Roland KC-350 Preamp Problem

[Enzo]

Prime suspects are the ICs, and the bypass caps. Each IC has a small cap from each rail to ground, plus the overall rails each have a bypass cap to ground.

Look at each IC closely, are any cracked or discolored? Look on the underside of the board under each IC, any signs of browning? Find the little bypass caps, and black spots on any?

Securely clip your ground probe to ground near the connector. Now probe the -15v line at various places, note the reading carefully. The traces will have a small but readable resistance, so if you see something like 115 ohms here, and 116 ohms there, then here is closer than there to the actual problem. Remember, one of the ICs is on the jack board, so unplug that board from the preamp board and see if it affects your 115 ohms.

If you are reduced to opening jumpers, start near the middle of th board, so if possible you can divide the board into two halves. Then take the half with the problem and try to divide that in half., etc.

[RWood]

Very good tips, thanks. I do have a couple of developments to report on this project.

First, after moving the ground probe from the power amp chassis over to the preamp chassis I now measured the -15 rail at only 15 ohms away from ground. But it resorted back to 115 ohms when I disconnected the connector that goes to the jack board. Measuring on the jack board itself now, the -15 rail indeed sits only 15 ohms from ground. So I am thinking the problem started there on the input jack.

Then I got this new intel from the owner of the amp. He said the problem occurred when he turned on the amp with a reverb pedal plugged into Ch 4. And the reverb pedal is now blown, too! It sounds to me like the reverb pedal's problem put voltage on the Ch 4 input and took the Roland out. So now I have a channel to work with and a path to chase. Is my logic here sound?

I see that the schematic for the jack board is not included in the KC350 section of the manual, so here is the input jack board from the KC550 (which I believe is the same as that of the KC350). The first place I see the -15 introduced on Ch4 is D29 and D31.

[RWood]

Smoking Gun

....or at least Jumper. And maybe more.

I had marked the jumpers on the Pre Board that were along the B- line, measuring resistance to ground, as Enzo suggested. I came across J36 and saw that it been part of some kind of "event". Check it out:



J36 connects, on its left end, Pin 4 of IC2 (which is its' B- supply) and C206 which is a cap to ground. On the right end of the jumper it attaches to IC3 Pin 4, C284 and C207, both of which are caps to ground.

I lifted one end of J36 and the components on the left which had been 125Ω from ground were now 345Ω away. The components to the right, which are closer to the B- source, were still 125Ω away from ground. I did not lift any of those caps or ICs (its getting too late) but I wanted to post what I found here in case anyone has ideas of what I should do next.

Thanks

[Enzo]

OK, so the supply side remains low resistance, leave that jumper open, and move on to another. We want to divide and conquer.

It does appear something happened there, looks like sparks flew, and they may have affected the -15 line, though the actual failure may be elsewhere also connected to -15. Don't try to puzzle it out, just isolate the problem.

[The Dude]

Turn that chassis upside down and give it a good shake. I'm wondering if there isn't a screw, nut, or something rolling around inside that may have caused the arc. It may be lodged somewhere. Have a careful look around.

[RWood]

C284 (right above the burnt jumper wire in the photo above) was short - I removed it and it was the reason for the low resistance. I replaced that cap, put in a TIP42 in place of the Darlington, and at low voltage it looked promising. I put the meter on the TIP42's emitter and raised voltage up to about -7.5vdc there, before things started to go bad, high current draw enough to blow the two 1/2A fuses on the low voltage line.

I did have all the connectors in place; next time I will remove the one to the pre-board and see if the current draw is okay. At least from a resistance standpoint, that is not among the problems. I checked the areas that had previously had low resistance and they were all ok.

I am wondering if, when that cap shorted, it took out the IC that is right next to it. Or all of them. Can they be tested? It's an M5218AL.

Next steps?

Dude - I've had the boards completely out of the amp, nothing rolling around. Thanks though.

[The Dude]

The M5218AL is an op amp. The data sheet is here:

www.datasheets.pl/integrated_circuits/M/M/M5218al.pdf

You may want to build a light bulb limiter to save fuses and hassle. If you can get the unit powered even partially (not full voltage and holding fuses), check to see if any of the op amps are getting hot- or anything else for that matter. Also check to see that only pins 4 & 8 of the op amps have supply voltage on them. If an op amp is shorted, it will likely have rail voltage on other pins where it should not be (normally).

[J M Fahey]

If you can't get it, M5218 is close to 4558 , same pinout version of course, and in many applications can be freely substituted.
Is yours DIP or SIP?
You should have no problem except, maybe, if used at extremes of capability (100X gain, driving a very load impedance, etc.)

[RWood]

Thanks for all the responses and help. I had to step away from the amp for a few days.

I am able to power things up, using about 1/3 of the wall voltage. The Dude, I do have a dim bulb and used it (Post #1) but with this being on the low voltage and low current side it didn't make much of a difference. Anyway, I was able to measure some voltages this week.

What was strange about measuring both the - rail at the new PNP on the power supply, as well as the existing NPN on the + side, was that the meter itself really loaded down the circuit and the voltage would drop any time I probed it. Meter is a Fluke 115. If I waited a few moments and tried again, the voltage will have crept back up. I don't know if this is normal; it surely isn't with the tube amps that are my usual fare.

Speaking of tube amps, I treated the op amp pins much like the pins on a tube socket and wrote down all the voltages. Although this amp uses eleven M5218s, I focused on the first four, corresponding to the four input channels. Despite the ever-fluctuating voltages due to the meter load-down, I was able to compare the four channels' results and what jumped out was that Channel 1 had four pins that were showing much lower readings that those same pins on the other channels. And of those four pins, two were to one side of the pin for the -15V source, and two were to the other side of it.

I was a little surprised that it was channel 1 that seems bad, since it was Ch 2 that had the shorted cap and Ch4 which was plugged into when the amp went out. But it is the farthest downstream, so maybe that counts.

I'm thinking of doing any or all of the following:
- pulling Ch 1 op amp out completely and seeing if the other channels work,
- putting an 8 pin IC socket in so that I can plug-n-pull the op amps as needed,
- ordering the NTE equivalent to try in the socket.
What do you think?

I also measured for temperature (different meter with a temp probe) but nothing at all jumped out on the ICs however the TIP42 that I installed, with the metal tab, got very very hot, like 60°C compared to the 40° or so on the plastic-tabbed NPN or the 32° of the opamps. I haven't yet placed an order with MCM (their $11.99 minimum on shipping is causing me to make sure I get everything I need[yeah right]). But I do intend on replacing that PNP and also I am using a 14V zener diode there as I did not have a 16V. I don't expect that would be causing some of these conditions but I report it here in case it might. And I will be ordering some 16s

That's the update from Richmond, let me know any thoughts or things I can try. Thanks!

[The Dude]

If you don't have an op amp, yes,....I would pull the suspect one out and see if the unit becomes functional. No need to replace it until you verify.

[RWood]

I pulled out the op amp for Ch1 but still couldn't get any signal through the other channels.

I think there continues to be something wrong with the power supply, as the negative voltage only gets to about half of what the positive voltage is. And the TIP42 gets exceedingly hot (burned my fingertip). Perhaps it needs the "C" designated TIP42, and I still have the 14v zener in there. I will replace those two parts and report back.

One positive is that with the op amp removed, the voltages were no longer being loaded down by the meter. It sounds like there a few problems going on at once here.

[Enzo]

TIP42C is the 100v version, the plain old TIP42 has plenty of voltage rating for this 15v supply. it I shot because something is loading down its supply. If that transistor gets hot, feel each IC, any get warm? If you drive down the highway with the parking brake on, the engine will overheat. It doesn't mean there is something wrong with the engine. If no ICs are getting warm, is anything else getting warm on the board?

I seem to be missing part of the schematic, but if pulling IC1 restores the power supply voltage, then it is likely defective. The fact it doesn't allow signal to pass might just mean is is in the signal path for all channels.

[km6xz]

IC1 a and b sections are the right and left out of the preamp so if that is bad, no channel will pass signal. It is a UPC4570

Just simplify the current status, what is the voltage to ground at the emitters of both Q1 and Q2 with the cable W7 is connected to CN12 and when the cable is unplugged. With it unplugged, it should be 15 volts positive or negative depending on the transistor. If that is good working without the cable connected to CN12 move on to the preamp board. If the -15 is low or missing. Stop working with the preamp until the supply is repaired.
When it is working, isolate the branch of the high current drain on the -15v rail by doing as Enzo said, remove jumpers or break the rails in ever reduced sized of 1/2 the remaining branch that is pulling current.

[RWood]

If you can't get it, M5218 is close to 4558 , same pinout version of course, and in many applications can be freely substituted.
Is yours DIP or SIP?
[/IMG]

Thanks for the sub suggestion. It is a SIP 8. And ....... it is not even a 5218. It is an NJM4580L! The 5218 came from the schem but the actual part from the amp is a 4580. Which is in stock at Mouser for 54c!

[RWood]

Thanks km, I was working on the amp when you replied so I have some of that info. Btw, it was IC2 that I removed a few days ago, not IC1. IC2 is for Channel 1. And it is an NJM4580L despite being listed in the schem as an M5218AL.

Here's where I stand with the amp:

With the cable to the preamp board unplugged, the power supply is very close to right-on and the fuses don't blow, even at full wall voltage. The positive rail is at 15.67vdc and the negative is at -14.18vdc. Keep in mind that the zeners are not the same value from the + rail to the -. On the - I am now using three 5.1V zeners in series. Also, I do notice that when I cut the power off the negative rail drops voltage more quickly than the positive does.

Plugging in the preamp board with full voltage will blow the fuses, so I dialed them up to around 6v (+6.3 and -6.1vdc on Q1 and Q2 respectively). Plugging in the board now, the negative rail loaded down and now measured-3.3vdc while the positive went down to only 5.1vdc.

With the low voltage having run for around five minutes I probed the temperature of every cap, IC and Q on the preamp board. At first I didn't notice, but then confirmed that while every other component was sitting at 29 or 30°C, one of the ICs and two ceramic caps next to it were at 34°. They are IC13 (another NJM4580L), C223 and C217. Possible culprits? It looks like these feed the headphone jack, and I did have phones plugged in (I could hear them tic-tic when I plugged in the preamp board, and I also could hear RF broadcast when I temp probed the metal casings of the electrolytic caps near the headphone amp. Amusing.

Enzo, the service manual does not have a schematic for the KC350 jack board and pre-board, as it does for the KC550. And there are some differences. But I have attached the one that we do have, for the 550.

Thanks everyone for the help.