AC... to bipolar?... to.. uh...? Hmm... little help?

[NaBo]

It's been a while and I still need help understanding the power supply here: Crate GFX-15

What happened is the transformer, with an internal fuse, got toasted.

I dont have a 15V transformer to replace it with, so would I be able to just pull the 4 rectification diodes, build a 15V bipolar supply (based around a 24V 400mA transformer, 7815 and 7915 regs), and attach it to the 15V+ and 15V- points?

I really wanna get this up and running so I can hook up my horn driver when it arrives and enter talkbox country.  PLEASE help me understand this power supply

[R.G.]

PLEASE help me understand this power supply

Did you read anything at GEO (http://www.geofex.com)?

Try the "Power Supplies Basics" articles, then post back.

[NaBo]

Mmhmm, I've plowed through that article a couple times, and I think I've grasped that much.

I unsderstand what's happening up until the resistors... It looks like a standard bp supply til then (which I'm planning on replacing for the repair).  I guess the next two test points are what's really confusing me... what does the 15VA signify?  Is it converting the 15V+/- DC back to an AC signal or something?  The waveforms given in the top right look like it...  Come to think of it, yeah, thats what the resistor-capacitor configuration would do, no?  But isn't that what it started with in the first place?  Hmm...  :?

As for restoring this to working order... I would just pull everything before the resistors, build a regulated bipolar supply (this time with a replaceable fuse...), and then hook that up before the resistors, right?  That the best way short of having a direct replacement for the transformer?

Thanks for reading, R.G., I'm sure I'll have this down with your next reply.  

[niftydog]

would I be able to just pull the 4 rectification diodes, build a 15V bipolar supply (based around a 24V 400mA transformer, 7815 and 7915 regs), and attach it to the 15V+ and 15V- points?

you could potentially do that, but you could also just replace the transformer. It wouldn't have to be the exact part, just get a similarly sized 15V center tapped transformer.

BUT, you have to ask yourself why the transformer went kaput in the the first place! You could end up just toasting the transformer or power supply that you retro fit without actually fixing the problem.

Looks like they're just trying to separate the op amp supply from the power amp supply. The 15VA goes to all the op amps, but the plain 15V goes to the power amp. Maybe those resistors are some current limiting or something.

Ok, the schematic is a little hard to read, but I think you're not taking into account the amplitude of those waveforms. From what I can make out, test points 9,10,11 and 12 are power supply, yes? Those waveforms are given at a DC offset (­±14V or something, can't read it) and show a very small AC ripple component (looks like 16mV or something). That's entirely normal and I wouldn't call that "converting it back to AC".

I would first try to replace the transformer, then work out the "real" cause of the fault.

Failing that, making a ±15V unregulated supply and feeding it in at those resistors would probably work. However, keep in mind that the supply needs to source the appropriate amount of current, and it's higly likely that a simple regulated supply (LM317 based) would fail given this task unless it's really well heatsinked... and even then there's some doubt.

[R.G.]

Nifty's correct. It looks like they're just separating the opamp supply from the power amp supply.

R41 and R42 separate whatever's happening on the +15 and -15 from +15Va and -15Va. The capacitors on the ...Va lines hold them firm in the face of wiggling around on the +/-15. The "a" sections are a bit cleaner than the entry.

The power used prevents you from taking any old 11Vac wall wart and powering this. You'll need to find a 20Vct transformer to run it.

If you found a 24Vct transformer, that would give you +/- 17V nominal,  but I'm afraid that any high line voltage or low power usage would make it go over +/-18V and that will likely kill all the opamps.

Notice that you can't use a 30Vct transformer. That's 15V rms on each side of the center tap and when you full wave rectify that you get 15*1.414 = +21.2 and -21.2V, which will most likely kill your opamps.

So you need 20 or 22Vct, no more. 1A should be plenty of current.

You *could* feed it +/-15Vdc and the full wave rectifier will put the + and - where it needs to go, but again, Nifty's right - before you put much of anything in there, figure out why it popped in the first place.

- Shorted output?
- shorted power jack?
- bad power filter cap? (C28/C30)
- bad diode? (DL1... DL4)

[NaBo]

Aw nuts... just when I think I'm getting something, it turns out to be more complicated than I thought!  

I've found the cause of the short... actually I realized it while I was doing it... idiocy on my part.  Last year I was attempting to test a feedback loop but I didnt have any pedals home with me, so I decided to try using my practice amp which I didn't bring to school with me.  I kept the volume low, and it worked fine when the loop was out of the box... but when I boxed it up, it stopped working...  "hmm, I'll just crank the knobs a bit. *pop* ... *the smell of burning* ... oh, I guess I'm effectively plugging a patch cord from the output to the input.  amps don't like stuff like that."  I just couldn't wait a few days to try it with my pedals, or check for continuity between the tip of the plugs and the jack (which was the real problem... output jack was bent a bit).  Stupid stupid stupid.  At least those days are behind me... taught me a valuable lesson... If you don't know for sure, ASK.  :shock:

Anyhow, if the current handling capabilities should be that high, it'd probably be best just to find a replacement transformer afterall.  Looks like the hammond 166L20 is my best bet, since it's in stock at digikey.ca and is rated for 2A.  I'll fuse it at 1A so i never have to deal with this crap again

In the meantime, I think I'll take this opportunity to replace all the electrolytics since this thing is pushing 10 years now.

Thanks for all the help guys!!!

[toneman]

Nabo,
i'll chime in here, quickly..
everything everyone else said is absoutlely valid.
Even, your idea of building your own bipolar supply is valid.
*IF* U know absolutely what U R doing.
Since U *did* find the fault, the amp should B" easy" to rebuild.
Here;s my suggestion:
why not contact Crate??  order an original xfmr??
chauk up the expense to experience.
U know what "they" say---
"Ass, gas or grass, no one rides 4 free"   LOL
or, Knowledge is power.  Power is expensive.
Therefore, Knowledge costs U....
stayKnowledgable
tone

[NaBo]

Even, your idea of building your own bipolar supply is valid.
*IF* U know absolutely what U R doing.

Yeah, the main reason I was gonna go that route is cuz I have five 24vac transformers that I picked up for 2 bucks apiece...  But an unregulated supply with a centre-tapped transformer could dish out more than the .4A of my transformers or the less-than-1A limit of regulators, which it seems this thing could require...

Here;s my suggestion:
why not contact Crate?? order an original xfmr??

Well, that would certainly get it working... but again, the original had an internal fuse that couldn't just be replaced... I guess it did its job of keeping things, you know, not on fire............. but otherwise pretty stupid!  That's the main reason behind my problem... if it were possible, I would've only had to replace a 15 cent fuse, instead of replacing a 20$ transformer.  Thanks for the ingenious design, Crate

But yeah... I learned my lesson.  Test things thoroughly before hooking them up to quasi-expensive things. :x

And amps aren't just loud stompboxes...   :wink: