Yet another GK MB150 combo PS has failed...

[Liron_al]

Hi,

I have decided to revive my beloved GK MB150 bass amp.
It is old, really old, the board says 1990, it has the switching power supply and it has failed.

I took it to a local repair guy which suppose to be good at repairing guitar amps but unfortunately, even I know more then this guy.. and I'm a complete idiot
as for now, and because I have no idea how to debug this thing, i'm going to replace the two transistors (BUK456-800) that I already have,
the Zener diode (1n4741) cause I have no idea how to test it,
the IC PWM controller and some resistors that appears to be burnt.

any other parts you would replace while its open on the operations table?

also, a question for those who knows this thing,
the two BUK456-800 transistors should be attached to the heat sink. the questions are:
1) are they suppose to be forced against a heat conducting pad or directly touching the heat sink? i mean should they "short" the transistor's heat sink to the case heat sink? or should they be pressed with plastic tabs to prevent shorts.
2) same question for the two BYW-29-200 diodes..

I found this schematic a while back if it helps.. 206-0071-E switching supply.pdf
Thanks for all who can help I really like this amp and would hate to throw it to the bin..
L

[52 Bill]

Do you have a meter with a diode test function? If you do you can use it to test all of the diodes in the power supply as well as the two mosfet transistors. If not, then an ohm meter will probably suffice for the diodes.

The back surface of the mosfet is a metal plate than must be isolated from the surface of the heatsink. Check the grey silicon pads for holes and squeeze through points.

[nosaj]

Welcome to the Forum.
First off the guys here will tell you not to replace parts until they are determined faulty.
As for your transistor questions, How were the originals mounted? Are the replacements the same? I had a Hartke module I purchased from Enzo and it used the same model transistor as my old module but the originals had a plastic back so they could mount directly to the heatsink. The replacements had a metal back so I had to use an insulating sheet so they would not be grounded to the heatsink but still pass off heat.

And last but not least what does failed mean? No power?. lights? etc. Do you have power on the other side of the main fuse?

Thanks,
nosaj

[Liron_al]

Yes, I do have a multimeter with diode test.
The zener is bad, it reads 4.5k and 6k when off the board.
Do I need to test all diodes off the board or van I leave them on?

Also, how do I test the transistors?

L

[Liron_al]

Thanks.
Main fuse blew once. No lights after replacing with the exact same one.
One resistor blew.

10 4 on the heat sink. The transistors I got are the same ones but lost the pads and I couldn't remember how they were mounted

L

[52 Bill]

I always suggest testing the diodes in circuit first and then if there is a perceived problem, lift one leg to test out of circuit.

The mosfets will not test like a normal transistor, but you can basically test them to see if they are shorted.

[Liron_al]

All parts are in the unit,
I took the boards off about two years ago and now decided to fix the damn thing..
I lost the pads and the transistors are not mounted to the heat sink, that's why I ask.

As for the transistors, they are not N and P type, they are the same, wired in parallel (Q522 and Q513 in the schematics).
The guy who fixed it last time replaced only one with a different ("but close in values".. pffff) one, it worked for 10 minutes until the first slap.
I believe that one failed due to uneven load sharing and the other failed right after, but I'm no expert.

L

[olddawg]

If it's a stand alone power supply and not integrated to the main board you can work on it in isolation. If you were going to shotgun it. Rebuilding this type of power supply is a valid course of action. I would replace all of the active components, diodes, and electrolytic caps. Some of the diodes are probably special "fast switching" or other types. Make sure you check all of the resistors with one leg lifted if you don't replace them. . An open 500k resistor can burn you down. Also check the output stage if you can. An overcurrent condition could have taken it out in the first place.

[lowell]

Yes, I do have a multimeter with diode test.
The zener is bad, it reads 4.5k and 6k when off the board.
Do I need to test all diodes off the board or van I leave them on?

This does not mean that a zener is bad. A "semi" test function your meter, the diode symbol, is somewhat useful but not always. Be sure for zeners that the red, positive probe, is on the cathode. Zeners are operated backwards from regular diodes...in their reverse breakdown region. Your semi function maybe not read higher voltage drops and if it's a higher voltage zener it may show OL. Not sure zeners can really be tested accurately without applying a voltage above their rating. IE 15v for a 12v zener. Someone else here may have a better suggestion.

Switch mode supplies aren't easy but I'd be happy to assist in any way I can.

[lowell]

Leave the mosfets out and see if you have +/-165v after the bridge rectifier. And ofcourse be extremely careful as there is no isolation transformer. You'll want to use the junction of the two filters as your ground reference.

[nickb]

This does not mean that a zener is bad. A "semi" test function your meter, the diode symbol, is somewhat useful but not always. Be sure for zeners that the red, positive probe, is on the cathode. Zeners are operated backwards from regular diodes...in their reverse breakdown region. Your semi function maybe not read higher voltage drops and if it's a higher voltage zener it may show OL. Not sure zeners can really be tested accurately without applying a voltage above their rating. IE 15v for a 12v zener. Someone else here may have a better suggestion.

Switch mode supplies aren't easy but I'd be happy to assist in any way I can.

A zener, measured by the average low voltage diode test function will read just like a regular diode. In one direction the forward voltage is usually displayed and "infinity" in the reverse direction.

You measured 4.6k & 6k? They don't usually display ohms, well none of the ones I've ever had did. A busted diode will usually be either measure open or short. That measurement is just weird. Best just replace if there is any doubt.

As for the FETS, I think they would have matched them prior to assembly to get stand a chance of getting the current to share. I would not buy almost certainly fake replacements off eBay, but rather use a single modern FET of known type with ratings equal to or better than the two BUK465's which will have otherwise better performance.

[Liron_al]

thanks guys for trying to help me solve this one..
it seems that i missed your recent replies, i think i've made some progress, but still confused:
when measuring the voltage between the DC output after the bridge rectifier, i get 320VDC.
the problem is that when i test the positive and negative rail in reference of the center of the two large capacitors (330uf), i only get 106VDC and -106VDC (????).
how can that be possible? bad rectifier?
when i move forward, i get only 8VDC across the zener diode instead of 11.4v, even when the zener was taken out of the circuit.
i think that because of the low voltage, the UC3843 chip does not operate (rated to 12V) thus not driving the mosfets.
any thoughts?
BTW, board's jumper set to 234V (220V here..). even tried 117V setup but still the same results.

Thanks
L

[nickb]

thanks guys for trying to help me solve this one..
it seems that i missed your recent replies, i think i've made some progress, but still confused:
when measuring the voltage between the DC output after the bridge rectifier, i get 320VDC.
the problem is that when i test the positive and negative rail in reference of the center of the two large capacitors (330uf), i only get 106VDC and -106VDC (????).
how can that be possible? bad rectifier?
when i move forward, i get only 8VDC across the zener diode instead of 11.4v, even when the zener was taken out of the circuit.
i think that because of the low voltage, the UC3843 chip does not operate (rated to 12V) thus not driving the mosfets.
any thoughts?
BTW, board's jumper set to 234V (220V here..). even tried 117V setup but still the same results.

Thanks
L

Are you sure you are seeing +/-106V and not 160V or was that a typo?

Did I read that correctly? You switched the link to 117VAC and applied 220V to the input? If you did then instead of 320V your would get 640V making it an incredibly stupid thing to do. That much I'd expect would likely blow the fuse, the bridge rectifier, the main electrolytics and the FETs. If one or more of those things didn't blow then I suspect something is not right with the primary DC power supply and if there wasn't there probably is now.

[g1]

If memory serves, these are the units that GK offered the linear supply retrofit for, due to reliability issues with this smps.
Not that they can't be repaired, but just to make you feel a bit better if you did run it set for 117V.

[Liron_al]

I guess I didn't explain my self correctly,
When tried the 117v, it was hooked to a 220 to 110 transformer 
Nothing blew.
106v. That is not a typo. Have no idea what's going on.
Wouldn't that indicate that the bridge rectifier is bad?