Power Amp Board Smoked

[BobHelms]

I completed recap of this SX-3800. Did the DC balance procedure with no problems. Set first idle current on left channel without any problems and was in the process of setting the first right channel idle current and it went poof, lots of smoke. What is the best way to recover from here? TIA for any help.
Bob Helms

 

[Ths364]

Build a dim-bulb tester (there is a thread here on AK about how-to- it's in a "sticky" on one of the forums). It is a current limiter that will allow you to check an amp without the "poof" and also to prevent the next "poof" in your situation. You probably toasted the outputs for that channel.

 

[BobHelms]

I have a DBT it will not bring the SX-3800 power supply assembly up to full voltage even with a 100 watt bulb. Can you set idle current with low voltages from the power supply assembly? Thanks.
Bob Helms

 

[zebulon1]

Did you replace the trimmers?
It may of opened on you.
Check the trimmer off the board.
Everything on that channel needs to be checked.
You won't get the Idle to adjust to specs on the DBT but you can see if it reacts to your trimmer movement.

 

[BobHelms]

No, I not sure what to replace them with. They were not on the recap list.

I have 'EW 6-way' tested the finals, they are gone. I tested them before I started the recap they were OK then.

 

[zebulon1]

I feel for you, been there. :yes:
Nothing that can't be fixed! :thmbsp:
I'll start looking for replacement trimmers.

 

[KingBubba]

The SX-3800 is an absolute bear to work on. I have to give techs, like markthefixer, all the kudos in the world for working on these damn things. I tried to deoxit the switches on mine and somehow ended up killing it. The process of dealing with the switches alone took me two weeks and I have no idea what I did to cause the implosion. The series of linkages connecting the buttons to the switches in the back of the receiver are a product of hell itself. I swore one week in that I would never work on that receiver ever again. I haven't. It sits in the garage collecting dust. I'd probably still be listening to it, if I had just left it alone. I wish you all the luck in the world.

 

[glen65]

I completed recap of this SX-3800. Did the DC balance procedure with no problems. Set first idle current on left channel without any problems and was in the process of setting the first right channel idle current and it went poof, lots of smoke. What is the best way to recover from here? TIA for any help.
Bob Helms

Start first by accessing the damage,
Don't make assumptions, thinking about what you already have replaced.
Start at the outputs and work your way back checking components, making sure none get missed. Check transistors, resistors (especially the emitter power resistors), then diodes and ect..
It would be wise to replace the adjustment pots, but don't worry about that until you have first accessed the current damage.

 

[Ths364]

The DBT allows you to check for dead spots in the adjustment pots with lowered risk of smoking the outputs. IF you can adjust them while on the DBT without incident, turn them back to zero and then power up the amp.

 

[Watthour]

Did you replace the trimmers?
It may of opened on you.

Lacking further information, that's probably the best guess. When that happens it just starts the suicidal race between the outputs.

I hope your outputs are still available from somewhere like B&D. If not, you'll have to get creative.

I suspect you'll find other carnage as well, due to the runaway.

 

[glen65]

Currently available MT-200 case Sanken's are a replacement.

2SA1075 use Sanken 2SA1216
2SC2525 Use Sanken 2SC2922

 

[BobHelms]

Here is what I know for sure;
The picture shows R162 & R164 fried. C122 is right in the middle of those 2 so I would guess it's toast also. Maybe mtf can provide a few comments here.
Q1
E+ C- OL
E+ B- OL
C+ E- OL
B+ E- .537
C+ B- OL
B+ C- .522
Q3
E+ C- OL
E+ B- .562
C+ E- OL
B+ E- OL
C+ B- .558
B+ C- OL
Q2
E+ C- OL
E+ B- OL
C+ E- OL
B+ E- OL
C+ B- .012 Dead Short
B+ C- .013 Dead Short
Q4
E+ C- .000 Dead Short
E+ B- .001 Dead Short
C+ E- .000 Dead Short
B+ E- .001 Dead Short
C+ B- .001 Dead Short
B+ C- .001 Dead Short
What's the best way to begin isolating all the other damaged components? IMO If I have to unsolder every thing to test it then I might as well replace it. For sure I want to replace all the trimmers, with multi-turn versions. The OEM single turn ones I've used so far are PIA. Can someone please recommend which way to go on the trimmers. Thanks.
Bob Helms

 

[glen65]

What's the best way to begin isolating all the other damaged components? IMO If I have to unsolder every thing to test it then I might as well replace it. For sure I want to replace all the trimmers, with multi-turn versions. The OEM single turn ones I've used so far are PIA. Can someone please recommend which way to go on the trimmers. Thanks.
Bob Helms

Who said you had to unsolder everything?
Most of the time you can take readings in circuit. If you do get an abnormal reading from a component, only then might it be necessary to pull a leg out of circuit.
It only takes one missed component and you can smoke your work all over again. As for your readings....

Q1
E+ C- OL
E+ B- OL
C+ E- OL
B+ E- .537
C+ B- OL
B+ C- .522
Q3
E+ C- OL
E+ B- .562
C+ E- OL
B+ E- OL
C+ B- .558
B+ C- OL

Both good

Q2
E+ C- OL
E+ B- OL
C+ E- OL
B+ E- OL
C+ B- .012 Dead Short
B+ C- .013 Dead Short
Q4
E+ C- .000 Dead Short
E+ B- .001 Dead Short
C+ E- .000 Dead Short
B+ E- .001 Dead Short
C+ B- .001 Dead Short
B+ C- .001 Dead Short

Both junk,

Also if Q126 and 128 (the drivers) aren't bad they will have taken a hit. Replace them as well.
Below is a schematic of the area where the bad outputs Q2 and Q4 are located. Check Q130 and the other transistors
as well. Make sure you don't miss any diodes, resistors or caps, ect.. Remember, it only takes one.

The OEM single turn ones I've used so far are PIA. Can someone please recommend which way to go on the trimmers. Thanks.

You're not having trouble because they are single turn. The OEM pots are simply junk. Bourns 3386 series are good.
http://www.digikey.com/catalog/en/partgroup/trimpot-3386-series/5834

 

[BobHelms]

Glen65,
I've tested 'in circuit' all resistors on the right channel. Other than R164 I didn't find any that showed abnormal readings. Even R162 was OK even though it was burnt from all the residual glue on and around it. I will replace R162 & R164. I did remove Q126 & Q128 and did the EW 6 way test.
Q126
B+ C- .604
B+ E- .602
C+ E- OL
B- C+ OL
B- E+ OL
C- E+ OL
Q128
B+ C- OL
B+ E- OL
C+ E- OL
B- C+ .583
B- E+ .593
C- E+ OL
I agree with your assessment and will replace all the drivers, finals and those crappy trimmers. Thanks for the info on Bournes 3386. Can Q130 be tested in circuit? Do you know if replacements for the drivers have been called out in a previous post?

 

[glen65]

Glen65,
Can Q130 be tested in circuit?

Try it in circuit first. If you get strange readings then you can pull leads and repeat. You won't need to pull more than two of the leads.
Start with the C to E test first. If it fails that then the transistor is junk, and no point spending time testing it any further.


Concerning the driver replacements, Q126 is a 2SC2275 or equivalent. Replace with 2SC5171 . Q128 is a 2SA975, replace with 2SA1930.

Don't forget to check the diodes. D110 will check higher because it has two in series within the same package.

 

[glen65]

Glen65,
I've tested 'in circuit' all resistors on the right channel.

Make sure you tested the dual emitter power resistor as well.
Also I don't like the looks of any of the resistors marked below. All of the smaller ones are 1/4 watt. The larger ones are 1/2 watt.
Use metal film for the replacements.

 

[BobHelms]

I will start a BOM today. Here is the right channel. Let me know is you see any candidates for replacement there. I will clean up all of "Glue Gun George's" residue. Why don't I just replace Q130 & Q129 with KSC2383YTA's while I'm in there. Can I reuse the existing finals mica insulators for the new Sanken's? They are in good shape, I just changed the silicone compound. Thanks.
Bob Helms

 

[glen65]

I will start a BOM today.

I didn't give you the specifics earlier, but when ordering the pots make sure
they are the side adjustment type the 3386H series.

Here is the right channel. Let me know is you see any candidates for replacement there.

Looks like there is some enamel deterioration on the ends of a couple of those carbon film resistors. One looks badly discolored or burnt.
Also change the potentiometers.

I will clean up all of "Glue Gun George's" residue. Why don't I just replace Q130 & Q129 with KSC2383YTA's while I'm in there.

Certainly won't hurt.

Can I reuse the existing finals mica insulators for the new Sanken's? They are in good shape, I just changed the silicone compound. Thanks.
Bob Helms

Generally micas are reusable unless they are damaged. You need to make sure they are clean before trying to reuse them. However there may be an issue concerning their size. The new devices are going to be just a bit taller than the originals, which might cause an issue. Mouser and digikey for whatever reason don't carry insulators for them.
I'm out of them myself and need to look into a source.

 

[mech986]

 

[BobHelms]

glen65,
Some good news. The finals came in from Digi-Key WITH micas! I need to confirm the drivers and their replacements. you said...

"Concerning the driver replacements, Q126 is a 2SC2275 or equivalent. Replace with 2SA1930. Q128 is a 2SA975, replace with 2SC5171."

My OEM Q127 and Q128 are in fact 2SA985's. Does that make any difference with regard to the 2SC5171 replacements? They are mfg. by NEC. Does that make a difference with regard to the "A" and "C" designations? Thanks.
Bob Helms