KC350 - Blown Amp? Preamp Seams OK

[Jung4g]

I picked up a "broken" KC350 off Craigslist yesterday as a project to maybe help a buddy who needs (wants) a keyboard amp and would trade for a working Marshall AS50D.... Note that it would be for very occassional (likely never) use on a old family Rhodes or similar vintage keyboard by a music fan that can't play an instrument (but for some reason collects cool old gear). It'll never leave his basement, see a gig, or go more than a crack above minimum... IF I get it working again.

The previous owner had bought it used 8 years ago, and she said after plugging a source into the Aux In (JK14), it stopped working and now just Hums all the time, plugging it in with no source and all volumes at minimum confirmed a hum. Changing volume levels of various inputs did change the hum level a bit.
I had printed the schematics (got on this forum and reattached below) at work before bringing it home and set into opening it up to see what I was dealing with.
Roland_kc-350 schematic.pdf

- Step one was pulling the pre-amp section to do a visual for any obviously popped caps, burnt anything, melted wires, etc. All good there.
- Step two was pulling the woofer to get at the power supply and amp boards. First thing I noted was that the woofer cone was very stiff (and suspected blown). I'm used to working on long throw Hi-Fi woofers that move a bit easier compared to Pro Audio counterparts, but it wasn't totally frozen, so I moved on to the internals. Again, nothing obviously wrong. All fuses were intact and nothing was obviously blown.
- Step three was just connected the woofer again to see what happened and the first thing that surprised me was a spark when connecting the wire to the terminal, along with an audible pop and a slight movement of the woofer cone.
- Step four I pulled another woofer over and tried with that (just in case the woofer coil was shorted) and when just tapping the terminal of the alternate woofer it pushed out hard and drew the same spark. We're talking a 1 way motion of the cone, just extending as far out as possible, a light hum, and then it returned to resting state when disconnected. I did these in very short bursts, as it was a known good woofer. Note that I had not yet disconnected the tweeter at this point.
- Step four was measuring a no-load voltage at the speaker outputs after disconnecting the preamp feed (CN9) and retrying the above test with no change.
Woofer: 10-15mV DC (0V AC) Tweeter: Started @ ~3V DC (0V AC) and then as I left the meter on it it dropped over 30 seconds or so to ~350mV and settled there. Note that I wasn't getting any sound from the tweeter, I suspect that is blown as well. From those results and the sparking, I assumed the Amp board is the culprit.
- Step five was testing the preamp at least at the headphone output. Doing a VERY basic test as I don't have many working sources around the shop at the moment. Using an Aux In and the Headphone out, I was getting clear audio. Of course, those two functions don't test much of the Pre Amp board (main out, main volume, etc), but it is getting power and that basic function worked fine. By cranking any of the channels individually, you could get some some hiss out as well. Not in a bad way, just the normal noise from going to 11 without a source.
- Step six (as far as I got for night one) was testing the voltage output of the Power Supply feeding the amp board. As expected, I got + 55.1VDC and - 55.2VDC at the WP25 connector. I didn't test it on the Amp board (WP26) as that's 2" away and nothing was wrong with the connector. I didn't test the 24v or 15v circuits on the power supply since the Pre-amp seemed to be working fine.

Ok, now a few basic questions:
Is there anything that is obviously from the above that says "The amp board is blown, don't bother moving on with it" OR "It sounds like the Q17 blew, likely from hooking up a hot Aux In Source while it was on and cranked"?
From the schematic, it looks like there is a first order high pass filter with pad (C18 and R31) on the tweeter output, but no low pass on the woofer, is that correct?
If Fuse 4 or Fuse 5 on the Amp Board are blown, would that negate any output at all and cause the sparked output issue I was seeing on the woofer terminal?
Any way to isolate if the issue is in the gain stage of the amp or elsewhere?
Am I way off base on anything so far?
Am I being too hopeful that the preamp section is really ok?


If I can't fix the amp board quickly, I'm thinking I scrap that part of it all together. What I'm considering as an alternative is a basic IRS2092 200w mono circuit (Class D). Something like this for $16 shipped: YJ IRS2092 200W Class D Amp Mono Amplifier Board | eBay
Yes, that may not be up to the rigors of club life, but for the cost, I figure it's worth a try AND considering the amps likely home, I think it'll be fine.
That amp board wants up to +/- 60VDC power (+/- 50VDC optimal), so I figure the +/- 55VDC (will be closer to +/- 50VDC under load and should work great). There are higher power, lower ohm capable circuits, but they cost more and I'm doubting the 100 watts (@ 8ohms) would get used in a basement setting anyway. The extra 6db output from a 350 watt (@ 8ohm) board just doesn't seem necessary (at least for this test).

If I go that route, I'll create a passive crossover to go with whatever new speakers I use. I'm considering a Peavey Pro 12 and a basic bullet tweeter like the Parts Express 270-055.
All told, I should be ~ $120 into the whole thing: $30 off craigslist, $16 amp board, $70 for the replacement speakers and few bucks for crossover parts if I don't like what I find in my parts bin.

Thanks for any insight. There seems to be some great minds on this board from what I've read so far.

[nickb]

First, some recriminations...

The first rule of fixing a solid state amp is don't connect a load (speaker) until you know there is no DC on the output. If you don't follow this rule you stand a good chance of blowing up an expensive speaker. It sounds like the original speaker is toast and if you had the test one on for more than a second or so you may have damaged that too.

It would have been a big help if you have mentioned the manufacturer, Roland, in the title and polite to include the schematics so we don't have to go looking for something you already have.

NOTE:I've edited this next paragraph to avoid future confusion as I was looking as the wrong schematic....

Start with the power amp on page 7. It's a simple one. First measure the DCV across the output (no speakers!) to confirm what I think we already know. Assuming you find more than 0.1V or so we can proceed to testing the output & driver transistors Q7, Q17, Q9, Q16. I prefer to remove them to test in case one has failed open. Also check the and resistors R34, R46, R30, R49, R33 and R45.

Report back and we'll go from there.

[Jung4g]

Thanks for the reply, nickb. It's been years since I've been on a forum (or dug this deep into an amp), so relearning the courtesies expected is part of that game... I updated the original post with your suggestion.

As for measuring the DC output voltage, I did measure that, just after I noted the speaker issue. From above:
Step four was measuring a no-load voltage at the speaker outputs after disconnecting the preamp feed (CN9) and retrying the above test with no change.
Woofer: 10-15mV DC (0V AC) Tweeter: Started @ ~3V DC (0V AC) and then as I left the meter on it it dropped over 30 seconds or so to ~350mV and settled there.


...we can proceed to testing the output & driver transistors Q210, Q212, Q211, Q213. I prefer to remove them to test in case one has failed open. Also check the emitter resistors R232 & R233.
We must be looking at a different Schematic. I don't see anything with those designations. Here's what I'm looking at.

[nickb]

No worries

Duh, sorry. I was looking at the KC500 It's page 30 of the manual I linked to above and is the same as the one you included - thanks.

So look at Q7, Q17, Q9, Q16 and check resistors R34, R46, R30, R49, R33 and R45. ALso fuses 4 and 5.

[Jung4g]

So look at Q7, Q17, Q9, Q16 and check resistors R34, R46, R30, R49, R33 and R45. ALso fuses 4 and 5.

OK, to be honest I've never tested transistors before. I looked it up, but I'm not certain I did it right I was just testing the readings between the center pin and the outer pins. Center is Base, right? So I'm testing the resistance between the base and the emitter or collector. I tried both ways on both (negative to base and positive to base) and from what I read briefly I was looking that they weren't shorted (0 Ohm) AND that they were at infinite ohms. From what I can tell, none of those 4 were dead, but the fact that they didn't just give a single reading (other than Q7) seemed odd to me. It's as if the DMM's load was changing something where the readings would go from 500 ohm and then up to 300k ohm over 10 seconds or something odd like that.

Also, I tested everything in the circuit, I didn't pull components via desoldering prior to testing. Does that matter?

Q7 - Not shorted, readings would fluctuate, but in the 250k ohm range
Q17 -Not shorted, readings would fluctuate too much to give a value
Q9 - Not shorted
Q16 - Not shorted
R34 .4ohm (Rated .33)
R46 .3ohm (Rated .33)
R30 3.5ohm (Rated 3.3)
R49 3.5ohm (Rated 3.3)
R33 216.2ohm (Rated 220)
R45 216.5ohm (Rated 220)
Fuses 4 & 5 are still intact

[The Dude]

It would be better if you tested the transistors using the diode check function. Here's more on checking diodes and transistors.

Electrical Measurements Part 3 | RSR Electronix Express

Edit: Also, the base is not always the center leg. One of the best ways to know for sure is to look up the datasheet for the transistor.

[Jung4g]

Ok, I'm not sure I did that right... again, still in circuit

Q7 and Q17 - Resistance between outer terminals 220ohm on both, I couldn't get a beep (that happens when I touch my leads) in any combination of B to E, B to C, C to E, reverse the polarity and the same thing happened on all of them.

Q9 and Q16 - Resistance between outer terminals 5.5M ohm on both. In Diode check mode, no beep in any combination.

If it matters, I'm using a Mastech MS8268 DMM. Just a cheap one off Amazon from years ago.

For the few bucks it would cost, is it worth just replacing the transistors?

Or since it wouldn't be hard or expensive, any thoughts on the IRS2092 amp board I posted originally as a replacement? For $16 shipped, if I'm replacing parts, it won't take long to exceed that price in parts plus shipping.

[The Dude]

The resistance readings don't help much. We need to know if the transistor junctions are as they should be. Don't worry about the beep. That just tells you if there is a short. We need to know the voltage drop when you check the junctions. It will show on your meter when you use the diode check function. It's all explained in the link I posted. Once you get the hang of it, it's really quite simple.

[nickb]

Ok, I'm not sure I did that right... again, still in circuit

Q7 and Q17 - Resistance between outer terminals 220ohm on both, I couldn't get a beep (that happens when I touch my leads) in any combination of B to E, B to C, C to E, reverse the polarity and the same thing happened on all of them.

Q9 and Q16 - Resistance between outer terminals 5.5M ohm on both. In Diode check mode, no beep in any combination.

If it matters, I'm using a Mastech MS8268 DMM. Just a cheap one off Amazon from years ago.

For the few bucks it would cost, is it worth just replacing the transistors?

Or since it wouldn't be hard or expensive, any thoughts on the IRS2092 amp board I posted originally as a replacement? For $16 shipped, if I'm replacing parts, it won't take long to exceed that price in parts plus shipping.

Let's figure out what needs doing before considering replacing the power amp. The one you linked will certainly work but I think it may need a fan for reliable operation - the KC350 may already have one. some effort will be required to mount and wire it in too.

Testing things in circuit can be confusing, it's best avoided. For example you can't test the base to emitter junction of Q17 or Q12 in circuit at all as there is a small resistance if parallel. Do what 'The Dude' said - put meter in diode test mode and measure both ways round on CE, BC and BE. You will need to research the datasheets for the transistors to find out the pinouts. I suggest you draw a table of the voltage readings with headings like this. It's very important to be consistent so B>E means red lead to Base and black to Emitter always and so on.

Reference	B>E	E>B	B>C	C>B	C>E	E>C
eg Q17	0.5V	OPEN	0,6V	OPEN	OPEN	OPEN

[Jung4g]

Good deal. I've got some work for the future. If not tonight, it'll be a couple weeks before I'm back in the shop to try this out.

I did pull data sheets for all of them and none of the transistors in question are center post as base, so that alone will make a difference. They're all BCE.

[gwell1]

Sorry if i ask questions that have been already clarified in the original post but I have A.D.D!
1) Are the other channels working?
2) The first thing i do is always verify you have all of your voltage rails working
3) If you suspect the original owner plugged in a hot signal into the Aux input, I would replace IC 6
4) It would be help full if you had a signal generator. Before I could afford mine, I downloaded one from the Android playstore on my phone, cut an 1/8" aux cable in
half and made a simple and cheap signal generator! Hey it works!
5) Check your main output transistors using a multimeter in diode mode

[Jung4g]

Ok, finally got some time in the shop again.

Here's the results from the requested transistors with common lead (black) first:



That looks good to me, but I could be off.

@gwell1:
1) Are the other channels working? I never got to test as what I had available before pulling it apart would work, but the speakers were blown anyway
2) The first thing i do is always verify you have all of your voltage rails working I believe the PS is ok, I was getting +/- 54-55V with no load and +/- 15V to the preamp
3) If you suspect the original owner plugged in a hot signal into the Aux input, I would replace IC 6 Ok, so grab a new pair of the M5218AL Op Amps
4) It would be help full if you had a signal generator. Before I could afford mine, I downloaded one from the Android playstore on my phone, cut an 1/8" aux cable in
half and made a simple and cheap signal generator! Hey it works! Easy enough to make. I've got a app that outputs any frequency I want for tuning Home Theater Crossovers, until the amp works, I don't know where I'd use that
5) Check your main output transistors using a multimeter in diode mode Done, results above

Next Steps?

[The Dude]

Those transistors look ok and good job measuring and posting the info. Seeing that, I read (quickly) through the thread and can't figure out what we are fixing. I may have missed it easily since it was a quick skim read. But, I see 1) no AC on the output, 2) acceptable DC levels on the output, 3) transistors good. So, and I'm not trying to be a smarty pants, what is wrong with the amp?

[Jung4g]

It just hums all of the times and the speakers are basically blown and the speaker outputs spark when connected.

Back to the potential of the IC6 Op amp being the issue, that seems odd, as the headphone output worked properly when I fead into the Aux input, which is where those op-amps are connected.

I'll try reassembling to test the power amp again and then connect the pre-amp output to a separate power amp input to see what's going on with an isolated test of the pre-amp.

[Jung4g]

OK, so after resoldering everything back in place, I tested it more, and now there is no buzz, no DC on the speaker outputs, no sparking when I hook up to the old "blown" woofer and I could even get some sound out. I hooked it up to another woofer (car subwoofer) that I had nearby and it'd make noise, though clearly not well tuned for the application. I think the amp itself is ok as of now.

The tweeter is blown AND the woofer is in rough shape. I'm going to go cut into the woofer to see what I see. if I can somehow fix it, great. Otherwise I'll put in a replacement speaker set and call it good.

Thanks for the help. I learned some stuff and I'm guessing the disassembly and reassembly had part of the fix built in somewhere... maybe just some cold joints?