9090DB, Went from Playing Normally to Quiet

[BearsFan]

Hey Guys, I do have one additional question for you. I had my 9090 restored about 12 years ago by a guy in PA who did a fantastic job. That unit is a gem, performance out the ying yang. When he redid the output transistors he used a greaseless shim under the outputs that resulted in a clean environment.... what are your thoughts on those devices vs the messy silicon? I've had zero problems with that unit and it has seen some serious Zeppelin and Doors over the years.

Last item, just an FYI, I am driving these amps with Klipsch KLF 20's, which are 100db efficient, so none of these receivers ever makes it near "12 o'clock" on the volume dial or past 25-30 watts on the power meters. I'm driving the inputs with a 24bit NAD transmitter, Amazon Echo and a Panasonic CD player. This is just for reference, a lot of times problems occur due to the outside influences. These speakers have been in place since 2006 and have all the Crites upgrades, so they are solid.

One last thing.... there's always one last thing, right? Shortly after I had this unit refurbed, the tuner went out. It would play for awhile then die out, both AM and FM no meter action. The tech told me to bring it in because a DC component had probably failed. I never brought it back, wasn't thrilled with the sloppy work. Well, eventually, that "DC component" failed completely and the tuner has not worked since. Any thoughts on where i might look to see if I can't resolve that issue. I'm going to take a peek at the schematics. I seem to recall he said something about a 15vdc supply, I could be off by a volt or two. Thanks again!

 

[maxhifi]

Hey Guys, I do have one additional question for you. I had my 9090 restored about 12 years ago by a guy in PA who did a fantastic job. That unit is a gem, performance out the ying yang. When he redid the output transistors he used a greaseless shim under the outputs that resulted in a clean environment.... what are your thoughts on those devices vs the messy silicon? I've had zero problems with that unit and it has seen some serious Zeppelin and Doors over the years.

Last item, just an FYI, I am driving these amps with Klipsch KLF 20's, which are 100db efficient, so none of these receivers ever makes it near "12 o'clock" on the volume dial or past 25-30 watts on the power meters. I'm driving the inputs with a 24bit NAD transmitter, Amazon Echo and a Panasonic CD player. This is just for reference, a lot of times problems occur due to the outside influences. These speakers have been in place since 2006 and have all the Crites upgrades, so they are solid.

One last thing.... there's always one last thing, right? Shortly after I had this unit refurbed, the tuner went out. It would play for awhile then die out, both AM and FM no meter action. The tech told me to bring it in because a DC component had probably failed. I never brought it back, wasn't thrilled with the sloppy work. Well, eventually, that "DC component" failed completely and the tuner has not worked since. Any thoughts on where i might look to see if I can't resolve that issue. I'm going to take a peek at the schematics. I seem to recall he said something about a 15vdc supply, I could be off by a volt or two. Thanks again!

Silicon pads are one way to isolate the transistors from the heat sinks. They are effective, but actually aren't as good heat transfer wise as the grease and mica. As for the mess, I have a tip for you. Install the transistors, then wipe off all the excess grease with paper towels and Q tips. It takes less than five minutes and you can make it look great. Excessive grease is useless, the only stuff that matters is what's under the transistor, and you won't wipe that away. This guy was just sloppy and skipped a step. I always slop a bit of grease like that when I replace transistors, but then I take the time to wipe it clean so It doesn't look like crap.

Remember, grease goes on both sides of the mica, and you need to wipe off all traces of old grease before installing new. If the mica insulator is damaged, replace it with a new one. If you hold the old mica down to a flat surface with a finger, you can wipe off the old grease without damaging the insulator.

For the missing radio, yes, you need to troubleshoot that. Start measuring power supply voltages and go from there.

 

[Skywatcher]

Hey Guys, I do have one additional question for you. I had my 9090 restored about 12 years ago by a guy in PA who did a fantastic job. That unit is a gem, performance out the ying yang. When he redid the output transistors he used a greaseless shim under the outputs that resulted in a clean environment.... what are your thoughts on those devices vs the messy silicon? I've had zero problems with that unit and it has seen some serious Zeppelin and Doors over the years.

Last item, just an FYI, I am driving these amps with Klipsch KLF 20's, which are 100db efficient, so none of these receivers ever makes it near "12 o'clock" on the volume dial or past 25-30 watts on the power meters. I'm driving the inputs with a 24bit NAD transmitter, Amazon Echo and a Panasonic CD player. This is just for reference, a lot of times problems occur due to the outside influences. These speakers have been in place since 2006 and have all the Crites upgrades, so they are solid.

One last thing.... there's always one last thing, right? Shortly after I had this unit refurbed, the tuner went out. It would play for awhile then die out, both AM and FM no meter action. The tech told me to bring it in because a DC component had probably failed. I never brought it back, wasn't thrilled with the sloppy work. Well, eventually, that "DC component" failed completely and the tuner has not worked since. Any thoughts on where i might look to see if I can't resolve that issue. I'm going to take a peek at the schematics. I seem to recall he said something about a 15vdc supply, I could be off by a volt or two. Thanks again!

Nice speakers! I personally don't care for the silicone pads. To get good ones with enough thermal transfer, you'll pay for them. The "cheapies" you get aren't up to the task. That tech must have used good ones before the chinese flooded the market with inferior pads. Plus they are a one-time deal, not reusable. I'll use mica and good thermal paste any day. The tuner we can address after we get it back up and running. The RF power supply may, or may not be the problem, but usually is in this situation. Check the solder joints on the 2 ceramic resistors, they run HOT and sometimes the solder joints crack. And the power (I think 2SD313 if memory serves me right) transistor runs hot and could be bad as well. There is one cap in front of the transistor, sitting on a plastic "spacer" (if it is original) that gets hot and dries out, plus a smaller cap right next to the heatsink that also dries out. It's a small board and cheap to rebuild and replace all the electrolytic caps. There is a pot to adjust the voltage to spec.if needed. I believe it is 12.5v.

 

[Skywatcher]

Silicon pads are one way to isolate the transistors from the heat sinks. They are effective, but actually aren't as good heat transfer wise as the grease and mica. As for the mess, I have a tip for you. Install the transistors, then wipe off all the excess grease with paper towels and Q tips. It takes less than five minutes and you can make it look great. Excessive grease is useless, the only stuff that matters is what's under the transistor, and you won't wipe that away. This guy was just sloppy and skipped a step. I always slop a bit of grease like that when I replace transistors, but then I take the time to wipe it clean so It doesn't look like crap.

Remember, grease goes on both sides of the mica, and you need to wipe off all traces of old grease before installing new. If the mica insulator is damaged, replace it with a new one. If you hold the old mica down to a flat surface with a finger, you can wipe off the old grease without damaging the insulator.

For the missing radio, yes, you need to troubleshoot that. Start measuring power supply voltages and go from there.

Agreed on the mica! I clean off all the old compound, then, using a small paint brush, apply some (not excessive, just to fill the pores in the metal) to the transistor spreading it evenly, then put the mica on the transistor, put some pressure to make full contact, then apply compound to the mica, again, just a thin layer. Then when you install the transistor, only a thin line will squeeze out around the edge making for a nice "clean" look. Too much compound and you will not only risk "bending" the tabs, but the transistor will "float" in top the compound resulting in poor thermal transfer. I've seen guys blob it on, then wonder why they have trouble.

 

[maxhifi]

Agreed on the mica! I clean off all the old compound, then, using a small paint brush, apply some (not excessive, just to fill the pores in the metal) to the transistor spreading it evenly, then put the mica on the transistor, put some pressure to make full contact, then apply compound to the mica, again, just a thin layer. Then when you install the transistor, only a thin line will squeeze out around the edge making for a nice "clean" look. Too much compound and you will not only risk "bending" the tabs, but the transistor will "float" in top the compound resulting in poor thermal transfer. I've seen guys blob it on, then wonder why they have trouble.

I put a little dab in the middle of the transistor, and then smear it around into a thin layer with my finger.... I hope nobody ever finds out this stuff is a carcinogen

Regardless of your method of application that's the thing - you want a thin even layer

 

[Skywatcher]

I put a little dab in the middle of the transistor, and then smear it around into a thin layer with my finger.... I hope nobody ever finds out this stuff is a carcinogen

Regardless of your method of application that's the thing - you want a thin even layer

I used to do the "finger spreader", but then I saw white fingerprints on things everywhere.

 

[Skywatcher]

Be aware that there are quite a few "mistakes" in the service manual, and some mistakes on the silkscreen markings on some boards. If you see something (like a resistor value/size) that doesn't jive with the service manual, STOP and verify it with one of us here before making any changes. There are quite a few resident experts here that can verify things for you. Sansui isn't the only company that has made mistakes in the documentation and/or board markings. Off the top of my head, there mistakes for the driver board, and the protection/relay board. So just be careful. If you find any resistor burned beyond recognition, ask and verify the correct component. Fusible resistors, which are a "bluish or more of a "tanish" color metal oxide, sometimes run hot enough to cause the color bands to fade.

 

[BearsFan]

Hifi, just wanted to bring you up to speed as a courtesy, Skywatcher is going to tackle the driver board, I have ordered a complete set of On Semi replacements and this project is going to go on hold until I get back from out of town on the 23rd. When i get back, I'll put things back together and step through the start up procedure. Will keep you up to date. Thanks again for hopping in, hopefully my dusty tech skills can be brought back, it's been 30 some years since I was doing Tacan and aircraft radar component level in the Navy...

 

[maxhifi]

Hifi, just wanted to bring you up to speed as a courtesy, Skywatcher is going to tackle the driver board, I have ordered a complete set of On Semi replacements and this project is going to go on hold until I get back from out of town on the 23rd. When i get back, I'll put things back together and step through the start up procedure. Will keep you up to date. Thanks again for hopping in, hopefully my dusty tech skills can be brought back, it's been 30 some years since I was doing Tacan and aircraft radar component level in the Navy...

I have no doubts your amp will be back up and running in no time! You guys are making me want a Sansui amp of my own!

 

[Skywatcher]

Hifi, just wanted to bring you up to speed as a courtesy, Skywatcher is going to tackle the driver board, I have ordered a complete set of On Semi replacements and this project is going to go on hold until I get back from out of town on the 23rd. When i get back, I'll put things back together and step through the start up procedure. Will keep you up to date. Thanks again for hopping in, hopefully my dusty tech skills can be brought back, it's been 30 some years since I was doing Tacan and aircraft radar component level in the Navy...

The board made it here today safe and sound. And still fresh!

 

[BearsFan]

Great Jeff! I am still recoering from St. Patricks Day here in Nashville. They certainly kmow how to do it up! I'm interested in what you end up finding when you dive into that board. As far as the component replacements go, I'd rather you err on the side of replacing more than less. If that means another $15-$20 of your labor, so be it. I trust your judgment and look forward to diving in when I get back. Thank you again for your help.
Dan

 

[Skywatcher]

Great Jeff! I am still recoering from St. Patricks Day here in Nashville. They certainly kmow how to do it up! I'm interested in what you end up finding when you dive into that board. As far as the component replacements go, I'd rather you err on the side of replacing more than less. If that means another $15-$20 of your labor, so be it. I trust your judgment and look forward to diving in when I get back. Thank you again for your help.
Dan

Oh, party time! lol! It is pretty much done but I am waiting on the new multi-turn pots to arrive. All the electrolytic caps are replaced with known quality (Panasonic FC, Nichicon low impedence, and 2 Meritek 0.47uf low leakage) and all are now 105°C. Whoever did this had 1uf where the .47 should have been, though it still worked. All the transistors "tested" fine (no shorted ones), but I did find 3 "iffy" diodes whos forward voltage was different than all the rest. I replaced all the diodes since I had to pull them to test them anyways. The N4148 are literally "pennies" each in bulk. All the glue is cleaned off too and anything that was corroded from the glue replaced.The VD1212 replaced as well. Here are a few photos of the progress so far.

 

[BearsFan]

It's visually 100% better than when you received it. I kinda miss the silicon fingerprints on the heat sinks though....

 

[Skywatcher]

It's visually 100% better than when you received it. I kinda miss the silicon fingerprints on the heat sinks though....

I can "reinstall" them if you like.

 

[BearsFan]

Based on what you see up til now what's your assessment on the correlation between the driver board cand the two failed outputs, or was it just "their time" due to 45 years of old tyme rock and roll"?

 

[Skywatcher]

Based on what you see up til now what's your assessment on the correlation between the driver board cand the two failed outputs, or was it just "their time" due to 45 years of old tyme rock and roll"?

Well those 3 "iffy" diodes could have played a role since 2 of them were in the input stage. I had a Yamaha recently that had 2 shorted outputs and the only thing wrong were two diodes in the input stage that were "iffy" just like yours with a higher Vf. However with yours, it might be an "effect" rather than a "cause". I haven't pulled the pots yet (waiting on multiturn) to test them. They have been known to just fail. If a BIAS pot is bad, I would say the pot would have been the cause. All 4 of your pots have an intentional "melted spot". I am not sure why that "tech" would have done so. Possibly as a "marker", or an attempt to "lock" them. . It's also possible that one of those (AHEM) speaker fuses you showed me that were cobbled in and never insulated, may have shorted out against a solder joint. You wouldn't necessarily see a "burn mark" if it had touched lead solder. Maybe there is a tiny "melt" mark in the lead under the offending fuse where you can't see it. When you get back, we'll take a look at the relay/protection board where the fuses *should* be if your revision has them. However, depending on *if* one did short out, we'll have to check out *where* it shorted in case it damaged something.But let's concentrate on first things first. And we'll have to test the 0.33 ohm 5 watt cement emitter resistors as well before even thinking of powering it on.

 

[BearsFan]

Thank you Jeff, good instruction. When I get back I'll follow your lead, nothing hasty without discussion first. This unit's a keeper so I want it to be right. Forgot about those fuses; he actually did that on all my receivers if the truth be told, I didn't question based on his years of experience. I know my 8080db has them. It will be interesting to see what you find on those pots when you pull them. IMHO just looking at them, they should have been replaced during the last board update, from my long ago experience, anything that "adjusts" has a lifespan and those critical pots fit that description.
The outputs arrived yesterday so I will have them going into phase 2 next week.

 

[Skywatcher]

Thank you Jeff, good instruction. When I get back I'll follow your lead, nothing hasty without discussion first. This unit's a keeper so I want it to be right. Forgot about those fuses; he actually did that on all my receivers if the truth be told, I didn't question based on his years of experience. I know my 8080db has them. It will be interesting to see what you find on those pots when you pull them. IMHO just looking at them, they should have been replaced during the last board update, from my long ago experience, anything that "adjusts" has a lifespan and those critical pots fit that description.
The outputs arrived yesterday so I will have them going into phase 2 next week.

The pots aren't sealed and just get dirty and oxidized over time. Anyways, here is a photo of one of the 9090db relay/protection board. We'll discuss it when the time comes.
No surprise the "butcher" treated your 8080db in the same way.

 

[BearsFan]

Hi Jeff, just a quick data point for you, I never did lose the ability for the protection circuit to go steady green, but the time to kick in was about 12 sec, post output failure. This unit was never fast to go green, the 8080db and rebuilt 9090 kick in under 5 seconds consistently. This is more an FYI than anything to help with the testing

 

[Skywatcher]

Hi Jeff, just a quick data point for you, I never did lose the ability for the protection circuit to go steady green, but the time to kick in was about 12 sec, post output failure. This unit was never fast to go green, the 8080db and rebuilt 9090 kick in under 5 seconds consistently. This is more an FYI than anything to help with the testing

That's good to know. It would seem something was causing it to take longer for the voltages/current to stabilize. It is very possible the BIAS was too high in at least the one channel. Or even the DC too high as well. We'll see how this board acts in my 9090db. Though it will not have the old pots in it for any comparison. I will take note of their current settings when I pull them, and compare after I get it tuned in with the new pots. The pots should be arriving any day now.