Another Crate SMPS issue (V100H)

[bakeaj]

Long time lurker, first time poster. I'm the proud owner of one of those Chinese Crate amps with the SMPS that likes to die (100 watt head). I humbly bow before you to ask for suggestions...on where to proceed; I realize the best suggestion may be to bash it mercilessly Office Space style.

A little background:
I'm very new to tearing apart amplifiers, and electronics in general, but I'm a good student. If fact, I bought this piece of machinery to teach myself how to do this stuff, only paid forty bucks for it too. I had no idea what a switching power supply was when I bought it, and given the complexity thus far, I would have bypassed it had I known. The kid said he took it to a tech, and it needed fuses and new power tubes. Wrong. I tested the fuses as soon as I got it home, they were fine. I should have known then that I was in for the long haul, but I proceeded to track down schematics, test components on the main board, and generally struggle, although I learned quite a bit in the process. Essentially, the thing powered up and lit all the tubes, but wouldn't make a sound, not even a hiss. After a couple of months, I decided that I was beyond my skill level and took it to a local tech. $80 and a couple transistors and a flameproof resistor later, I had an amp that still didn't make a sound. The tech wasn't sure what was going on, and I decided I would take another stab at fixing this thing myself, having done some more research.

So, I've had the amp for a few more months since then, and I've fiddled with it, cursed at it and threatened it, and finally I discovered that the SMPS is not putting out the correct voltages. In fact, there is almost no B+ voltage, and the other voltages are out of wack too, though not by nearly as much. Why the tech didn't realize this, or bother to tell me, I'm not sure. The way I figure it, no B+ means no sound, and the other voltages are close enough to their specifications that nothing is exploding and all the tubes are lighting up. I surmise that one of the power tubes shorted, sending power where it shouldn't; there was a fried resistor on the main board, and two fried transistors on the power supply that the tech replaced.

After reading multiple sites and pdf's about troubleshooting SMPS's, I'm inclined to believe that one of the electrolytics is faulty, either in one of the outputs or the feedback, causing problems to the reference voltage or excessive ripple. I'm fairly limited in equipment (and skills, haha); I tested all the components on the board using my DMM with whatever tests I could find on the 'net, but I don't have an oscilloscope or an ESR meter to test capacitors. I did test that the capacitors would charge (measuring resistance to infinity), but I've read that it is really only good to tell you whether the capacitor is totally open or shorted, not whether it is functioning properly under powered conditions. Obviously, there are no bulging caps anywhere.

So, at this point, I'm looking for suggestions as to what steps to take next. I intend to purchase an oscilloscope, but that will be many months away, and I still have to learn to use one. Most folks recommend replacing all the electrolytics, given that they are cheap, is that a reasonable next step? And should I replace the big filtering caps on the primary side as well? I'm open to any advice you might be willing to give.

I realize I'm a noob, so apologies if I've committed a sin against tech law. Thanks in advance.

[J M Fahey]

Hi bakeaj.
¿Why do you write using such narrow text columns?
The post becomes *very* long and phrase meaning is lost.
It's difficult to follow what you write.

[Enzo]

Look at the SMPS. There are some small inline connectors for most of the SMPS output voltages, and a couple push on connectors for the B+ wires. I forget, are they violet and orange? NO matter. Pull the B+ wires off and unplug the low voltage output connectors. Now the SMPS can power up with no load. Check for voltage at those B+ push-ons as convenient. Does the SMPS wake up?

As with any other tube amp, the most likely failure is a power tube. And that can damage the SMPS. If the SMPS works alone, then find the problem in the amp - pull the power tubes and reconnect the SMPS - power up?

If the SMPS doesn't fire up unloaded, then pull it. BE EXTREMELY CAREFUL, THIS THING RECTIFIES THE MAINS DIRECTLY. The primary side of the SMPS is NOT refernced to ground or chassis. The secondary side supplies are referenced to chassis though.

Most SMPS failures I have seen in these has been shorted rectifier diodes in the B+ supply. I don't have the V100 drawing handy, but seems to me ther are three secondary windings each with their own diodes and all wired in series. Those are 2A ultrafast rectifiers, so no 1N4007s please.

And if your SMPS is not the same as the one I am thinking of, the same sorts of things apply. I have been fixing switchers for 30+ years, and one of the most common failures regardles of what the SMPS is driving is shorted secondary rectifiers.

[bakeaj]

Hey folks, thanks for the quick replies.

JH Faley:
Is my formatting weird? Looks alright from my side of the screen. But that does lead me to my next point.

Enzo:
Maybe I wasn't clear in my description of what is going on, or it was lost in my diatribe. The SMPS fires up, fan runs, all the lights are on. Those are the B+ wires you are talking about, and I have disconnected them, as well as the other push on's to measure voltage. B+ voltage surges to ~2V when you flip off the standby switch, then drops to ~.3V.

There are a total of 4 outputs, the B+ at 420V, then 12.6, 63 and 15. The -63V measures around -53V, and the other two are within a volt or two of their spec. So, would you consider this SMPS working? Obviously the voltages are wrong, but there aren't what I have read to be classic signs of failure (i.e., no function, immediate power off, ticking fan, etc.) Would shorted rectifiers lead to a shutdown of the entire unit?

So, if I go about testing ultrafast diodes, I can just check for continuity correct? Yes one way, no the other? As I said, I have tested them, but none off the board as they all seem functional.

[Enzo]

The diodes will test just like diodes. Don;t worry, they won;t have slowed down or something. They will be shorted or open or OK. Your meter will be fine for that.

The SMPS is running then, but you have no B+, so I still bet my lunch money one of those diodes is bad. They are the multiple suppllies wired in series, there is not just one 450v winding. SO any one diode opens, it kills the string. I find them shorted more often than open, myself. SHorted would usually shut it down, but an open probably wouldn't. And to be fair, it may not be a diode, a broken solder joint to one secondary pin on the little transformer could kill it as well.

The fact you have the other secondaries working, at least being close, means the primary side is probably OK.

I am not a fan of wasting an hour testing a 20 cent part to see if it is bad or not, I just replace it if I suspect it. On the other hand I am also not a fan of just wholesale replacing parts. In the case of a 20-30 year old amp, sure, the e-caps are probably all tired. But this thing is not old, and we could replace 25 caps and find out that there was an inductor with a broken off lead. By the way, check the indictors for broken off leads.

No scope? OK, check each secondary with a volt meter. Each rectifier charges its own cap. See if they all are being charged. Look at the solder real close. YOu have the schemtics, look at the B+ circuit. The fact it is a switcher churning out odd square waves at 100kHz doesn't alter the fact that the secondaries take that odd waveform, rectify and filter it just like any other power supply.

[Steve Conner]

Does the B+ come off the same converter and transformer as the other supplies? If I were designing one of these things, I'd consider a separate one, it makes it easier to implement the standby switch.

If it is separate, then a shorted secondary rectifier could kill the B+ only.

I mention this because I'm suspicious of the open diode theory. As Enzo says, diodes almost always fail short, they only go open if the circuit has enough energy to explode them after they've shorted.

Also consider how the SMPS came to burn out: maybe the protection diodes across the power tubes on the amp board failed first. This shorts the B+, which would overstress the SMPS.

My guess is that this happened, and it blew a secondary rectifier, which then caused the chopper transistors to fry. The fact that the panel lights and heaters stayed on is what makes me think the B+ is a separate converter.

If the worst comes to the worst, you could always junk the SMPS module and put an old-fashioned iron cored transformer in there. :-)

[Enzo]

I don't have the V100 drawing, so I am imagining that it looks like the other V series SMPSs. Here is the V50 watt version. Opens in the B+ section I would think more likely a broken lead or solder joint to the inductor or the transformer secondary. But the diodes are testable in seconds with a hand meter, so...

And yes, I have no trouble thinking a power tube failure shorted the B+ rail, which could have damaged the SMPS.

But ther is only one primary circuit in the SMPS, no dual conversions here.

[Enzo]

Dammit, upload manager not cooperating.

Here is V50 SMPS

[Steve Conner]

Thanks for the schematic Enzo. That's quite a cheap and nasty SMPS there :-/

There is only one converter, so all the problem can be is:

D103, D104 or D105 failed open (I'd expect one of C106, 107 or 108 to explode in this case as it gets hit with the remaining B+ in reverse)

One of the transformer secondaries burnt (again I'd expect an exploded cap)

Bad solder, cracked PCB etc

Unless of course the V100 has a different topology. It's not impossible that they could have kept the V50 circuit for heaters etc. and added a second converter for the B+.

[J M Fahey]

Hi bakej.
I'm not blming you on anything, far fro it, but with this new Forum software, some posts (including yours) come out formatted into narrow columns, others come out normal seemingly at random.
Obviously nothing you (or others) did, just buggy new software.
As an example, see how 3 posts are dendered in a different fashion, obviously all three at the same time, same OS, same Browser, etc. ¿¿¿¿¿?????

[Enzo]

And I just took in a V100H, so pretty soon I will get a chance to look at the SMPS and see if it is the same as the others.

[Steve Conner]

Re the formatting issue:

On my screen, bakeaj's post looks wider than JM's post complaining about it.

Anyway, when posting on a forum etc: Don't press return when the cursor gets to the end of the box. Let it wrap, and press return twice when you want to start a new paragraph. That way your post will adapt itself to the screen size of whoever is viewing it.

[J M Fahey]

Hi Steve, thanks, maybe that's what happened.
I have the distinct impresion, although I should look at more posts to confirm it, that posts made under the old system, come narrower under the new one, at least for me.
New ones look normal.
We'll see, and thanks again.

[Enzo]

You have to be pretty old school to hit return at the end of each line, even I don't do that. And I even still own a typewriter.

Steve, you were right, I was wrong. The V100 does have a dual converter SMPS, unlike the V50. In the schematic request section the V100 drawings were already there from January.

[Steve Conner]

Sorry Enzo!
*Kachunk... sound of carriage return*

I looked at the V100 drawings there, but couldn't see the SMPS amongst them, it only seems to be the amp part. Unless I missed it somehow.