Kenwood KR-9400 Biasing, Service Bulletin Mod Problem

[kanurys]

I'm working on a very nicely kept KR-9400. Cleaning up the pots/switches, checking bias, etc...

The heat sinks get pretty warm. I can keep my hand there but it's at my wimpy limit. I pulled the driver boards to see if the mods have been done. They had been, but the last point, "6) Short pin 3 to pin 2 via a jumper wire." had not been done. Before and after shorting the pin the reading across the emitter resistors remained at around 1.6mV.

The other channel reads around 13.4mV and adjusts to about 20mV max. That's better, but I'd like to get it closer to stock (somewhere between 30mV and 50mV).

The amp does pass audio at the rated power with a clean sine wave. I just really don't like the reading or how warm they're getting.

Any thoughts as to why this amp is reading like this? I'm definitely missing something and need fresh eyes. Maybe the bias pot is bad. I'll check that.

Manual: https://www.hifiengine.com/manual_library/kenwood/kr-9400.shtml

 

[hopjohn]

In my experience amps run hotter when the bias is above that recommended, not below. I would associate a lower bias reading with distortion before rated power..

The service bulletin notes that the procedure should only be done if the original power transistors have blown. Are the 4 replacements installed in both channels? (2SB539, 2SD287).

If the service bulletin modifications are employed the new bias should be at 30mV, not the original 50mV, so forget the range you mention.

If adjusting VRe1 does nothing to effect the bias, under load, with the 10Khz signal injected as prescribed. I'd suspect you may have other problems and checking the pot would be a good place to start.

 

[gort69]

Also - at some point during production Kenwood changed the amp boards to incorporate the changes called for in the service bulletin.. As I recall it's pretty obvious which board you have by comparing the foil sides on yours to the service manual and you won't see a cut mark on the trace from pin 3, but rather the trace stops short of Re7. It was my understanding from reading through pertinent threads here that no further mod are necessary to this later board. (It's been awhile)



I remember it being pretty cramped in there for connecting the mini-grabbers and with 2 pairs of outputs per channel it can get confusing. You might not be connected to the proper test points. Seems it was easier to connect to the pair accessible from the bottom.

 

[kanurys]

In my experience amps run hotter when the bias is above that recommended, not below. I would associate a lower bias reading with distortion before rated power..

The service bulletin notes that the procedure should only be done if the original power transistors have blown. Are the 4 replacements installed in both channels? (2SB539, 2SD287).

If the service bulletin modifications are employed the new bias should be at 30mV, not the original 50mV, so forget the range you mention.

If adjusting VRe1 does nothing to effect the bias, under load, with the 10Khz signal injected as prescribed. I'd suspect you may have other problems and checking the pot would be a good place to start.

Thank you for your replies. I interpreted the instructions for the modification in the same way you did - that the bias should be set to 30mV. I just saw another post on here talking about alternative biasing. Lets stick to the instructions, shall we? I'll get some proper 1K pots and swap them out to see if the bias is more in spec.

All the transistors have been changed. The mod has been done as instructed. I can see a dremel cut in the trace on pin 3 and all the caps and resistors have been installed. The jumper was not installed, as stated in the original post. I'm not sure what effect this had on the amp over time.

Is there anything else on the driver boards that tend to go bad on these amps?

 

[hopjohn]

If the trimmer doesn't seem to be the issue and you're not able to adjust try pulling and checking the bias transistor Qe5. Among the other transistors on the board Qe3 2SC1451 (CBE) is the probably most suspect in terms of reliability. I replace them on sight. and I've been using KSC3503 (ECB) as a replacement. I'd probably also replace Qe6,7 2SC1213A while I was in there as preventative maintenance, but they should have nothing to do with (ASO protection) the issues you are having. I replace those with KSD1616AYA.

 

[kanurys]

If the trimmer doesn't seem to be the issue and you're not able to adjust try pulling and checking the bias transistor Qe5. Among the other transistors on the board Qe3 2SC1451 (CBE) is the probably most suspect in terms of reliability. I replace them on sight. and I've been using KSC3503 (ECB) as a replacement. I'd probably also replace Qe6,7 2SC1213A while I was in there as preventative maintenance, but they should have nothing to do with (ASO protection) the issues you are having. I replace those with KSD1616AYA.

Awesome. I will check through the transistors and see if anything is failing. A few questions:
1) what would be a good available substitute for Qe5?
2)Qe6 and Qe7 are different types of transistors stated in the service manual as 2SA673A and 2SC1213A, respectively. What available substitutes are there for these? Should I also replace Qe8 and Qe9 since they're the same type? I suppose that would be good form.
3)I was not able to see any sort of AYA variant of KSD1616AYA in the datasheets. Is this a correct part number?

Thanks for reading through my questions.

 

[hopjohn]

Forget KSD1616AYA

2SA673A (Qe6, Qe9) replace with KSA733C and 2SC1213A (Qe7,Qe8) with KSC945C.
Make sure you use the C suffix parts here so that pinout matches the original.

Replace Qe5, 2SC1416 ONLY if it proves to be bad, then KSC1845F.

 

[kanurys]

Forget KSD1616AYA

2SA673A (Qe6, Qe9) replace with KSA733C and 2SC1213A (Qe7,Qe8) with KSC945C.
Make sure you use the C suffix parts here so that pinout matches the original.

Replace Qe5, 2SC1416 ONLY if it proves to be bad, then KSC1845F.

Ok thanks. I appreciate you rechecking those part numbers.

 

[kanurys]

Forget KSD1616AYA


Replace Qe5, 2SC1416 ONLY if it proves to be bad, then KSC1845F.

Just curious- why ONLY if Qe5 is bad, replace it? What is the harm of replacing it if I'm already replacing other transistors on the board? It's $0.30.

I can see how Qe6 through 9 can take a beating in this circuit.

 

[hopjohn]

Just curious- why ONLY if Qe5 is bad, replace it?

Because It's potentially problematic to swap it and you'd only want to do so if you HAVE to. It's not an upgrade or preventive swap. BTW if you do find Qe5 is bad you should have a dim-bulb tester (search threads for how to build one, if you do not) to use after the replacement so you can set the bias without damaging anything.

You should probably focus your efforts on Qe3 first, as they do tend to fail with regularity. Again use a KSC3503 there and watch the pinout differences.

 

[kanurys]

Because It's potentially problematic to swap it and you'd only want to do so if you HAVE to. It's not an upgrade or preventive swap. BTW if you do find Qe5 is bad you should have a dim-bulb tester (search threads for how to build one, if you do not) to use after the replacement so you can set the bias without damaging anything.

You should probably focus your efforts on Qe3 first, as they do tend to fail with regularity. Again use a KSC3503 there and watch the pinout differences.

I see. Thank you for your explanation and advice. We'll see how this thing comes out after testing.