Sansui 9090db Protection

[markallen]

Another thread on a 9090db that won't come out of protection.

Got this unit with failed L channel. Easy to see some of the problem on F-2624 board, as a number of components clearly smoked. Replaced burned driver board components and tested/replaced others described here on AK as usual suspects. Checked DC & bias VRs off board, all seem to range as they're supposed to, though I plan to replace them. Also tested main outputs, some also failed, so replaced all with MJ21193G/94G. Won't come out of protection, red LED flashes steadily. I'm using a DBT, and the bulb glows, but not brightly.

1. Can't set DC at speaker terminals, so checked bias. Still using the DBT, F4 = 43v, F5 = 84v. Bias trimmers don't seem to affect readings. While nothing was obviously smoking/burning, some driver board components got hot enough during this short time so you could smell them. Wondering if I now have more failed component(s)!

2. After reviewing another thread, I unplugged the driver board and the unit comes out of protection. Does this definitively isolate the problem to the driver board? I've resisted the temptation to plug in the F-2624 from a working 9090db I have to see if the non-working unit comes out of protection; I suppose the wisdom of such an action depends on the answer to my question of the previous sentence.

Also, likely unrelated: I've been reading about R13/14 on the F-2657 board. Per the SM, these are supposed to be 4.7 ohm 1/2w fusible resistors. On both the units I have, R13/14 are 10 ohm 3w ceramics. Are these mislabeled in the SM? Seems they might be confused with R25/26, which are listed as 10 ohm MR (not sure what type that is).

Would really welcome help on next steps to take, thanks!

 

[DougBrewster]

When checking bias you're checking current, not voltage. Your meter should be set up to measure mA and set it using the trimmer to 30mA. You can check offset when in protection by measuring at the blue and grey leads under the power supply board but I forget which channel is blue and which is grey.. What wattage lamp are you using in the DBT? Should be 60W

 

[markallen]

When checking bias you're checking current, not voltage. Your meter should be set up to measure mA and set it using the trimmer to 30mA. You can check offset when in protection by measuring at the blue and grey leads under the power supply board but I forget which channel is blue and which is grey.. What wattage lamp are you using in the DBT? Should be 60W

Wow. Really missed that important detail on the bias measurement. Thanks - I'd try it again, but I'm wondering about the hot components I mentioned. I've been checking components on the "bad" driver board, comparing against the same components on the known good board, and I've got some differences. Going to have to start pulling components for off-board measurements. So many components on that small board it's kind of intimidating!

DBT: Was using a 100w bulb, but will change it out. Any thoughts on the other items I mentioned?

 

[markallen]

Follow up: Changed out the DBT to a 60w and properly checked the DC & bias per Doug's comments. Result was that the R channel came into adjustment. Alternately, the L bias won't adjust at all, with 26v DC at the power supply board (grey wire). At least the problem appears to be exclusive to the L channel.

Still would like to know if I'm endangering a known good F-2624 driver board by plugging it in to the unit with the failed board channel. Any other advice on solving this welcome!

 

[LBPete]

Yes, you would endanger a good driver board if you plug it in to this unit. Don't do it. Also, don't sweat the resistors on the protection board. Sansui made many revisions to that board during production. If you have 26v on the gray wire, you have work to do. If there were burned component on the driver board you need to start by checking the output transistors for that channel.

- Pete

 

[markallen]

Yes, you would endanger a good driver board if you plug it in to this unit. Don't do it. Also, don't sweat the resistors on the protection board. Sansui made many revisions to that board during production. If you have 26v on the gray wire, you have work to do. If there were burned component on the driver board you need to start by checking the output transistors for that channel.
- Pete

Thanks for that confirmation - so I shouldn't assume the problem is exclusively in L channel of F-2426? Quite sure the outputs are all good. I'd replaced them all (see first entry, above), and then checked each one individually off the unit again when I re-opened this project a couple days ago. Though I doubt my testing while using the 100w DBT would have damaged them, I will re-check them just in case. I'm thinking I'll have to go through the driver board component by component . . .

 

[LBPete]

Don't assume your new outputs are good if that was all you replaced. You need to determine what fried the output transistors in the first place. A fault on the driver board could damage the new outputs. In addition to rechecking the outputs in the bad channel, check the driver transistors mounted on the driver board heat sinks. After that, check TR15 and TR17. It's also good idea to check the emitter resistors too. They are the white ceramic resistors lined up on the power supply board.

Also be sure to replace the two VD1212 "red dot" diodes. Replace each one with a pair of 1N4148 diodes in series.

- Pete

 

[markallen]

Don't assume your new outputs are good if that was all you replaced. You need to determine what fried the output transistors in the first place. A fault on the driver board could damage the new outputs. In addition to rechecking the outputs in the bad channel, check the driver transistors mounted on the driver board heat sinks. After that, check TR15 and TR17. It's also good idea to check the emitter resistors too. They are the white ceramic resistors lined up on the power supply board.

Also be sure to replace the two VD1212 "red dot" diodes. Replace each one with a pair of 1N4148 diodes in series. - Pete

Finally able to get back to this receiver. Checked the VD1212 diodes, and they measure good, but will replace when I get some 1N4148. All eight emitter resistors measure the same, must assume they're good.

However, identified failed TR03 (2SC1400) on affected channel - somehow missed checking this one before when I first discovered the burned components. Replaced it with a C1845. Plugged everything back in, halfway expected it to come out of protection - nope! Checking voltage at gray wire on protection board, got it to set down to zero. Next tried to set bias current, but wouldn't display in mA range. Using Amp range, turned it down but it gets to a point where it won't drop any further despite more available turns (the bias trimpots are new Bourne multi-turn) so it stops at .006.

Feel I'm going in right direction, but at a standstill. 1) Is bias adjustment problem on the F-2656 PS board? 2) If somewhere below "dangerous" voltage & bias measurements, would the unit be prevented from coming out of protection due to power being run through the 60w DBT?

 

[DougBrewster]

Next tried to set bias current, but wouldn't display in mA range. Using Amp range, turned it down but it gets to a point where it won't drop any further despite more available turns (the bias trimpots are new Bourne multi-turn) so it stops at .006.

0.006A is 6mA It would be a good idea to fix the blown fuse in the mA range of your meter..

 

[markallen]

0.006A is 6mA It would be a good idea to fix the blown fuse in the mA range of your meter..

Thanks - great catch! Blown internal fuse - that's a first on this meter. Never would have thought of that - odd to me that the only function protected by a separate fuse is this one (there is another larger fuse I assume protects the whole meter). Oh well, live and learn.

So, now both channels are adjusted - but still not coming out of protection. DBT bulb leaning towards dim, but still lit. Protection board problem?

 

[DougBrewster]

so if both channels are adjusted for bias and offset to the zero and 30mA respectively go to 100W on the DBT and check if it comes out of protection but watch for the bias going higher.

 

[markallen]

so if both channels are adjusted for bias and offset to the zero and 30mA respectively go to 100W on the DBT and check if it comes out of protection but watch for the bias going higher.

Doesn't come out of protection - but the red protection LED quits flashing, goes steady. Didn't check bias as I wasn't sure what that meant, shut it down. Also, when the LED goes steady red, the transformer seems to generate a slight buzz.

 

[DougBrewster]

the transformer buzz tells you it's supplying excess current. Did you re-check the output transistors as LBPete advised in post #7?

 

[markallen]

the transformer buzz tells you it's supplying excess current. Did you re-check the output transistors as LBPete advised in post #7?

Well, I can't explain it, but I went out to move snow, and when I came back in and tried it, the receiver came out of protection. Is it necessary to recheck the amp settings from when it was still connected to the 60w DBT?

 

[DougBrewster]

Yes.. Go back to the 60W lamp in the DBT and set the bias low, about 20mA and check the offset again. If it is stable over time, say 30 minutes then go on to the 100W lamp again and set the bias back to 20mA and recheck the offset adjusting as necessary. Observe again for stability over time. It will drift up and down somewhat.

 

[markallen]

Yes.. Go back to the 60W lamp in the DBT and set the bias low, about 20mA and check the offset again. If it is stable over time, say 30 minutes then go on to the 100W lamp again and set the bias back to 20mA and recheck the offset adjusting as necessary. Observe again for stability over time. It will drift up and down somewhat.

Thanks Doug & LBPete. Followed instructions, then graduated to full power for final adjustment, and unit seems to work as designed. It's evident, though, that the controls needs cleaning. Contingent on running the unit for awhile, I'm cautiously optimistic we have success. Unless you have further advice (always welcome!), will post final later.

 

[LBPete]

Glad to hear it worked out. As a side note, only the ammeter circuits of a multimeter have fuses. Better meters will have separate fuses for the mA and A scales. If you have a Fluke meter replacing the fuses will set you back $10 or more.

Have you posted the serial number?

- Pete