Rockford Fosgate Power 650 Mosfet "rejuvification"

[pgerry]

I've had this amp stowed away since '95, bought for a project that never happened before I sold the car. I finally decided to put it to good use, then discovered an issue while setting the gains.
Using a 1k test tone and a scope on the outputs, I noticed an enormous amount of noise in the sine wave. It appears there was a 1.5MHz "carrier" also present in the output. The 1.5MHz signal was present regardless of input connection and persisted with just B+ and GND applied, but was significantly more prominent on one channel than the others regardless. So, I opened it up to have a look....then the fun began.
I first suspected it might be an oscillation problem when I saw a trace repair on 2 of the outputs (the same channel where the 1.5MHz was strongest). The repair had been made with a coil of 26 ga insulated wire, that when straightened out was ~6" long. I replaced them with leads that were formed to match the original trace but saw no improvement. A few days later I decided to dive into this again and found little to no output from 3 of the 4 channels. I found a few bad FET's (some N some P) on the 3 weak channels as well as a failed 3055 and the associated zener. I don't see how these could have been bad initially but they measure bad now. This amp came with RFP18N10 / 12P10's which are less than readily available and I know the later models came with IRF540 / 9540's in them which are available most anywhere. What I'm unsure of is if the gate resistor value remained at 75 Ohms for both models and if the 540/9540 is really the best substitute.
I beginning to think I've opened Pandora's box, maybe hope is still inside. Any thoughts on any of the above are welcomed.

 

[seankane]

Have you found any bad capacitors in this amp, as old as it is they are probably leaking or dried up by now.

I have had good luck substituting IRF540 and IRF9540 in these amps.

 

[Perry Babin]

These amps can be noisy if not properly biased and they bias controls are very touchy because there are no proper source resistors.

Also, look at the signal with the crossover set to low-pass. Many Rockford amps look noisy but the noise isn't audible.

 

[pgerry]

The only pair of caps I formally tested were the 16V 2200uf at the input and surprisingly they tested good. The other caps have not been, but hold voltage well for quite a while. I've ordered replacements as a precaution anyway.

I did not get to the point of verifying the bias settings before component failures began. I expect a little noise from these amps but was a bit alarmed when (with the inputs disconnected) I had just over 4v P-P at 1.54MHz on the output. Isn't that roughly the switching frequency? could I possibly have a failed LP filter?

Do you know if the gate resistance on the later (green board) models with the 540 / 9540's was still 75 Ohms?

Thanks for your help. I'm determined to get this amplifier back in good order.

 

[Perry Babin]

I don't have the value of the gate resistors. The older ones didn't have conventional gate resistors. Contact Rockford with the board number to get the schematic diagram.

The power supply operates at ~25kHz. The higher frequency noise is possibly generated by the power supply but it's not from the main oscillator. It's not likely to be audible and could be common.

When you removed the board from the heatsink, did you see any spots on the sink where the black coating peeled off?

 

[seankane]

I have a schematic for this amp whats your e-mail?

 

[pgerry]

I have the schematic that Fosgate has for this amp dated 1984. My particular board is a rev B and so far all components and values are a match with those noted on the schematic. I don't know the exact year of the amp but would guess it to be early. If you have a different one I'd love a copy. philgerry@gmail.com

Thanks

 

[pgerry]

Perry - The heatsink is pristine with no evidence of damage to the anodized coating. Rockford provided me the schematic and claim it's the only one they have available. It appears to be identical to my board.

 

[pgerry]

Ok - I've made a little progress but wanted to verify a couple things before I install the new outputs. With all of the outputs removed I power the amp up and noticed the current draw to be ~1A through 1 Ohm current limiting resistor which seemed a bit high. I removed power ,and before discharging the caps. checked voltages in the power supply. The negative side was near 0 within a few seconds while the + side held voltage longer than I was willing to wait.
I then removed the 2955 from the circuit and powered the amp again. Current draw was ~.5A and when power was removed, the negative side of the power supply retained voltage like the + side.

The MJE2955T tests good as does the associated zener and resistor. Voltage on the pads (with transistor removed) are:

B = -17.9
C= -40.6
E= 8.85

Seems like something on the negative supply side is faulty but I'm not sure where to go from here.

Thanks for any advice or suggestions.

 

[Perry Babin]

1 amp doesn't seem excessive.

This isn't a common way to troubleshoot an amp. For someone to answer this type of question (if one rail drops more quickly), someone would have had to have tried that. It's possible that there is a leaky capacitor but it's also possible that there is no problem.

I'd suggest installing the new parts for one channel and reinstalling the regulator transistors. Does the one channel work properly?

 

[pgerry]

I was preparing to install the outputs but wanted to make sure the regulators were behaving 1st. The voltage measurement was done to ensure I had bled the caps. I said the measurement was taken before removing B+, what I meant was after. I normally attach a meter while bleeding a power supply to ensure it's truly discharged. I just thought it was odd to see such a discrepancy between the two sides. I'll proceed with one channel and see how it goes.

Thanks

 

[pgerry]

Finally found some time to work on this again. Installed the outputs for one channel (Grey/Green), powered the amp up and all looked fine. injecting a 1k sine wave produced a nice clean output. I moved on to the second channel (White/Blue) and had good results as well. The third and forth (orange/black and yellow/brown) produce no audio output. idle current is at ~2.5 amps. Where should I be looking next?

 

[Perry Babin]

Do you have both positive and negative rail voltage on the output transistors for those channels?

 

[pgerry]

Yes I have -40.9 and 40.7 present for all outputs.

 

[Perry Babin]

Are you sure that you're sending signal to the correct inputs on the amp?

Is the input selector switch in the correct position?

 

[pgerry]

I wondered the same and cleaned / verified the selector switch and ohm'd out the inputs on the DIN input. I do get a sinewave output from the two channels but it's ~1v P-P and at a fixed 1.5 MHz.

 

[pgerry]

Perry - you nailed it. The selector switch has a third position that is normally locked out with the cover installed. I was going past the second position by mistake. I'm going to let it burn in for a while and re-check before closing it up, but so far it looks good. Any other areas on this amp that I should look out for before it goes back in service?

Thanks again for the help Perry & Sean.