Power Acoustik BAMF 1250/2

[Dr Zeus]

Working on this amp, I originally found all power fets blown and one bad IRF540N output transistor in the right channel. I reaplced all the IRFZ44 power fets and also the all fets in the rt channel which were IRF540N with IRF540 (Vishay brand).

-Right channel outputs are set of four bottom left of this photo:

Testing the amp out of the sync yielded an ok amp so I put it back in the sync and set the bias pots carefully. The right channel started to get warm when source resistor voltage got to 0.001v so I backed the bias pots out about 1/10 a turn and left the bias voltage to 0.000. Under about a 7 amp power draw from my supply, the right channel starts to clip on the possitive half of the waveform with a test speaker connected:

First reading is a 100hz sin wave, scope set to 5v/div and 2mS:

Second reading is a 70hz sin wave, scope still set to 5v/div and 2ms:

The left channel does not experience this behavior and can hit about ~33vAC on the scope with a test speaker attached. The rail voltage of this amp is +-33.5v.

 

[Perry Babin]

Were the originals IRL540Ns or IRF540Ns?

Have you compared the left channel gate drive signal with the right channel gate drive signal for the high-side FETs?

 

[Dr Zeus]

Whoops.. The originals were IRL540N. I didnt see that before. They are 100v 36A mosfets where the IRF540 are 100v 33A. The charts look pretty close between the two but I guess I'm at the mercy of not only the L, but also the N in the part number.

I'll try testing out the drive signals tonight. Thanks

 

[Perry Babin]

The L version is a logic level transistor that requires less drive voltage. Although the non-L versions have worked in the past, it's possible that there isn't enough drive voltage in this amp. It's also possible that the drive circuit is damaged so that has to be checked also

 

[Dr Zeus]

Have you compared the left channel gate drive signal with the right channel gate drive signal for the high-side FETs?

I turned the bias voltage all the way down. The left channel shows a decent sin wave, and best when the bias voltage set just under 0.001vDC. The right channel is showing defective drive signal. This is with all fets installed but no test load attached.


These images are taken with the same input voltage and frequency, with the scope set to 5v/Div

Left channel drive signal:

Right channel drive signal:

 

[Perry Babin]

Was this taken from the high-side FETs or the low-side FETs?

Was the scope set to AC or DC coupling?

 

[Dr Zeus]

Readings were taken from the fets with positive rail voltage on their center legs. Scope set to AC coupling. DC coupling yeilds similar graphs with the straight horizontal sections rising +1v off of 0 with input source. All referencing RCA shields for ground.

I've poked around a bit further and found various bad states in the drive signal at multiple transistors in the channel. I've just got to search for the bad one.

 

[Dr Zeus]

I thought the BSP254 TO-92 mosfets in Q142/Q242 were bad but they both measure the same out of the PCB. They ocasionally get a 27 ohm reading across leg 1&2. They are probably OK. I pulled them out because the waveform and voltages didnt seem right on Q242's legs:

Q242
Pin 1: -30.54
Pin 2: -31.25
Pin 3: -33.20

Q142
Pin 1: 0.896
Pin 2: -31.71
Pin 3: -1.610

 

[Perry Babin]

You have to check it like you would check a P-channel FET (like an IRF9540).
http://datasheet.octopart.com/BSP254A-Philips-datasheet-147743.pdf

What did the waveforms look like with these transistors out of the circuit?

 

[Dr Zeus]

With these fets removed

Q242
Pin 1: 9.17
Pin 2: -34.05
Pin 3: -33.68

Q142
Pin 1: 31.41
Pin 2: -34.05
Pin 3: -33.61

No AC voltage observed on pads when the BSP254s are removed.

 

[Perry Babin]

Did you check them out of the circuit?

 

[Dr Zeus]

I checked the BSP254s out of circuit and in general they seemed alright. If I I switched and swapped the dvm probes for a bit then I could get both fets to occasionally read exactly 27 ohms in either direction across legs 1&2.