Kenwood KR-720 diagnostics

Sam Cogley

Sam Cogley

Last of the Time Lords
Subscriber
I bought this last week at a moving sale along with some other gear and a rack. I knew going in that it had a blown fuse, first assumption is that it's got some bad transistors. A multimeter confirmed that diagnosis, the primary output transistors on one channel were shorted.

Tonight, I desoldered the leads from those transistors and plugged the receiver into the dim-bulb variac. It lit up and everything else seems to be working OK. If I simply replace the two bad transistors with appropriate units, should I be good to go, as nothing else seems to be causing a dead short?

This is an interesting little transitional receiver. It has both a digital and analog tuning display, yet it retains the 70s silverface look.

The toasted transistors are a 2SA1102 and a 2SC2577. It looks like Mouser has the NTE37 and NTE36 cross-referenced parts in stock.
 
Register to hide this ad
Sam, Do you need the basic specs on that? I found the page from a brochure on that year's Kenwood receivers online, I forgot where I found the page, but I copied the image, since it also has the KR-750 which I have. Does yours also have that strange automatic tuning mechanism?

If you want the that spec sheet, send me a private message with your email address in it, okay?
 
The KR-720 is, put simply, a "bad receiver." This is due to an error in the design. The channel that shorted is very likely the one where the drivers and outputs are separated from the pre-drivers by a long multi-conductor cable.

The engineers who designed the piece did not include the reactance of the multi-conductor cable in their design equations. This reactance causes a phase shift in the transfer of the signal that is not taken into account in the frequency compensation in the negative feedback loop. As a result, the amplifier is only marginally stable, and is inclined to oscillations somewhere between 500kHz and 1mKz.

http://en.wikipedia.org/wiki/Bode_plot

This condition is further complicated by the fact that, over the decades component values drift, further rendering the phase-compensation in the feedback loop incorrect.

It is possible to add reactance to the feedback loop to "null out" the untoward effects of the multi-conductor cable and lumped component drift, but this requires some experimentation.

* * * * *

Separate from the above, I highly recommend that you NOT use NTE devices. Wherever possible, use the original OEM parts. They are still available. I only use NTE parts, when the originals or close substitutes are no longer available in the secondary market. I will sometimes use NTE devices for troubleshooting purposes, but will remove them and install the correct parts later on.

Fred
 
I haven't been able to find the SK transistors anywhere - if there is a source where I can readily order them, I'd appreciate the info.

Is there anywhere I can read up on what I need to do to dampen the oscillation - what circuits I need to be messing with and how much resistance is needed (in general)? I've got a service manual coming to me, so I should have schematics shortly. Something that was certainly not helping matters was the notorious glue that turns black and becomes conductive. Two blobs of it were holding a disc capacitor and a resistor to the board on top of traces for the dead channel. I popped the components loose and scraped that junk off.

The operational channel does sound good, I'd like to get the little oddball running properly again.

The engineers who designed the piece did not include the reactance of the multi-conductor cable in their design equations. This reactance causes a phase shift in the transfer of the signal that is not taken into account in the frequency compensation in the negative feedback loop. As a result, the amplifier is only marginally stable, and is inclined to oscillations somewhere between 500kHz and 1mKz.

http://en.wikipedia.org/wiki/Bode_plot
Click to expand...

That math was way over my head. I'm a lawyer, not Stephen Hawking. :D
 
These were wide bandwidth designs, and did sound pretty good. I have one in my back room museum. You can use any set of transistors in that family, for instance 2SA1106 2SC2579 and have a beefier output. I have a pair of these in stock. I'll bet drivers are bad, as Fred points out.
 
Which transistors are the drivers? I know the output transistors on the right channel (left side of the case, by the transformer) are dead shorts. It seems that if I had any more shorted semiconductors that it shouldn't power up even with the outputs out of the circuit.

I still haven't actually figured out of the dead channel is the one with the cable in the middle of it. There are a lot of ribbon cables running around in that case.

If replace the dead parts with the 2SA1106/2SC2579 combo, shouldn't I do both channels?

Mouser doesn't have a listing for the 2SA1106, and the only thing that comes up for the 2SC2579 is the NTE36. DigiKey comes up with nothing for either number.

Found someone on the auction site selling 2 2SA1106 and 2 2SC2581 transistors as a 4-piece set. What's the difference between the 79 and the 81 (besides 2...)? The same seller in Hong Kong has pairs of the 2SA1106 and 2SC2579, that looks like it might be the way to go.
 
Last edited:
Fred Longworth said:
The KR-720 is, put simply, a "bad receiver." This is due to an error in the design. The channel that shorted is very likely the one where the drivers and outputs are separated from the pre-drivers by a long multi-conductor cable.

The engineers who designed the piece did not include the reactance of the multi-conductor cable in their design equations. This reactance causes a phase shift in the transfer of the signal that is not taken into account in the frequency compensation in the negative feedback loop. As a result, the amplifier is only marginally stable, and is inclined to oscillations somewhere between 500kHz and 1mKz.

http://en.wikipedia.org/wiki/Bode_plot

This condition is further complicated by the fact that, over the decades component values drift, further rendering the phase-compensation in the feedback loop incorrect.

It is possible to add reactance to the feedback loop to "null out" the untoward effects of the multi-conductor cable and lumped component drift, but this requires some experimentation.

* * * * *

Separate from the above, I highly recommend that you NOT use NTE devices. Wherever possible, use the original OEM parts. They are still available. I only use NTE parts, when the originals or close substitutes are no longer available in the secondary market. I will sometimes use NTE devices for troubleshooting purposes, but will remove them and install the correct parts later on.

Fred
Click to expand...

The service manual came in today (I have to give Kenwood much credit, this is one of the best audio service manuals I've ever seen), and I've been poring over the schematics for the past 45 minutes. I've come to one, inescapable conclusion - the channel with the shorted outputs isn't the channel with the long cable stuck in the middle of it. Something other than oscillation induced by bad design ate those transistors, apparently. It also doesn't appear that anything other than the outputs let go, as the receiver powers up just fine with them out of the circuit. I would assume that any other shorted transistors would continue to register as a fuse-blowing short even with the outputs removed.
 
Last edited:
You are VERY lucky. As I recall the channel with the oscillation problem is located on the right side of the machine, as viewed from the front.

The correct transistors are available from places like MCM, Parts Express, B&D Electronics, and Electronica in Denver. I prefer B&D and Electronica, if only because they make an extra effort not to buy and sell counterfeit transistors. (Counterfeit transistors are becoming more and more common these days.)

In a typical output transistor job, I will replace the outputs and drivers, and often the thermal feedback device (diode array or transistor mounted on heatsink).

Check the emitter resistors.

Fred
 
That's it. The pre-out circuitry is all bundled up on the left front of the unit, then a cable runs across to the driver and output transistors on the right side. I don't have a clue what popped the left channel, the guy I bought it from said the fuse just blew. Once the bad parts are replaced, I'll make sure I test the DC offset.

I didn't find the transistors on the PE site, I'll try the others you mentioned. (edit: B&D's site isn't working, and Electronica's online catalog doesn't show either of those transistors. I'll give them a call tomorrow.)

Thanks for the parts replacement advice. In other repair news, my JR-S100 still has a short even after replacing the outputs, I'll track down the drivers and thermal feedback device to see if one of them is causing the problem.
 
Last edited:
Everyone talks about the oscillation issues on these units, but my father bought one (actually, a KR-710, but close enough) new in 1981 and used it every day for 25 years without any problems (driving somewhat inefficient speakers in a big room with a CD player sitting on top of it, no less). I had it serviced for good measure, and aside from a few bad caps in the phono stage, the tech said it tested just fine.

Incidentally, if you can get it fixed for minimal cost, you may be surprised by the sound. Clean, fast, and punchy within its limited power, and a decent tuner.
 
It does sound good on the operational channel. I'm going to get the components (as well as a complete cap set while I'm in there) and get it running again. I have no idea why the "good" channel is the one that fried, but everything seems to check out OK except the output transistors.
 
Another Kenny lives again!

I just finished replacing the driver and output transistors on the dead channel (I didn't order the emitter resistors yet). Success! The little oddball is singing away through my Minimus 7s. I should order new emitter resistors for that side (and probably another complete set of the obsolete output transistors in case of future trouble), along with a complete recap...but for now, all is well. I think this is going to become my desk unit at work if I can find some relatively small speakers to go with it. I had a line on a pair of Paradigm Atoms locally, but the eBay sale went far higher than I wanted to pay.
 
New issue - the left channel (the one I repaired) has a noticeably lower volume level than the right. I'm going to take it home tonight to check the DC offset, is there anything else I should look for?

I also just realized that I've never DeoxITed the controls on this unit. That will be done tonight.

Update: got the DC offset under 10mV on each side (the repaired channel was running around 74mV). The volume is now equal on each side, but at extremely low volumes, the left channel takes a tiny bit more knob movement to kick in.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom