Kustom 400, burned 100 ohm resistor

[davohilts]

Hello,
A friend brought over a Kustom 400 amp which began to smoke and flame inside (he saw this thru the crack in the back of the chassis and tuck and roll). From what I've been able to gather, this is a stereo/mono guitar amp and it looks like it is two Kustom 200's combined into one big amp. There is a stereo/mono dial below the blue Kustom pilot lamp and has 2 inputs for each channel and two power amp boards.
The circuit board numbers are PC 303, PC 403, PC 704A and PC 704B (these two boards are the mirror image of eachother with some small differences).
When I opened up the amp I looked around inside and finally noticed the burned up resistor, a 100 ohm, 1/2 watt. The 1 K resistor next to it was also toasted and the board has a nice black streak up it where the fire and smoke went. The burned resistor is on the PC704A board and is next to a RCA 38736 CXJ 8L transistor.
I'm wondering is there is anyone who can help me in trouble shooting this issue. I am not familiar with solid state electronics but do work on a lot of tube amps.
I'll attach some pics of the board.
Thanks,
DaveKustom PC703 power amp+Vreg.pdf

[52 Bill]

Your description is correct, it is just two separate power amps built into one case. The 100 and 1K ohm resistors are related to the driver/output transistor that they connect to. One or both will be shorted.

Test each power transistor for shorts and then test each driver transistor for shorts. Use the diode test function on your meter. Pull off the connector on each power transistor which will remove it from the circuit. Now take readings from the base to the emitter, from the base to the collector and most importantly from the collector to the emitter. You may get lucky and only have one or two bad transistors. Worst case is that all of the drivers will be shorted.

Test each of the driver transistors (the ones with the heatsinks) the same way. Because the 1K resistor is part of the protection circuit, you should check those transistors as well. Well for that matter you should just test all of the transistors on the board.

You should take some time to remove that carbonized areas from the pc board. Because there are two separate power amps, you can use the second channel to compare your readings to. These are some of the most basic power amp designs out there, so it will make for a good learning experience for you.

[davohilts]

Thanks for the help 52 Bill,
I am pretty much in the dark on transistors so this is going to be a real learning experience. It looks like there are 19 transistors on the board and 4 additional power transistors (big ones in large heatsinks) off the board along the bottom. I've desoldered these to allow me to get the board out and get a better look at it. Now, the 19 transistors on the board, you advise to "disconnect the connector on each power transistor which will remove it form the circuit" and I am uncertain about this. Do you mean disconnect the 'collector' and if so, do I desolder it and pull it out? I'm kinda lost about how to proceed. I know the collector is the lead of the transistor that is opposite the emitter but does that mean it is the positive side? Is there a way to tell by looking? I'm sorry for such a basic question but I am not familiar with transistors at all.
Dave

[52 Bill]

Please do yourself a favor and don't unsolder anything that you don't need to, you'll just make things more difficult for yourself.

I think I can see 10 transistors on the board and the 4 power transistors on the chassis. The black connector with the blue and yellow wires will pull off the pins of the power transistors. Pulling the connector will leave the Collector (the case) connected but the transistor will be out of circuit. The pin that the blue wire connects to is the Base and the yellow wire connects to the Emitter. Touch one lead from your meter to the base pin and then touch the other lead to the emitter pin. You should either get no reading or a low reading. Now reverse the leads, if you had no reading you should now get a low reading or if you had a low reading, you should now get a no reading. If you get a reading in both directions, the transistor is shorted. Repeat the test with one lead on the base and the second on the collector (the red wire on the case mounting screws), and again test from the emitter to the collector.

You will then need to test the transistors on the board the same way.

[davohilts]

I have followed your instructions and have tested the transistors and the one power transistor below the fried resistor is in fact bad. Now my task will be to find a replacement. The bad transistor is made by rca and has the following written on it: 36892 CUC 9B. Randomly I happen to have two transistors which are made by NTE and has NTE130 042F written on it. Is this a suitable substitute?
Another issue I noticed when I pulled out the bad transistor is that there is some kind of grease and a clear washer between the transistor and the heat sink. I assume this is something to help transfer the heat from the transistor to the heatsink.
Thanks for all your help.
Dave

[52 Bill]

Yes, the NTE130 will probably work as a replacement. The original transistor is a house numbered RCA 2N3055. If you go this route, I'll suggest replacing both of the paralleled outputs on that side of the amp so that they will work together better than a mismatched pair.

There is heat sink grease or compound as well as a thin Mica insulator on each of the power transistors. You should clean up the old stuff and replace it with new compound. Inspect the insulators and see that they are not cracked or otherwise damaged before you reuse them. You can probably get heatsink compound at Radio Shack if you don't want to order from your regular suppliers. When you mount the new transistors, check with your meter to see that there is no continuity from the chassis to the transistor cases before you connect the red wires.

It is important that you also check the driver transistors as well as the protection circuit transistor that is on the board before you power up the amp. It is fairly common for the drivers to short when the output transistor is shorted. In fact with the burned 100 ohm resistor I'd be more surprised if the driver wasn't shorted.

In fact, I think that I mentioned before that you should just go ahead and check all of the transistors and diodes on the pc board before you go any further. And seeing that there was a dead power transistor be sure to check the 1 ohm 5 watt ballast resistors as well.

This might be a good time to look into a light bulb limiter, as it will save your new parts if you didn't find all of the shorted things.

[davohilts]

I have a light bulb limiter and have been using it, the light glows nice and bright so I knew I had a serious issue.
I do have 2 of the NTE130 transistors so will do as you suggest.
Now on to the other transistors on the board. Should I remove and test them one at a time or can I test them on the board. There are 4 in smaller heatsinks on the board that have 2 each of rca transistors with numbers 2N38736 CXJ 8L and 2N38737 CXC 8G, one RCA marked 38735 CXV 9C (on board without heatsink), two marked 2N4248 S 81 835Z, one marked 2N3638 81 836Z, one marked PET 8002 81 834Z, one marked PET 3567-S 81 821, and a two lead RCA 1N3754 FR 9A.
I have tested the diodes and one tests bad, it is a clear small diode next to the row of burned resistors.
So the transistor in the heatsink next to the burned resistor is suspect for sure, owing to the bad diode and bad power transistor. Should I pull all these and test? What about replacement? Do they still make something I could replace them with?? Should I replace both that are associated with the two power transistors I'm replacing?
Thanks!
Dave

[52 Bill]

Should I remove and test them one at a time or can I test them on the board.

I will always recommend that you start by testing in circuit and then if there is a problem due to in circuit resistances or if you find something suspect, you pull the transistor to do a final test. Pulling parts for testing usually causes more harm than good.

The drivers are house numbered RCA 40409 and 40410 transistors, which haven't been made since the 1980's. They can be replaced with other TO-39 type transistors or can be replaced with TO-220 cased parts.

[J M Fahey]

The drivers are house numbered RCA 40409 and 40410 transistors, which haven't been made since the 1980's. They can be replaced with other TO-39 type transistors or can be replaced with TO-220 cased parts.

Agree but add: 40409/10 have built-in tinned iron heatsinks, with their own mounting legs, which obviously are connected to collectors and to each other.
*Often* some designers use such heatsinks as links in the PCB, meaning that a track reaches one of them, another track continues from the other side.
If you replace the transistor by a modern TO220 one, part of your PCB may be disconnected .
Just don't ask how "I" found out .
So check that and if necessary restore continuity with a small piece of wire.

[nickb]

To all this good advice may add the suggestion to replace the 100 ohm and 1K resistors with flameproof types. Regular carbon ones are a fire waiting to happen.

[davohilts]

by paralleled do you mean Q1 and Q2?

Yes, the NTE130 will probably work as a replacement. The original transistor is a house numbered RCA 2N3055. If you go this route, I'll suggest replacing both of the paralleled outputs on that side of the amp so that they will work together better than a mismatched pair.

There is heat sink grease or compound as well as a thin Mica insulator on each of the power transistors. You should clean up the old stuff and replace it with new compound. Inspect the insulators and see that they are not cracked or otherwise damaged before you reuse them. You can probably get heatsink compound at Radio Shack if you don't want to order from your regular suppliers. When you mount the new transistors, check with your meter to see that there is no continuity from the chassis to the transistor cases before you connect the red wires.

It is important that you also check the driver transistors as well as the protection circuit transistor that is on the board before you power up the amp. It is fairly common for the drivers to short when the output transistor is shorted. In fact with the burned 100 ohm resistor I'd be more surprised if the driver wasn't shorted.

In fact, I think that I mentioned before that you should just go ahead and check all of the transistors and diodes on the pc board before you go any further. And seeing that there was a dead power transistor be sure to check the 1 ohm 5 watt ballast resistors as well.

This might be a good time to look into a light bulb limiter, as it will save your new parts if you didn't find all of the shorted things.

[davohilts]

I've gone thru and tested the transistors and diodes and have found that CR703 has failed as well as the associated transistor Q706. The two 38736 transistors Q704 and Q708 are associated with the area around the bad power transistor Q1 and the blown resistor R722 and I tested them in place. On Q704, I'm not sure which pins are which but with the neg. lead of the dmm on pin 1 and the positive on pin 2 I get no reading, when I switch them I get .64 VDC. Neg. lead on pin one and pos. lead on pin 3 I get no reading, when I switch the leads I get .87 VDC. Pos. lead on pin 2 and neg. lead on pin 3 I get .89 VDC and with leads switched I get .96 VDC. Does this seem to show the transistor is good? or is my wishful thinking.

I will always recommend that you start by testing in circuit and then if there is a problem due to in circuit resistances or if you find something suspect, you pull the transistor to do a final test. Pulling parts for testing usually causes more harm than good.

The drivers are house numbered RCA 40409 and 40410 transistors, which haven't been made since the 1980's. They can be replaced with other TO-39 type transistors or can be replaced with TO-220 cased parts.

[52 Bill]

On Q704, I'm not sure which pins are which but with the neg. lead of the dmm on pin 1 and the positive on pin 2 I get no reading, when I switch them I get .64 VDC. Neg. lead on pin one and pos. lead on pin 3 I get no reading, when I switch the leads I get .87 VDC. Pos. lead on pin 2 and neg. lead on pin 3 I get .89 VDC and with leads switched I get .96 VDC. Does this seem to show the transistor is good? or is my wishful thinking.

On all of the small transistors the leads form a triangle shape which is off center from the round shape of the case. The center point or tip of the arrow is the base, the other two pins are the collector and the emitter. If you look at the driver transistors there is a small metal tab that protrudes from the edge of the transistor case itself, not the heatsink. This tab indicates the emitter pin. On the plastic case transistors there is a small flat spot along the outer edge of the case that indicates the emitter pin. This information is available all over the web and the case drawings will explain this better than I can here.

Test the small transistors the same way that you tested the power transistors. Touch one lead to the base and then test to the collector and the emitter. The tests that you ran seem to show that the transistor Q704 is ok, but the collector to emitter test is showing continuity in both directions. This may be caused by the resistors in the circuit or by a bad transistor. To check this, you can remove the transistor and test it out of circuit, or you could try testing the driver transistor on the working side to see if you get a similar reading on the E to C connection.

And to answer the earlier question, yes Q1 and Q2 are on the same side of the circuit.

[davohilts]

I collected all the required replacement parts for the repair and installed them into the circuit board and heat sink. I have replaced the following components: CR703, Q706, Q1, Q2, R726 (it tested at 2.6M ohms not 1 ohm). I put the circuit board back onto the chassis, double checked my installation and connected the amp to my voltage limiter and flipped the switch on and off quickly. The light bulb glowed brightly so I assume I've got further issues. What would be the proper steps that I should take to try and localize my problem without blowing things up? I feel like I must be making some sort of headway because I've replace a lot of failed components, I just haven't found the root of the problem yet.
Thanks for any help.
Dave

[davohilts]

Not to muddy the waters but I'm having trouble matching my board marked PC704A and its almost mirror image PC704B with the schematic for PC704 posted here. There are 3 more circuit boards in the amp, PC 303, PC 403 and PC 504. I'm wondering if the PC704 schematic is a combination of PC 704A (or B) and PC 504.