Memphis 16-PR1KD

[RiLoWa]

This amp will briefly cycle audio output and then quit cycling and remain silent. Only the power LED is lit, the other LEDs are off.

Here is what I have on the 494
1: .620V
2: 4.21V
3: .076V
4: .469V
5: 1.531V
6: 3.487V
7: .0157V
8: 13.66V
9: .0156V
10: .0156V
11: 13.66V
12: 13.68V
13: 4.97V
14: 4.97V
15: 4.97V
16: .0156

Q19 and Q33 get hot to the touch.

 

[Perry Babin]

Do you have rail-rail oscillation on the output transistors when it's producing audio and when it's not producing audio?

 

[RiLoWa]

When this amp produces audio, it is very brief. It will be difficult to test it while audio is present.

I did see oscillation (very briefly) and then nothing. The scope was set at 50V div. and 10uS and I would be guessing that it was about 1 division peak to peak that I saw.

 

[Perry Babin]

These amplifiers (like many of the clones) have problems with transistors overheating. I'd replace any of the transistors that have solder connections that show signs of heat stress. The drivers are also suspect. I'd suggest replacing those as well. They are the 8 transistors closest to the output transistors. If that doesn't solve the problem, you'll need to check voltage across various components.

To make this amp reliable, you'll need to install heatsinks on the driver transistors.

 

[RiLoWa]

In another thread, it was recommended to replace the MPSA42s with MPSW42s. Would it be the same logic switching the MPSA56s with MPSW56s also?

Are there suitable substitutes for the C3227, C3200, and A1275?
If so, would it make it better to go with a slightly larger package like a TO-126 for the C3227 and A1275?

 

[Perry Babin]

The W42s operate at the highest temperatures for the drivers so those should be replaced with the W version. You can replace the A56s with W56s if you have them.

I think the substitution would depend on the application. Transistors used in the switching part of the class D circuit will be more critical than those used in non-switching applications.

Larger packages will be able to dissipate more heat. Install the transistors a bit higher than you normally would to prevent the board from being damaged from the heat.

 

[RiLoWa]

I come up with KSA1013OTA for the 2SA1275 as a good sub.

It would appear that an MPSW06 might sub for the 2SC3227 if the legs were ECB instead of EBC.

I do not find a suitable sub for the 2SC3200.

The two harder to find parts are the two that heat up the most as the amp sits right now.

 

[Perry Babin]

In some of these amps, I used a KSA1220A in place of the a1275.

What's the circuit board designation for the c3200?

 

[RiLoWa]

There are two C3200s, Q19 and Q18. It is Q19 that gets hot.

Q33 is the C3227 that gets hot.

The data sheets look almost identical on the 1013 to replace 1275 (Q21) unless I am not looking far enough into the specification. I will note the A1220 as a possible.

All of these are in the transistor array behind the eight driver A56s & A42s.

I found a place that claims to have the correct part numbers on hand (C3200 & C3227), but I have never done business with them.

 

[RiLoWa]

Maybe a BD139 in place of the C3227 at Q33?

 

[Perry Babin]

I think that should work. Also confirm that the Zener connected to the base of that transistor is within tolerance (6.2vZ).

 

[RiLoWa]

I will check that diode.

I am not sure if the MJE344 is a possible substitute for the 2SC3200. Otherwise, I am not finding a good one. The 3200s check ok out of circuit, but the one at Q19 is heating the board below it.

 

[RiLoWa]

I measure 5.93V accross D8. Not sure of the tolerance but that is within 10% of 6.2V.

Should be OK?

 

[Perry Babin]

I think it should be close enough. Did you replace the drivers yet?

 

[RiLoWa]

No sir, I have not. I will have to order those parts.

Should I order the BD139 and the KSA1220A along with them?

What about the the 2SC3200? They test OK out of circuit.

 

[Perry Babin]

I'd suggest ordering the following just to have them on-hand.

BD139 and 140
KSA1220a and KSC2690A
MPSW42 and MPSW92

Also order the clip-on heatsinks for the TO-92 transistors.

 

[RiLoWa]

Replacing the eight drivers with w42s and w56s did not change the symptom. The original set checked out ok out of the circuit. I forgot to get the parts you posted just above, so I will have to get those if they become vital to continued testing.

 

[Perry Babin]

Did you replace all of the other transistors that showed signs of overheating (the ones with bad solder joints)?

 

[RiLoWa]

After replacing all of the transistors except for a pair of 2N5401s, the amp appears to be working correctly. I think it may have just been a broken leg on Q21 (A1275) that was the problem which was the last transistor I replaced. Hard to know for sure, I will have to do more testing on it tomorrow.

 

[Perry Babin]

If you want this amp to be reliable, long term, install heatsinks on the transistors that get too hot to touch within 1 minute. This will help reduce their operating temperature which will reduce the thermal stress on the circuit board and will help prevent premature failure of the transistors.