPT" and go find the parts. Probably means they're out with the output jack or output transistors.

I tested the filter caps. One read ok with my ohm meter and cap meter. The other read ok with the cap meter but not the ohm meter. Replaced the weird one but still couldn't get resolution.

I apologize if this sounds flippant, but you didn't test the caps in a meaningful way. Whether a cap meter reads OK or not means nothing if it will not support a voltage. With a possibly shorted load on a filter cap, you cannot tell whether the cap or the load is shorted. Well, yes, you can, but not with a cap meter and DMM. Open the load on the cap, see if the cap can be (gently) ramped up to the design voltage without a load on it. If not, the cap is bad whatever the cap meter and DMM says. If yes, then the load is shorted and the cap is at least able to withstand voltage and not melt your rectifier bridge.

I'll keep testing individual components until I hear more. Is that normal though? I'd like to think I knew how to test a capacitor!

It's a more elementary test than your cap meter. Capacitance does not matter if the insulation punctures at 10V (for instance) and the cap meter tests at half a volt. Some caps have the sneaky trick of being partially self-healing to some degree. They still won't necessarily work til you find that they will in fact stand the system voltage by themselves.

Caps **might** be fine, might not. Can't tell til you separate the load from them and try.

[Enzo]

Well, now...

There are two versions of the B100R - one designed in 1997 or so, and the later one from 2005 or so. When was yours made? Look on the controls and other components for date codes. The newer one does not have the OPT6 part, for example.

[SYSDOA]

Enzo -

This unit is REV1 made in 1997.

[SYSDOA]

"Run up the (current limited) primary voltage again and see if you get DC on the + and - terminals of the bridge. When that works, connect the bridge to the first filter caps and disconnect the rest of the unit from the filter caps. Bring up the (current limited) primary voltage again and see if the filter caps will charge to correct voltages by themselves. Notice that if the filter caps can't get to correct operating voltages, it doesn't matter whether something past them is bad or not. Once you're getting good power voltages on the filter caps, reconnect the first thing after them and again bring up a current limited primary voltage."

Secondaries tested good. I'm trying now to figure out the best/safest way to test the caps disconnected from the rest of the amp. Actually, any suggestions on doing so would be awesome (typical SLM board. Everything is packed together and power leads such that the board has to be reinstalled to power back up). Thank you and thank you to Enzo for explanation regarding using a bulb limiter as well. I wasn't measuring current correctly.

"It's a more elementary test than your cap meter. Capacitance does not matter if the insulation punctures at 10V (for instance) and the cap meter tests at half a volt. Some caps have the sneaky trick of being partially self-healing to some degree. They still won't necessarily work til you find that they will in fact stand the system voltage by themselves."

I always wondered about this. At least I only spent $15 on the meter, but here it sounds pretty useless.

I'll do as you have outlined this afternoon and report back.

[R.G.]

I'm trying now to figure out the best/safest way to test the caps disconnected from the rest of the amp. Actually, any suggestions on doing so would be awesome (typical SLM board. Everything is packed together and power leads such that the board has to be reinstalled to power back up).

Ah, yes. The more tightly integrated, the harder to repair.

The problem with shorts is that they are not amenable to testing with voltage-based instruments of the normal kind everyone has. To find a short you have to find where the current's going. There are esoteric ways to measure this, like a meter which can measure the milli-volts or microvolts of DC voltage drop along the length of a conductor to tell directly which way current is flowing, or an oscillator/pickup to sense where AC current is flowing in which conductor. I did this once way back when. I used a lab sine wave oscillator in series with a resistor (to keep from burning out the oscillator) and a cassette tape player head to follow the AC current path. Signal is strongest right over the conductor where the AC current is going, and you can tell where and how it divides by the signal strength in the tape head. A modern variant of that would be with a linear hall-effect probe. You could measure the magnetic field that even a DC current gave you as you traced the probe along the conductor. Any branch that shows no magnetic field is not allowing current through.

But back in the real world repair shop...

The fastest way to figure out where current's going is to open conductor lines and see if it stops. Cutting PCB traces is handy, but can ruin the PCB. I always look for what can I pull out to stop current flow and isolate what's sucking the current.

In your case there are a few highly likely suspects and a host of other shady characters. The most likely suspects are, in order:

Q5-Q8; C33-C38; C2/C3; D2, the rectifier bridge; and the transformer. You've kind of eliminated the transformer. The rectifier bridge has been replaced, I think. That leaves you with capacitors and output transistors as the next likely suspects.

At this point, You have to guess what to remove or open next. Given that highly integrated setup, the logical choice is either Q5-Q8 or the capacitors; for ease of work from your description, I'd probably remove Q5-Q8, and leave the caps in place. Then bring up the power slowly and current-limited-ly and see if the capacitors can come up to full voltage. If not, replace them in +/- pairs, electrolytics first. When the caps *can* come up to full working voltage, then start worrying about the power amp. If the caps were the problem, it may be fully fixed. A shorted cap would have kept disastrous currents out of the power amp.

If the caps do come up fully, then the overcurrent was almost certainly going through the output transistors. The question is - why? Could be bad transistors, could be they were being TOLD to pull current. If none of them were clearly shorted, suspect the circuitry in front of them too.

IMHO, this is a fundamentally incompetent power amplifier design. It uses opamps which can't swing the full power supply voltages in a floating configuration to drive the outputs at full power supply. This can be made to work, but it's pushing closer to the edge of unreliability. IC2 is specified for absolute max power supplies of +/-22V, and it's being run at +/-20V. A sneeze on the opamp power line kills it. The power supply also floats on the output of the power amp. R47 allows the output voltage to drag the power supplies for IC2 up and down following the output. It works. Mostly.

Also, if Q9 or R40 were open, the outputs would do their best to eat the entire power supply. Or they'd do it until they shorted, then eat the whole power supply.

I have (marginally at least) more money than time these days. I would pull out all the semiconductors in the power amp and replace them with new, known-good ones, and test every resistor and cap in place. All of the money is in the output transistors, and replacing all the cheaper stuff will save enough bench time to pay for itself. IMHO, the smaller stuff is not to be trusted after a rectifier-melting event.

I always wondered about this. At least I only spent $15 on the meter, but here it sounds pretty useless.

Not useless - it just measures something that's not what's hurting you now. For measuring capacitor values, it's just the ticked. I keep a capacitance meter tucked away for when I have to know the capacitance.

[SYSDOA]

RG -

As always, excellent direction. I'll focus on the areas mentioned and report back. To this point, the power section behaves past the filters. Thank you for the guidance and thank you for your patience.

[R.G.]

As always, excellent direction. I'll focus on the areas mentioned and report back. To this point, the power section behaves past the filters.

Hmmm. Interesting. Do I translate that correctly as "The rectifier gets hot enough to melt, but the power output amplifier still amplifies audio OK and doesn't itself get too hot."?

[SYSDOA]

Pulling the output transistors helped. The amp didn't want to blow a fuse with or without the bulb limiter. There's higher than normal voltage at IC2 (-22VDC @ P4?) and -6V on J6. The voltages are dropping as I'm testing... why is that?