SX-727 amp board issue

nd0001

nd0001

New Member
I am new to AK and to vintage receivers, but I picked up this 727 and have dove head first into the repair/restoration process. I bought this one in working condition (after a little Deoxit) but the bass/treble pots were basically stuck. So I got a couple used ones off eBay and at the same time bought a re cap kit. I first just replaced the bass/treble pots and checked DC offset and bias, made the necessary adjustments. At this point I decided I would wait on the re cap for now. after I got it back together, I hooked it up and all was working perfectly for a few minutes on the radio, then the right speaker made a pop noise and the unit went into protection. So I figured id take it apart again and go ahead on the re cap, and see if I could figure out what happened. Ultimately I found one bad transistor, Q12, on the AWH-011 amp board. After scouring AK I decided on a replacement of KSA1220. I figured at the same time I would replace the opposite side Q11, and also Q9 and Q10 with KSC2690. I left the other transistors alone for now. I put it back together and turned it on, relay clicked on. Woohoo I fixed it! well, I went straight to DC offset again and was trying to adjust, but the meter didn't move. then I heard the relay click off again and started to smell heat. before I could test anything I got the dreaded puff of smoke from resistor 39 on the amp board. I triple checked pinout on the transistors I replaced and I'm certain they are right, did the resistor maybe get damaged from the original bad transistor? I am going to order up a new 100 ohm 1/2 watt resistor but what should I check next? thanks, sorry for the long post, just wanted to answer as many questions as possible ahead of time.20220421_180350.jpg 20220421_180409.jpg 20220421_180419.jpg
 
Register to hide this ad
any time you work in an amplifier,
you need to turn the idle current (vr3, vr4) down to minimum,
then after DC offset is adjusted, idle current is turned up while being closely watched.

how did you test the transistors? 6 way diode test can spot bad but cannot confirm good.

old trimmers can go open circuit, and I'm not sure what happens in a Vbe multiplier circuit like that.
 
Last edited:
markthefixer said:
any time you work in an amplifier,
you need to turn the idle current (vr3, vr4) down to minimum,
then after DC offset is adjusted, idle current is turned up while being closely watched.

how did you test the transistors? 6 way diode test can spot bad but cannot confirm good.

old trimmers can go open circuit, and I'm not sure what happens in a Vbe multiplier circuit like that.
Click to expand...

Ok, I did not turn them down to minimum first.

I just did the 6 way test with my DMM, but I just ordered an actual transistor tester. I have been buying equipment as I need it, learning a lot haha.

I have read that those trimmers don't have a very long lifespan, should I go through and replace them?

Thank you very much for the help.
 
Update: so I got my transistor tester and I proceeded to pull and test all of them on the left channel of the amp board. Whatever happened ended up burning out Q12 and Q10, and shorted Q14 output transistor between collector and emitter. And also, Q8 is testing faulty. I have all the other replacement part numbers except for Q8... Scouring other threads I came up with KSC2383 or BD137 as possible replacements that are still in production, Can anyone verify or have a better suggestion?
Thanks in advance.

EDIT: actually just read on a totally different thread that another AK member sells NOS 2SC984's on ebay, and sure enough Peace.Love.And.Music had them available. Purchased immediately.
 
Last edited:
Also, is there anything I need to look for outside of the amp board as a possible cause for this problem? or is it possible one of these transistors quit and caused a chain reaction? Just want to nail down the actual problem so I don't burn up more components.
 
Q8 could cause such destruction. even if just dialed up too far. if the trimpot opened up (corrosion on the commutator for the sliding center tap) it could do that just like in other technically different models when the thermal diode wire breaks.

It's the electronic equivalent of a runaway or wide open throttle engine.

Shorts in the speaker wires might do it. Cross connected speaker wires too. Or a voltage (battery or line power) being fed into the amp through the speaker wires.

check the 0.5 ohm resistors carefully.
 
markthefixer said:
Q8 could cause such destruction. even if just dialed up too far. if the trimpot opened up (corrosion on the commutator for the sliding center tap) it could do that just like in other technically different models when the thermal diode wire breaks.

It's the electronic equivalent of a runaway or wide open throttle engine.

Shorts in the speaker wires might do it. Cross connected speaker wires too. Or a voltage (battery or line power) being fed into the amp through the speaker wires.

check the 0.5 ohm resistors carefully.
Click to expand...

Ok, I'm guessing that trimmer was turned up too high. Rookie mistake lol live and learn. Good call on checking those resistors, they are out of spec. Hopefully I'll have all the new stuff by early next week and we will see what happens! Thank you for the guidance
.
 
Well, after replacing all the bad transistors and resistors mentioned before as well as all four trimmer pots, I turned it on and everything seemed fine. I let it sit for 10-15 minutes and then set the dc offset and bias, both set perfectly. I figured I was in the clear so I put it all back together and put it back in its home in my office. I had it on for maybe 5 minutes as I was tuning in the radio and POP out of the left channel again and silence. Seems like I'm back at square one with the original problem I took it apart for in the first place, Q12 is shorted across all three pins. So far all the other parts I replaced seem fine with the DMM but I haven't removed them except for Q12. I am thinking my next step is remove Q2-4 and see how they test out since those are the only transistors on the board I have not replaced yet. Any other suggestions?
 
Unless you want to continue replacing parts I suggest you build a Dim Bulb Tester. You need a way to stop excessive current from frying components until you figure out which component is the root cause.
 
petes'picks said:
Unless you want to continue replacing parts I suggest you build a Dim Bulb Tester. You need a way to stop excessive current from frying components until you figure out which component is the root cause.
Click to expand...
Yes. After the last smoke show I put on, I did build one, and I was plugged into it for the first 15-20 minutes I turned it on. All seemed well, so I set bias and dc offset and called it good. That's when I put it all back together and moved it from the bench back to its home in my office, and it sat playing the radio for another 5 minutes before it cut out again and fried Q12. I guess next time I'll let it sit longer on the dimbulb.
 
Let it sit longer while watching the dc offset and idle current. If you only have 1 DMM, prioritize to idle current monitoring.

in the last "frying" event:

were these the transistors you used?
SX-727,AWH-011,Q1 PNP ,,2SA572-5A,512-KSA992FBU, (still original)
SX-727,AWH-011,Q2 PNP ,,2SA572-5A,512-KSA992FBU, (still original)
SX-727,AWH-011,Q3 PNP ,,2SA572-5A,512-KSA992FBU, (still original)
SX-727,AWH-011,Q4 PNP ,,2SA572-5A,512-KSA992FBU, (still original) (input diff)

SX-727,AWH-011,Q5 NPN,,2SC1124-3,512-KSC2690AYS,
SX-727,AWH-011,Q6 NPN,,2SC1124-3,512-KSC2690AYS, (vas stage)

SX-727,AWH-011,Q7 NPN,,2SC984-B,512-KSC2383YTA, .p.l.m.'s 2sc984
SX-727,AWH-011,Q8 NPN,,2SC984-B,512-KSC2383YTA, .p.l.m.'s 2sc984 idle current stretcher

SX-727,AWH-011,Q9 NPN,,2SC1157-D,512-KSC2690AYS,
SX-727,AWH-011,Q10 NPN,,2SC1157-D,512-KSC2690AYS,
SX-727,AWH-011,Q11 PNP ,,2SA647-D,512-KSA1220AYS,
SX-727,AWH-011,Q12 PNP ,,2SA647-D,512-KSA1220AYS, (drivers)

SX-727,AWH-011,Q13 NPN ,,2SD323-F,863-MJ21194G,10mhz,
SX-727,AWH-011,Q14 NPN ,,2SD323-F,863-MJ21194G,80?mhz
SX-727,AWH-011,Q15 PNP ,,2SA649-L,863-MJ21193G,10mhz,
SX-727,AWH-011,Q16 PNP ,,2SA649-L,863-MJ21193G,80?mhz,(output transistors)

SX-727,AWH-011,D1,,13V,,ZENER DIODE,,WZ-130,512-1N5243B,
SX-727,AWH-011,D2,,13V,,ZENER DIODE,,WZ-130,512-1N5243B,
SX-727,AWH-011,D3,,,,,,1S1212,512-1N4148,

did any 0.5 ohm or 100 ohm emitter resistors fry?
new trimpots?


upload_2022-5-4_19-38-15.png
 
markthefixer said:
were these the transistors you used?
SX-727,AWH-011,Q1 PNP ,,2SA572-5A,512-KSA992FBU, (still original)
SX-727,AWH-011,Q2 PNP ,,2SA572-5A,512-KSA992FBU, (still original)
SX-727,AWH-011,Q3 PNP ,,2SA572-5A,512-KSA992FBU, (still original)
SX-727,AWH-011,Q4 PNP ,,2SA572-5A,512-KSA992FBU, (still original) (input diff)
Click to expand...

Yes, these are still original.

markthefixer said:
SX-727,AWH-011,Q5 NPN,,2SC1124-3,512-KSC2690AYS,
SX-727,AWH-011,Q6 NPN,,2SC1124-3,512-KSC2690AYS, (vas stage)
Click to expand...

These are replaced with KSC2690AYS

markthefixer said:
SX-727,AWH-011,Q9 NPN,,2SC1157-D,512-KSC2690AYS,
SX-727,AWH-011,Q10 NPN,,2SC1157-D,512-KSC2690AYS,
SX-727,AWH-011,Q11 PNP ,,2SA647-D,512-KSA1220AYS,
SX-727,AWH-011,Q12 PNP ,,2SA647-D,512-KSA1220AYS, (drivers)
Click to expand...

9 and 10 have been replaced to KSC2690AYS, as for 11 and 12, both Mouser and Digi-key list KSA1220AYS as unavailable, so according to another thread on here, someone used TTA004B as a substitute. So that is what I installed.

markthefixer said:
SX-727,AWH-011,Q13 NPN ,,2SD323-F,863-MJ21194G,10mhz,
SX-727,AWH-011,Q14 NPN ,,2SD323-F,863-MJ21194G,80?mhz
SX-727,AWH-011,Q15 PNP ,,2SA649-L,863-MJ21193G,10mhz,
SX-727,AWH-011,Q16 PNP ,,2SA649-L,863-MJ21193G,80?mhz,(output transistors
Click to expand...

I replaced Q14 with MJ21194G. I didn't change the opposite Q13 because it tested good and I didn't want to change too much stuff without knowing the problem.

markthefixer said:
SX-727,AWH-011,D1,,13V,,ZENER DIODE,,WZ-130,512-1N5243B,
SX-727,AWH-011,D2,,13V,,ZENER DIODE,,WZ-130,512-1N5243B,
SX-727,AWH-011,D3,,,,,,1S1212,512-1N4148,
Click to expand...

I did not change these, also I did not check them either.

markthefixer said:
did any 0.5 ohm or 100 ohm emitter resistors fry?
new trimpots?
Click to expand...

The previous frying event, one 100 ohm resistor burned up, so I replaced both on the left channel. the .5 ohm resisters tested good when I removed them, but I changed them anyway since they were out. I installed all new trim pots, Vishay T7YA-2.2k-ND for VR3 and 4, and Bourns 3386-P-1-472LF-ND for VR1 and 2.
 
TTA004B is fine.

You skipped Q7 & Q8...
SX-727,AWH-011,Q7 NPN,,2SC984-B,512-KSC2383YTA, .p.l.m.'s 2sc984
SX-727,AWH-011,Q8 NPN,,2SC984-B,512-KSC2383YTA, .p.l.m.'s 2sc984 idle current stretcher
Click to expand...

uh, we usually go in pairs, because the opposite of the pair could have had a lightning bolt of destruction go through it.
the question I always ask myself: Where else DID or could the killing current (Q14's) go?

usually the best way to deal with that type of total destruction is dealing a new deck of silicon.(deck = diodes & transistors)
Two observations:
in a destructive event there is always a race to failure, and the second place losers can be sitting there,
apparently externally undamaged, but a ticking time bomb waiting for power.
The highest incidence of additional semiconductor destruction comes when a mix of (especially untested) old and new is used.

It's hard to separate out a partially working transistor that will fail when full performance is called for.

my mantra: transistors are far far cheaper than callbacks.
 
markthefixer said:
You skipped Q7 & Q8...
Click to expand...

Sorry, I replaced Q8 with an NOS 2SC984 that I got on ebay from apparently another member of this site. I did not replace Q7, but I bought two so I have one to replace it.

Ok, sounds like a plan. I'll replace everything and see what happens. Thanks again for the insight.
 
It has a +/- supply so it would be an old term, OCL design. It is a bootstrapped Voltage Amplifier gain stage
 
I'm surprised MTF didn't mention this. DBT strangles the current, so when you set the Idle offset and BIAS on DBT (Bias actually lower than you think), you need to back off the BIAS to minimum setting, before the unit goes on straight Wall Voltage. Then reset Bias after you go on Wall voltage. Otherwise the Bias set on the DBT will be proportionately HIGHER and WILL cause bad damage. It's called "Tickling the Dragon" . Looks like you got grilled a couple times. Bad Knight, No princess.:)

Don't presume because it works (you fixed it) it'll be ok in the main system right now. Always let a unit "cook" for a day or so before you call it done. Everyone of my units (both tube and SS) get a 96 Hour Engineering run @ approx 1/3 to 1/2 output power. If they can handle that strain for 96 hours with no problem, then they get put in my rotation.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom