Memphis 16-ST1000D

[Dr Zeus]

I've worked on a few of this series already, but this one is the worst. I'm not even close to done digging through but here the starting hit-list:

All IRFZ44 fets are bad. 22 ohm gate resistors are good.

All outputs are bad IRF640/9640

IRF640s are in Q215, Q216, Q225, and Q226.

I'll be placing one of each bank of transistors at a time for evaluation and function and will report back what I find.

 

[Dr Zeus]

Found Q15 and Q16 drivers on the PS driver board out of spec. A1266 transistors. Amp now holds +-60vDC at the rail with the outputs out of circuit.

I'll work on this little guy some more tomorrow. peace!

 

[amc32]

I personally love working on these little amps. PS and Output Driver Cards unplug I have tons of rebuildable cards. The key is to keep a functioning amplifier on hand makes troubleshooting much easier. If you need rebuildable cards or photos let me know. Forget the IRFZ44's and go with the IRF3205's or it will be back again as soon as the customer gets drunk one night.

 

[Dr Zeus]

What originally came in this amp? I could barely see the identification marking on the PS Fets and could only barely make out IRFZ44. This amp is running with 22ohm gate resistors for the PS fets.

 

[Perry Babin]

As far back as the board revision PAM280-09, the 3205s were the original parts.

 

[Dr Zeus]

Hmm so the IRFZ44s are likely not going to work if 3205s were original. I wonder why this amp had them in here? Or perhaps they were IRFZ48s - I cannot tell because all the fets were distroyed.

 

[Dr Zeus]

In powering up this amp with several IRF640/9640s installed (And IRFZ44s), the new IRF9640s get hot within seconds. Pulled both driver boards and they stay cool. I've narrowed the drive circuit i nthe following photo where you can see the lead is attached, and the trace<s> are leading to the three drivers on the top right corner of the board.

-From Left to Right last three on the right-
Q201: C3503-D
Q202: A1027
Q203: A1023

 

[Perry Babin]

Is there any DC on the tabs of the outputs with no driver boards installed?

 

[Dr Zeus]

Yes as follows:

IRF9640s have +64vDC on pads 1 & 3, and 0vDC on pad 2.
IRF640s have -64vDC on pads 1 & 3, and 0vDC on pad 2.

 

[Perry Babin]

Tab = top of the transistor.

It appears that you have none.

The PNP drivers, the 1 ohm resistors and the 12v zeners are all suspect.

 

[Dr Zeus]

A 1ohm resistor just popped off the board with minimal pressure. its location is R202 and is in the very corner, component side, to the right of the above 3 marked resistors. This 1 ohm resistor is actually still measuring OK off the board - Its probably damaged through. Seems only one driver board is like this.

Seems all the through-hole diodes on both driver boards are measuring ~0.650 and checking OK while in circuit via my Fluke 10. i think they are OK.

Next up: testing the drivers out...

 

[Dr Zeus]

All surface mount diodes seem to be OK.

 

[Perry Babin]

The Zeners are generally silver or orange leaded components, not SMD. Sometimes you have to apply power and confirm that the voltage across them is OK. 11v-13v is OK

 

[Dr Zeus]

I got this amp to power and produce clean audio with all boards installed. All PNP drivers checked out OK - All the problem was - that single 1ohm surface mount resistor on one of the driver board.

Time to put it back together and order some new PS fets! Thanks!

 

[Dr Zeus]

Befor now, I was able to get the amp to play clean audio through a test speaker; but that was when only 2 output transistors were installed on each side. Now with all 12 output transistors installed the amp experiences problems.

Hmm I'm stumped! Q214 / Q254 outputs get hot after just a few seconds of 5A supplied power. With the driver boards removed they stay cool. I pulled all 12 small transistors off one of the driver boards and they are all checking out.

 

[Perry Babin]

Are the gate resistors within tolerance?

Do you have a good gate signal on the gate leg (not on the board)?

Is the amp drawing excessive current?

It's possible that the transistors in question have a slightly lower threshold and are turning on a bit quicker than the others. Swap these with an adjacent transistor to see if the heat follows the transistor or the location.

 

[Dr Zeus]

Gate resistors are all measuring up fine. There are 10 ohm resistors and 100k ohm resistors. In circuit the 100k ohm resistors are measuring up at 47k ohms.

Amp is drawing more than 5A of current as my power supply has a really tough time powering the amplifier up. After a few seconds of it power cycling the outputs in question are already hotter than can be touched. They are still measuring OK via DVM.

I removed the warming pair of IRF9640 and the heating problem just jumps to the next transistor. It appears the defective range and heating problem is at Q213/Q214 and Q253/Q254. It appears to me that the driver boards have three driver sections each pertaining to a pair of outputs.

Im not 100% sure how to measure the gate signal. I set my scope to AC coupling and shielded the probe to tap. Leg 1 of the outputs only shows some basic static and nothing really signal related.

I'm still trying a few things though.

 

[Dr Zeus]

Now Ive got a wierd one going on. I'm trying to get the amp to power with just one or two IRF9540 outputs and they are just dropping like flies. I power up the amp and instandly the IRF9540 I put in goes dead shorted without heating at all. This is definately wierd.

 

[Perry Babin]

Did you confirm that the 12v Zeners were all within tolerance?

 

[Dr Zeus]

If the 12v zeners are D201 and D202 then they measure OK via the DVM (Beeps and 0.700). While powered they measure ~12 across them. There are other through-hole diodes on the driver boards and they all seem to be ok for what I can tell.

With one driver board installed, I am getting what I think is erroneous voltage readings on the gate pads. Whereas before leg 1 & 3 were meauring +63vDC, with the driver board installed leg 1 measures +54vDC on Q213 and Q214 and +56vDC on Q211 and Q212. Leg 3 always measures +63vDC on these. When the driver boards are removed the voltages on Leg1&3 are +63vDC.