If your output transistors are shorted or open, then all sorts of semiconductor, resistor, and even PCB trace damage may have resulted. Amps for bar bands tend to have this happen so often that some shops add the cost of all new output transistors to the estimate even if the complaint is "a buzzing sound". Bar bands set up & tear down so much, shorted speakers or wires often occur, which shorts the output transistors. I thought from your original post that the perhaps th amp had died of old age, which is different. In that case randomly placed electrolytic caps go low resistance and blow the semiconductor driving them, or refuse to pass sound, or sound funny. Old amps can have the fan stick and then overheat the O.T's, but it is a 20 year race between the fan and the wet capacitors.
In an O.T. short event, you remove all the O.T.'s and drivers and predrivers, and check them on your transistor tester. My transistor tester is a DVM, I do the double-diode test. This can prove them bad, but doesn't prove them good. I replaced 3 groups of 5 O.T.s on the current amp I am fixing, tested the other group good, and replaced all the damaged drivers, predrivers, resistors, burnt traces, diodes, etc. Then when I turned it on another of the old 5 output transistors exploded and took out a PCB trace, emitter resistor, opposing new output transistor, and another handful of drivers etc. I replace output transistors in the groups, since I can't match Hfe or Vbe and my amp has 0.5 ohm emitter resistors which allows for some O.T. mismatch. Other amps with .3 ohm or .1 ohm emitter resistors are more sensitive to O.T. mismatch.
Some people on here are forever looking for the exact same transistors their amp had. I don't bother. I match package, Vce, power rating. The first amp I worked on had house numbered transistors that nobody made anyway because they were so slow and obsolete. New fast transistors can replace old slow ones if you add enough base to emitter capacitance to slow them down and not let them oscillate. This practice was pioneered by dynaco that made up the "TIP" mod to put newer transistors in place of the 1966 original ones. My current stock selection includes MJ21195 and MJ21194 TO3 (<$4), TIP41C and TIP42C in TO220($.40), and MPS8099 and MPSA56 in TO92($.11). My amps have collector voltage of 85 and 95, so I accentuate the Vce over people that work amps in the 50V range. The TIP41-42's blew up after the last event so I am looking for something tougher in TO220 and have found D44R2 at $.50 but havn't found a >150 V PNP TO220 that is under $2 each yet. See the 10 favorite parts for other people's mainstream (and more expensive) choices of parts to keep in stock. I've also had to replace a lot of 1n4148 diodes, some 3 amp rectifiers and 6 amp bridges, and a half dozen 15 V zeners, a bunch of emitter resistors, and a half dozen other resistors and a few capacitors. After an O.T. meltdown, pretty much every part touching them has to be checked, back into the amp far enough that you quit finding blown up stuff. Not a lot of money involved, but a lot of time.
After an OT meltdown, you should really take your dummy resistors and check the power transformer for current at rated voltage for a while to make sure it hasn't been damaged before you fix the amp.
In an O.T. short event, you remove all the O.T.'s and drivers and predrivers, and check them on your transistor tester. My transistor tester is a DVM, I do the double-diode test. This can prove them bad, but doesn't prove them good. I replaced 3 groups of 5 O.T.s on the current amp I am fixing, tested the other group good, and replaced all the damaged drivers, predrivers, resistors, burnt traces, diodes, etc. Then when I turned it on another of the old 5 output transistors exploded and took out a PCB trace, emitter resistor, opposing new output transistor, and another handful of drivers etc. I replace output transistors in the groups, since I can't match Hfe or Vbe and my amp has 0.5 ohm emitter resistors which allows for some O.T. mismatch. Other amps with .3 ohm or .1 ohm emitter resistors are more sensitive to O.T. mismatch.
Some people on here are forever looking for the exact same transistors their amp had. I don't bother. I match package, Vce, power rating. The first amp I worked on had house numbered transistors that nobody made anyway because they were so slow and obsolete. New fast transistors can replace old slow ones if you add enough base to emitter capacitance to slow them down and not let them oscillate. This practice was pioneered by dynaco that made up the "TIP" mod to put newer transistors in place of the 1966 original ones. My current stock selection includes MJ21195 and MJ21194 TO3 (<$4), TIP41C and TIP42C in TO220($.40), and MPS8099 and MPSA56 in TO92($.11). My amps have collector voltage of 85 and 95, so I accentuate the Vce over people that work amps in the 50V range. The TIP41-42's blew up after the last event so I am looking for something tougher in TO220 and have found D44R2 at $.50 but havn't found a >150 V PNP TO220 that is under $2 each yet. See the 10 favorite parts for other people's mainstream (and more expensive) choices of parts to keep in stock. I've also had to replace a lot of 1n4148 diodes, some 3 amp rectifiers and 6 amp bridges, and a half dozen 15 V zeners, a bunch of emitter resistors, and a half dozen other resistors and a few capacitors. After an O.T. meltdown, pretty much every part touching them has to be checked, back into the amp far enough that you quit finding blown up stuff. Not a lot of money involved, but a lot of time.
After an OT meltdown, you should really take your dummy resistors and check the power transformer for current at rated voltage for a while to make sure it hasn't been damaged before you fix the amp.
Last edited:
That should still be tolerable, depends on how warm the whole affair got though. Wouldn't be the first transistors killed by exceeding SOA under load.
Anyway, the classic KA-701 is by no means a "low-end" amp. While it is true that Kenwood made their share of mass-market "consumer crap" in the '80s in particular (as did Pioneer or Yamaha - so what?), they also produced a number of quality components. This particular unit was one of their larger integrated amps back in 1980, playing "third fiddle" to KA-801 and KA-907. An 80 watter (FTC rating) wasn't too shabby back then, and 13.5 kg isn't a featherweight. Only 100 W into 4 ohms indicates it's happier with 8-ohm loads though.
Circuit topology is a typical "Japanese" affair with triple differential stages (cascoded FET input stage), similar to what you'll find in a modern-day Denon PMA-1500AE. Finals are 2SC2525 + 2SA1075 (matched I suppose), drivers are 2SC1913 + 2SA913. Input FETs are 2SK150A doubles.
As for what to do next, I'd suggest a number of tests in UNPOWERED state (you know what they say about the magic smoke - once it's out it's plenty hard to put back in).
Using diode test mode, check all transistors on the heatsink (after looking up their pinout in case it's not silkscreened on the PCB). They should measure like a double diode (np-pn for npn and pn-np for pnp), not like a dead short. Note down any suspicious parts. This should allow estimating which channel has a problem.
Ultimately, all the transistors in the power amp need to be checked, plus a few diodes and passives. That would be up to CONTROL board Q1..Q12 and POWER AMP board Q1..Q16. Testing FETs isn't quite as easy, they definitely shouldn't be low resistance from gate to anywhere else though.
Once the dead power amp has been identified, all the diodes and resistors should also be checked. Resistors go up in value when damaged. In circuit, they may measure lower than they should be (due to other stuff being in parallel), but they should never measure high. Finally the dead parts also want to be replaced. Substitutes for some Qs may have to be found.
There are a number of small electrolytics around the differential stages on the control PCB, some of them could get the amp to oscillate when dried out. As mentioned, these should be replaced with quality new parts, preferably with permitted operating voltage one step higher. In addition, operation of the bias setting pots on the power amp board should be checked, maybe they can also use some cleaning. The amp should NOT be turned on until these steps are completed. As a final tweak, I'd replace the 43 ohm gain-setting resistors on the control PCB with metal film types.
Yes, you're looking at a bunch of work. Repairs in DC-coupled power amps require patience and diligence (experience doesn't hurt either), as they tend to blow up in your face and take your $$$ new parts right to semiconductor heaven otherwise. To minimize damage, either use a variac or fashion a bulb tester, which puts a lightbulb socket in series with the mains live so the bulb acts as a current limiter (the requirement for adequate insulation need not be mentioned, I hope).
Anyway, the classic KA-701 is by no means a "low-end" amp. While it is true that Kenwood made their share of mass-market "consumer crap" in the '80s in particular (as did Pioneer or Yamaha - so what?), they also produced a number of quality components. This particular unit was one of their larger integrated amps back in 1980, playing "third fiddle" to KA-801 and KA-907. An 80 watter (FTC rating) wasn't too shabby back then, and 13.5 kg isn't a featherweight. Only 100 W into 4 ohms indicates it's happier with 8-ohm loads though.
Circuit topology is a typical "Japanese" affair with triple differential stages (cascoded FET input stage), similar to what you'll find in a modern-day Denon PMA-1500AE. Finals are 2SC2525 + 2SA1075 (matched I suppose), drivers are 2SC1913 + 2SA913. Input FETs are 2SK150A doubles.
As for what to do next, I'd suggest a number of tests in UNPOWERED state (you know what they say about the magic smoke - once it's out it's plenty hard to put back in).
Using diode test mode, check all transistors on the heatsink (after looking up their pinout in case it's not silkscreened on the PCB). They should measure like a double diode (np-pn for npn and pn-np for pnp), not like a dead short. Note down any suspicious parts. This should allow estimating which channel has a problem.
Ultimately, all the transistors in the power amp need to be checked, plus a few diodes and passives. That would be up to CONTROL board Q1..Q12 and POWER AMP board Q1..Q16. Testing FETs isn't quite as easy, they definitely shouldn't be low resistance from gate to anywhere else though.
Once the dead power amp has been identified, all the diodes and resistors should also be checked. Resistors go up in value when damaged. In circuit, they may measure lower than they should be (due to other stuff being in parallel), but they should never measure high. Finally the dead parts also want to be replaced. Substitutes for some Qs may have to be found.
There are a number of small electrolytics around the differential stages on the control PCB, some of them could get the amp to oscillate when dried out. As mentioned, these should be replaced with quality new parts, preferably with permitted operating voltage one step higher. In addition, operation of the bias setting pots on the power amp board should be checked, maybe they can also use some cleaning. The amp should NOT be turned on until these steps are completed. As a final tweak, I'd replace the 43 ohm gain-setting resistors on the control PCB with metal film types.
Yes, you're looking at a bunch of work. Repairs in DC-coupled power amps require patience and diligence (experience doesn't hurt either), as they tend to blow up in your face and take your $$$ new parts right to semiconductor heaven otherwise. To minimize damage, either use a variac or fashion a bulb tester, which puts a lightbulb socket in series with the mains live so the bulb acts as a current limiter (the requirement for adequate insulation need not be mentioned, I hope).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.