Throwing in the towel, need help with bias issue on a Sansui AU-719

[saabracer23]

I have given up on trying to figure this out and hope to get some fresh eyes on this. A little back story.

I picked up an AU-719 that made a popping noise through the left speaker right after power up and it would pop a few times and then be done. The pop could be pretty loud and sometimes quiet regardless of where the volume was set. Got some suggestions from you all and this is the progress I made.

Fully recapped the power supply/protection board, replaced all of the diodes with the exception of the zeners on the ps/protect board. Transistors I replaced on that board are

TR606: was 2sd356, now MJE15032G
TR613: was 2sb526, now MJE15033G
TR601-605: was 2sc945, now ksc1845FTA
TR607: was 2sa726, now ksa992
TR608-609: was 2sc1313, now ksc1845FTA
TR610-611: was 2sa726, now ksa992
TR612: was 2sc1313, now ksc1845FTA

All of the other 2sa726 and 2sc1313 transistors in the amp were replaced. All the the 22pF “black flag” caps were replaced with new ceramics.

After this work was done I powered up the amp and it helped some, but there was still occasionally popping. Not as much, but still some.
I put it on the scope and through many power cycles found the popping to be present at the output of the driver board (pin 13), but could not detect it at the output of the preamp section going to the driver board. So to me and others I asked, this isolated it to the driver board. Someone suggested the outputs or the drivers. I decided to replace all of the diodes on the driver boards with the exception of the zeners, including the MV103s which I used 3 1N4148s in series, and the MV12s in which I used two in series. I also fully recapped both driver boards.

I pulled the outputs and using the Peak Atlas DCA75 I did measure leakage current on one of them. So I replaced all of the outputs and all of the drivers.

Outputs:
TR701-702: was 2sc2581, now MJL21194
TR703-704: was 2sa1106, now MJL21193
These are bigger packages, but they fit just fine.

Drivers:
TR15: was 2sd382, now MJE15032G
TR16: was 2sb537, now MJE15033G

After this work, still had the freaking popping. By this time I was just done with it. I decided to replace every transistor on both driver boards. As I pulled them I measured them and did find a couple with leakage. Also when brought up on the dbt it would make my bulb pulse with light and sometimes go into protection. I read on another thread someone had this EXACT SAME problem with their AU-919. It ended up being their bias trimmer that caused the issue. So out came the 470 ohm trimmers and in went 500 ohm 25 turn Bourne’s.

Transistors replaced
TR01: was 2sc2071, now 2sc3503
TR02-05: was 2sc1845, now ksc1845tfa
TR06-09: was 2sa992, now ksa992
TR10: was 2sa939, now 2sa1381
TR11: was 2sc945, now ksc1845fta
TR12-13: was 2sc2071, now 2sc3503
TR14: was 2sa939, now 2sa1381
TR15-16: mentioned above, drivers
TR17: was 2sc1845, now ksa1845fta

After doing all this work including replacing the bias trimmers the dbt no longer pulses and the amp comes out of protect.

I am no longer able to measure bias voltage! I measure no voltage at the test points and if I turn the bias trimmers it’ll throw the amp into protect.

When the amp is not I protect I measure a very small amount of dc at the outputs, less than 10 mV.

I’ve gone over all of my work multiple times, all solder joints are good. All devices were put in the correct way (checked that like 20 times lol), I measured each and every resistor.

Looking at the schematic I see there is supposed to be input voltage from the rails and I do measure and equal + and - voltage, it comes straight from the filter caps into the driver boards (pins 10 and 12, pin 11 being ground). Only other pins would be pin 14 which goes to the protection circuit. Should I be measuring anything there? I get no measurements there at all.

Thoughts? I’m desperate!!!

Thank you all
Dan

 

[gort69]

I had the identical problem - same pulsing on DBT - but on a different amp (Kenwood 600). The drivers were 2sb536/2sd381 and I used the same replacements as you did. Same ON-Semi outputs too. The problem was the MJE15032G/15033G drivers. I recommend reinstalling your original drivers and see what you get.

 

[mbz]

Apologies, I'm a measurements guy, ie, measure, analyse results,,,
Regarding the bias issue. Set trimmer so it does not go into protect.
Suggest measure voltages at transistor base to GND/Chassis then repeat
for emitter. Please don't probe transistor legs, measure at adjoining
part.

Measure at the following
TR701base (pin 5)
TR701emitter (pin 6)
TR703b (pin 8)
TR703e (pin 7)
TR15b (R38)
TR16b (R41)
TR11b (R30)
TR11e (VR02wiper)
TR11c (R56)

Guiding/ballpark voltages are shown on the right channel.

 

[mbz]

Maybe before taking above voltage measurements, try adjusting VR02 so that TR11 collector and emitter voltages are the same but
opposite sign, ie sm has them at +/- 1.77Vdc. Get as close to same values eg, 1.5Vdc as you can then take above measurements.
Also note all measurements to GND.

 

[saabracer23]

I had the identical problem - same pulsing on DBT - but on a different amp (Kenwood 600). The drivers were 2sb536/2sd381 and I used the same replacements as you did. Same ON-Semi outputs too. The problem was the MJE15032G/15033G drivers. I recommend reinstalling your original drivers and see what you get.

Okay, thank you. What would cause those to not be good candidates for replacements?

Apologies, I'm a measurements guy, ie, measure, analyse results,,,
Regarding the bias issue. Set trimmer so it does not go into protect.
Suggest measure voltages at transistor base to GND/Chassis then repeat
for emitter. Please don't probe transistor legs, measure at adjoining
part.

Measure at the following
TR701base (pin 5)
TR701emitter (pin 6)
TR703b (pin 8)
TR703e (pin 7)
TR15b (R38)
TR16b (R41)
TR11b (R30)
TR11e (VR02wiper)
TR11c (R56)

Guiding/ballpark voltages are shown on the right channel.

Maybe before taking above voltage measurements, try adjusting VR02 so that TR11 collector and emitter voltages are the same but
opposite sign, ie sm has them at +/- 1.77Vdc. Get as close to same values eg, 1.5Vdc as you can then take above measurements.
Also note all measurements to GND.

Okay, I’ll get to measuring these, or at least try to. What would you suggest is the best method with tight quarters like this? And when you say adjoining part, you mean like the leg of the resistor or capacitor the leg of that transistor goes to? I have a bit more room if I try to measure from the top side of the board and navigate with the schematic.




—— NEW DEVELOPMENT ——-

I put the old drivers back into the boards and powered it up, still got no bias voltage showing up on my meters. I have two meters set up, one on each channel. I turn the bias trimmers all the way to one extreme, power it up and the bulb is dim. No bias voltage. The amp came out of protect for a split second and then went back into protect. I turn off the power and turn the trimmers to the other extreme. Power it up and the bulb is bright. All of a sudden after a few seconds I see bias voltage come up on the right channel, nearly 200 mV! I power it off and turn the bias down a bit. Power it up again and I have bias on the right channel, around 100mV and I turn it down to about 30mV.

The left channel is still showing no bias at this point so I turn it down about a 1/4 turn and then I have bias voltage showing up. It’s jumping all over the place, 111mV, then 22mV, then 60mV, then 13mV, then 47mV, etc. These aren’t actual numbers, just examples. The numbers were moving too quickly to write them down. While this is happening the bulb is going dim, bright, dim, bright, dim, bright. The amp comes out of protection and the bias then settles. I get both channels to about 30mV, but the bulb in the dbt seemed too bright.




I powered it down and when I powered it up again same thing. Right channel was stable, left was bouncing all over the place. Amp came out of protect and about 3 seconds later went back into protect.

Dan

 

[mbz]

Okay, I’ll get to measuring these, or at least try to. What would you suggest is the best method with tight quarters like this? And when you say adjoining part, you mean like the leg of the resistor or capacitor the leg of that transistor goes to

Fully understand that things are tight, seen too many posts where the probe slips and #$%@! The point is look for an alternative.
Yes, I mean a part connected to the same point, I've put examples in brackets above. Using minnigrabbers to a connection board
is probably the best way to take multiple measurements.

—— NEW DEVELOPMENT ——-

This is good news, means the amp responds to trimmer adjustment.
Almost sounds like a part is going intermittent eg, R45,46 or D03, or there is a bad solder joint/connection.
Still go ahead with the voltage measurement, maybe start with TR701, 703, it may be enough.

 

[mbz]

maybe need a bigger bulb

Yeah, this thought had also crossed my mind...
what size bulb are you using, if 30-50W, try with 100W

 

[saabracer23]

Yeah, this thought had also crossed my mind...
what size bulb are you using, if 30-50W, try with 100W

Currently there is a 60w bulb in it. Next size up I have is 150. I guess they stopped making 100w bulbs.

Dan

 

[bberkom]

I’ve had similar issues with a cracked trace on driver boards, and that was my first thought. I also think the suggestion of a bigger bulb is a good one. Intermittent problems can be hard to find.

 

[saabracer23]

Yeah, this thought had also crossed my mind...
what size bulb are you using, if 30-50W, try with 100W

I’ve had similar issues with a cracked trace on driver boards, and that was my first thought. I also think the suggestion of a bigger bulb is a good one. Intermittent problems can be hard to find.

Well I’m not sure which did it, but what both of you had said was really weighing on me. I switched over to my next largest bulb which is 150w and I also desoldered and resoldered every single joint on each driver board. Put everything back in, crossed my fingers and flipped the power switch. Bias voltage was immediately stable, no fluctuations whatsoever. Bulb went dim and once the relay clicked and the amp came out of protection it has stayed out of protection. Thank you very much!

Now after I get bias set I’ll have to let it cool and then power it up with a speaker connected to see if that dreaded pop is gone.

Would too small of a bulb in the dbt caused those issues?



Dan

 

[dr*audio]

Sometimes the bias test points become oxidized and you get no reading. Always check to make sure you have a good connection by turning the amp off and measuring resistance between the bias points. Should be less than 1 Ohm because you are reading across the emitter resistors. If it reads high, scrape the test points until you get a good connection.
The fact that your DBT started glowing when you turned the bias up indicates that there was bias current, you just weren't able to measure it.

 

[Blue Shadow]

Currently there is a 60w bulb in it. Next size up I have is 150. I guess they stopped making 100w bulbs.

Dan

Yes, not hat the 75 and up incandescent bulbs are no longer made one needs to think about grabbing a box of bulbs removed from service that show up on CL occassionally. Should be some 100s in there, enough to last a lifetime. I think you can still get rough service bulbs in higher wattages.

 

[Blue Shadow]

I hope the 719 I haven't figured out yet is as easy as redoing all the solder connections. I doubt it will be, though.

 

[saabracer23]

Sometimes the bias test points become oxidized and you get no reading. Always check to make sure you have a good connection by turning the amp off and measuring resistance between the bias points. Should be less than 1 Ohm because you are reading across the emitter resistors. If it reads high, scrape the test points until you get a good connection.
The fact that your DBT started glowing when you turned the bias up indicates that there was bias current, you just weren't able to measure it.

Well I had soldered leads off of the back of the test points and the pins are too close together for my liking. Oddly enough when I was removing my DMM leads one of the wires popped of. The solder joint was good, but the wire must have bent back an forth too much because it snapped off.

Yes, not hat the 75 and up incandescent bulbs are no longer made one needs to think about grabbing a box of bulbs removed from service that show up on CL occassionally. Should be some 100s in there, enough to last a lifetime. I think you can still get rough service bulbs in higher wattages.

I just jumped onto amazon and of course pretty much all “100w” bulbs were either power saving 72w or LED (even though I specifically typed in incandescent). They had one listing for 4 100w rough service bulbs for $9.99. They’re frosted though, bummer. I much prefer clear bulbs.

I hope the 719 I haven't figured out yet is as easy as redoing all the solder connections. I doubt it will be, though.

What’s wrong with your 719?

Good new though! I’ve power cycled it 4 times so far this morning and have yet to hear a single pop coming through the speaker. Yay so far!

Dan

 

[Blue Shadow]

What’s wrong with your 719?

One channel is way low on output. Been a few years since it was on the bench but it was there for at least three extended sessions. Might need another one since I've learned more. Swapped parts, traced signals and other such things but haven't found the before equal sign\al, after low signal, yet. Preamp seems to take a db or two but not the 10dB that seem to be missing. I need to get numbers to see what that drop it. I have a thread or two in Sansui on it. Maybe after the Kenwood amps get finished this amp will hit the bench because it will make a smaller dent than the ends I've been sent to make a power cord for Mike. Each end is over half a pound. One IEC female, one NEMA 15P, 8.4 and 8.6 ounces. That will be a dangerous power cord when completed.

 

[saabracer23]

Holy Crap! That’ll be a beast of a cord! Well good luck with your 719. I have a Kenwood 7002 that is behaving the same way. Several hours on the bench and it too has a channel with much lower output than the other. Oddly when I separated the preoit from the main in both the preamp section and the power amp section have the same symptom.

Dan

 

[elnaldo]

(you can always parallel 2 x 50W bulbs)