Crate (Loud Technologies Inc) V100H Power Supply (SMPS) Fault - Tearing out hair!

[lalunatech]

I have a Crate V100H with a very unreliable switch mode power supply on my bench and have reached a dead end with it. The symptoms are as follows:

B+ 420 Volts supply drops to nothing when the power supply is given a little tap with a screwdriver handle then comes back after the warm up delay.

This is quite a complicated supply with two HF power transformers each with it's own PWM controller circuit. One generates +/- 15V, -63V and the heater supply the other generates the anode voltage B+. After much head scratching I have discovered that the vibration makes the low voltage PWM controller shut down momentarily which also stops the B+ controller too. The B+ controller has a delayed start (555 timer chip) to allow time for the valves (tubes for you new worlders) to heat up. Sometimes however, the B+ does not come on at all but the low voltage supplies do so I haven't established the cause and effect of it.

I have desoldered and resoldered all the mosfets, bridge, the two HF transformers, heatsink mounted rectifier for the heater voltage and the two main reservoir caps. I checked over the transformers for bad winding connections but they looked OK. I have checked over the smd's and all other solder joints. The ICs have been resoldered along with any other suspect looking joints. It still goes off when you hit it!

It must be a mechanical fault of some kind and not a completely failed component because it works sometimes.

Has anyone found a similar fault on this amp because I am spending way too long on this and need to lay it to rest one way or the other. Any help would be much appreciated!

[soundguruman]

It makes a fine boat anchor, or door stop.

[Enzo]

PLEASE BE SAFE, SMPS are extremely dangerous. Look at all the primary side circuits, The ground symbol has a wavy line in it. That is NOT GROUND, it is the circuit commkon, but is rferenced to the negative side of the rectified mains voltage. DO NOT connect a scope ground or anything else to it. An isolationtransformer is recommended for servicing these.

Get out a wooden chopstick and push on each and every part in the HV supply.

It really doesn't matter if someone faced this before, because it is extremely unlikely their bad connection was in the exact same place.

The U7/T4 supplies have to come up first before the B+ supply is allowed to come on. After all, the -63v bias supply is there. Look at U7, see the three 150k in series above it? Those provide a kick start voltage from the main rectifier. Once U7 starts up, it powers itself through T4B and D20. But T4B also has D21 making the VCC supply. Watch that VCC, if it goes away, so will your B+. That is the first step.

VCC goes over to power up U1, which ultimately causes drive to the T1 and so your B+. U1 also puts out VREF. VREF powers U4, your 555, plus it is also the reference for the U3 comparators. U1 has a shutdown pin, pin 10, which must be held high. Low on pin 10 shuts it down. Left alone it goes low, it must be held high. Watch that pin, if it is dropping low, then find out why.

There is a +15 sensor and the 555 pulling up on the shutdown pin. If U1 is being shut down, look at both of those sources for the cause.

[soundguruman]

"PLEASE BE SAFE, SMPS are extremely dangerous."

The first thing to check is:
"Do I have a fire extinguisher?"

[drewl]

Nice help there "guru"...

I have to repair a variety of SMPS's for work, like Enzo said, be careful WATCH YOUR GROUNDING.

These can be a pain to work on, sounds like a flaky cap around the PWM, or opto coupler.

[lalunatech]

I'm very cautious with SMPS's don't worry, I come across them more and more nowadays unfortunately. I power it through an isolation transformer and am careful to discharge reservoir caps after every test. I have tested the opto couplers out of circuit, they are OK. I don't have much evidence but my hunch it is a transformer winding that is at fault. Like I say, it goes off when you tap it. Somehow the vibration upsets the little bugger. I have looked everywhere for a bad connection but haven't found it yet. I have checked all the parts that get hot and the big parts that vibrate and shear off their legs and have looked for bad soldering and nothing shows up. I am convincing myself now that if I crack this one I'll have a new tool in my fault finding armoury and be able to share my super fault finding skills with the world. If I charged by the hour, the customer would have a fit I've spent that long on it.

Dan.

[lalunatech]

I'm being ultra careful and I do have a fire extinguisher handy!

[soundguruman]

I'm being ultra careful and I do have a fire extinguisher handy!

See? Safety first Gentlemen.
"tearing out hair"
Well I have already given you the best advice possible, don't spend your time or money fixing it.
One of the most fun features of Crate is the production changes...
Which were frequently made at the factory. (changes made cause the amps were failing more than cats crap)
And when these changes to the design were made, these were never added to the schematic.
And so you may not even have the correct schematic, even though the label on the schematic "says" you do.
(the joke I think is on you this time, surprise.)

The techs at the Crate factory informed me: "Crate amps are made for occasional use only. These amps are not intended for continuous operation."
They told me this about 90 days before they went out of business.

Conclusion: even if you do get it working, it's going to blow, captain. It's not your fault. But when it does blow again, the customer will blame YOU.
A combination of (a) very poor design (b) really cheap parts (c) unbelievably bad thermal management [heat rises, somebody tell crate] (d) inaccurate schematics, see above. (e) no service notes, to upgrade existing known problems, (f) no parts support, or very poor parts support.

The Test: If the amp cannot run for 8 hours into a load at half sine wave power (without self immolation) then it's a paperweight.
Ask yourself before you repair any amplifier: does it pass this test when working correctly? If the answer is NO, don't waste your time on it.
AND don't lose any more hair either. It's NOT your fault!

[R.G.]

"PLEASE BE SAFE, SMPS are extremely dangerous."
The first thing to check is:
"Do I have a fire extinguisher?"

Nope. The first thing to check is "Do I have a buddy standing by to turn off the power and pull my unconscious body out of the burning building just in case CPR can revive me?". Fire extinguishers are number two, along with a second buddy to operate them.

I'm only mildly kidding. When I was designing switching power supplies, we all had to pass a CPR course before we were cleared to work in the lab.

Like I say, it goes off when you tap it. Somehow the vibration upsets the little bugger. I have looked everywhere for a bad connection but haven't found it yet. I have checked all the parts that get hot and the big parts that vibrate and shear off their legs and have looked for bad soldering and nothing shows up. I am convincing myself now that if I crack this one I'll have a new tool in my fault finding armoury and be able to share my super fault finding skills with the world.

If it was manufactured after RoHS was a requirement and uses lead-free solder, it just might have an internally tented solder joint. It took me three days to find this one. One of our switching power supplies was thermally intermittent. Worked fine cold, went into protection shutdown when it got warm.

There was a solder joint where the solder formed a tent-shaped covering OVER the clipped end of a resistor lead. When cold, the solder contracted and touched the end of the lead, and all was well. When it heated, the solder tent expanded and moved several microns (!!) off the end of the lead and the thing quit. I only found it because there was a microscopic crack part way around the periphery of the "tent" on the copper pad that I spotted under the stereo microscope.

If I were in your shoes, I would remelt 100% of the solder joints with some new, fresh, rosin-core 60-40 tin/lead solder. Assuming that's legal for you to possess any more.

The other issue with lead free solders is tin whiskers. Tin crystals too thin to see grow off lead free solder joints. If they touch...

Again, pollute the lead free solder with some lead bearing. Lead inhibits tin whisker growth.

These might not be a fix, but if you're down to shotgunning, it's quicker than solving it from first principles.

[Enzo]

Crate (and Ampeg) was a brand manufactured by Saint Louis Music, or SLM, which is still in business. Crate as a brand was sold to Loud technologies. Crate never went out of business, it was simply sold. SLM still exists, Crate still exists. What DID go out of business was the Crate FACILITY in St. Louis. SLM got out of the amp business to concentrate on their instrument business. I can imagine disgruntled employees there, who knew their jobs were about to disappear, might not have anything positive to say, but the Crate line never went under. It is still alive and selling plenty of gear.

[MarkusBass]

I think that Enzo explained how the SMPS works and what are the most important voltages to be checked when the high voltage controller does not start (VCC, VREF, OUT(NE555), SHDN(SG3525)). Can you verify them in case the low voltage controller starts and the high voltage controller doesn't? With the low voltage controller I would verify the supply voltage (VDD, VIN) - sometimes it may switch off due to undervoltage protection (check pdf what is allowed range) and if you think that the transformer is suspected, I would desolder it and check outside the circiut (with function generator). Of course, SLM will not send you such a transformer as a spare part but sometimes it can be fixed. I'm also a guy who would like to know why the SMPS does not start (even if I cannot fix it) - don't give up.
Ther is also a possibility that there is a short on the power amp board and the controllers do not turn on due to the overcurrent protection. Maybe this is not very likely but I would test the SMPS with dummy resistors - outside the amp.

Mark

[soundguruman]

The deal is you don't "LOOK" for a bad connection. The odds are you can't see it if it's there.
Just re-solder the entire circuit board. ALL the connections.
If that won't do it, Reynolds pays top dollar for scrap metal, and it's all cash.

[lalunatech]

Thanks to everyone for the help with this. I thought I'd put an update on here because I did manage to get it working in the end.

As soundguruman suggested, I fluxed and melted every solder joint on the board. Doesn't have the satisfaction of diagnosing the fault and tracking it down but it worked. The fault was making the 555 timer in the B+ 420 Volt supply re-trigger so the B+ never came on.

That's not the end of the story though. I left it running into a dummy load at about half power for a while then the B+ supply power mosfets Q1 and Q2 went bang along with their driver circuits. I replaced all these and powered up and the B+ worked but was at about 550 Volts! The B+ power supply does not have any voltage output error regulation. U1 pin 1 is grounded and pin 2 is at vref so it is running at full tilt all the time. I discovered that the capacitor C23 that sets the SMPS frequency had drifted down to half the correct value. Maybe I overheated it when resoldering. I replaced this and B+ dropped to 440V. Seems like bad design to have B+ at the mercy of the line voltage and passive component values.

I never want to see another one of these heaps of shite again!

Some way in to this job I emailed Loud Technologies UK but they said they'd checked UK and Europe stock for this power supply and there weren't any in stock at all. They did offer to send me the schematics though. Thanks!

Dan.

[soundguruman]

Thanks to everyone for the help with this. I thought I'd put an update on here because I did manage to get it working in the end.

As soundguruman suggested, I fluxed and melted every solder joint on the board. Doesn't have the satisfaction of diagnosing the fault and tracking it down but it worked. The fault was making the 555 timer in the B+ 420 Volt supply re-trigger so the B+ never came on.

That's not the end of the story though. I left it running into a dummy load at about half power for a while then the B+ supply power mosfets Q1 and Q2 went bang along with their driver circuits. I replaced all these and powered up and the B+ worked but was at about 550 Volts! The B+ power supply does not have any voltage output error regulation. U1 pin 1 is grounded and pin 2 is at vref so it is running at full tilt all the time. I discovered that the capacitor C23 that sets the SMPS frequency had drifted down to half the correct value. Maybe I overheated it when resoldering. I replaced this and B+ dropped to 440V. Seems like bad design to have B+ at the mercy of the line voltage and passive component values.

I never want to see another one of these heaps of shite again!

Some way in to this job I emailed Loud Technologies UK but they said they'd checked UK and Europe stock for this power supply and there weren't any in stock at all. They did offer to send me the schematics though. Thanks!

Dan.

That's funny-$6300 labor on a $200 amp...but You did not let it beat you! You WON.
Now you see why I don't want to work on them.
You will think twice about repairing another one, am I right?
NO! bring them all here. I will fix them all.

[lalunatech]

Oh no. customer called back, power supply has gone pop again, refund time. I hate that!

Blurb on Crate site is funny
Crate - The Best Value In Professional Tone :: Products
Especially the bit about how the the power section interacts with the power amplifier and the speaker to help characterize the amp's sound (ie. silence) and how it is called a switching power supply because it switches voltages when you cross borders!