• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

SX-3900 Powers on, but no audio output...

SolutionRoom said:
OK. Here are the other side of the amp readings. I turned both pots on the other side fully counterclockwise (VR3 and VR5):

I set the offset to 0 (or close as possible), then set the bias to 56mv, then did transistor readings (below). Then set the bias to 70mv and did another set of transistor readings (below)

SET AT 56MV
From top (closest to the Power Amp Board) from left to right:
+0.597v +54v +0.024v -0.566v -54v +0.009v -0.631v -54v 0.036v +0.629v +54v +0.030v

Then from left to right on the lower (furthest from the Power Amp Board) row:
+0.026v +54v +0.596v +0.005v -54v -0.564v -0.031v -54v -0.636v +0.028v +54v +0.633v

SET AT 70MV
From top (closest to the Power Amp Board) from left to right:
+0.594v +52v +0.021v -0.563v -52v +0.008v -0.648v -52v -0.038v +0.642v +52v +0.038v

Then from left to right on the lower (furthest from the Power Amp Board) row:
+0.025v +52v +0.595v +0.009v -52v -0.562v 0.033v -52v -0.644v +0.034v +52v +0.643v

Whewwww. That's everything.

POST HAS BEEN EDITED (IN green)
Click to expand...

The readings in red, I hope it's just a typo. What did you write on paper?

Everything else looks good, based on what I learned from my own measuring session.
 
markthefixer said:
Well, I went and took some readings myself on the sx-3900 on the bench:

It was set for 70mV a while back, it is cold at about 65f. It seems to be running at 50mV cold.

+0.601 +64.0 +0.006 -0.625 -63.7 -0.039 -0.658 -63.8 -0.071 +0.570 +63.9 -0.021

+0.006 +64.0 +0.601 -0.039 -63.7 -0.626 -0.073 -63.7 -0.656 -0.021 +63.9 +0.572

I didn't read the offset, I just remembered....

Since this unit has been working well, and the distortion figures bear this out, I guess that this type of amplifier circuit tosses out a few assumptions.

So since the readings are similar - it looks like yours is working. So far.
Click to expand...

OK. so what is the next step Mark? Getting late here, so add 2 hours to that and it is very late there. Let me know what to do next and I will take care of it first thing in the morning.
 
We're good to go, remove the dbt, then WITH all the bias adjustments counterclockwise set the bias and offset as per the manual.

edit:
Are my voltage readings (54 and 52v) low due to the DBT? Your readings on your unit are a lot higher.
Click to expand...
he he he he he I was JUST typing that answer B4 the question!! :D :lol:

You will notice your readings were about 54v at 56ma current and 52v at 70ma current.

My readings without the dbt are about +64 and -63.8, so your dbt is pulling down the voltage

and thus turning off the bias and resetting it without the dbt is prudent.
 
I will rest tonight, for the final battle tomorrow. Fingers are crossed for sure. I will let you know how it goes.

The case is almost done. Looking REALLY nice so far. Thanks Mark.
 
Oh, one more thing... PROVIDING ALL GOES WELL AND IT ADJUSTS FINE, Do those resistors we put in across the main caps stay on, or come off?

Also, can I reinstall the other (new) caps on the rectifier board on pins 7-10-14? Remember that I reinstalled the old ones...
 
SolutionRoom said:
Oh, one more thing... PROVIDING ALL GOES WELL AND IT ADJUSTS FINE, Do those resistors we put in across the main caps stay on, or come off?

Also, can I reinstall the other (new) caps on the rectifier board on pins 7-10-14? Remember that I reinstalled the old ones...
Click to expand...


email me pix of both sides of the old caps, then we'll see if they're safety caps.

Might as well keep the resistors on. I intend to.

well, I'm toasted for tonight.... at least I got the laundry done, 3 days late.... (not your fault btw - it just happens...)
 
SolutionRoom said:
Those caps are shown in a pic on POST #69. There is NOTHING written on the other side of them. Roger on the resistors staying.
Click to expand...

OK, I see them, the original C1 & C2 orange drop types didn't fit between the board and the transformer with the shorter main capacitors, right? So you put the square ones off the top - at least it's that way in the unit on the bench.

I'm slightly leery of the new ones, they SHOULD be fine, and they're on the secondary - thus conventional caps are ok, but the new guys are specified in DC volts and I'm not sure those old orange drop ones were an ac or dc specification. 250vac is 1.414 x 250 dc = 353. Plus the thinhs were included to eliminate incoming interference, and right now I'm not sure which is better.

Go with the new ones - but save the old ones.
 
OK. The DBT is OFF and the receiver powers up fine. Soft start relay, then protection relay.

Let it warm up for about an hour (garage is at 56 deg F right now.)

Adjusted the DC offset to as close to zero as possible (+- about 4mv) both sides.

Adjusted the 56mv bias on both sides.

Then adjusted the 70mv bias on both sides.

All is well.

Then checked the output transistors one more time with everything set.

Here are the readings:

From top (closest to the Power Amp Board) from left to right:
+0.641v +63v +0.037v -0.621v -63v -0.017v -0.632v -63v -0.022v +0.644v +63v +0.040v

Then from left to right on the lower (furthest from the Power Amp Board) row:
+0.042v +63v +0.642v -0.012v -63v -0.618v -0.026v -63v -0.625v +0.044v +63v +0.641v

Is it all good? And if so/not, what is the next step?

Fingers triple crossed here...:sigh:
 
Cut, print, It's a Wrap........

Enjoy the fruits of your labor. IF we were willing to go to the effort and expense of getting matched transistors by getting a group of NEW modern ON-Semi's AND installing the different package on the heat sink, the gains would not be worth it.

Plus the analysis is suspect at that level of "precision" as the numbers are shifting over time, the +0.644v and +0.641v are the same voltage reading off the same wire shared by both transistors.

One side comes out as 54 (5 mv high pnp)and 54 (5 mv high npn) which is outstanding.

The other side comes out as 62 and 70 (4 mv high pnp and 4 mv high npn).

markthefixer said:
To compare to the bench 3900 with the On-Semi's"
One side come out as 45 and 45 with no variations, (cold startup - no warmup)

The other side comes out as 50 and 52 ( 2mv high pnp).
Click to expand...

All this is just an indication of how the load is being shared at low power levels. It could change at higher levels.

What would I do if it were mine?
Enjoy the fruits of my labor. Button it up and leave it alone.

P.s. When the resistors arrive, I'll post the resistance readings. It should be interesting...
 
Wow. That's the greatest news I have heard for a long time. I hope everything works on it. I have NEVER heard sound come out of it yet.

If there is something wrong with it, I know where I can find outstanding assistance...

I will catch up on a new thread once the entire project is done with. It should be really nice.

EDIT: Sure sounds sweet! Just like I remember back in 1980.
 
The resistors arrived, and I have completed the 4 wire ohm tests:

R183 end:infinite, center: 0.467 ........ R186 center 0.471, end: 0.467
R185 end: 0.470, center: 0.474 ........ R184 center: 0.482, end: 0.485

These have a 5 percent tolerance, so 0.47 ohm at 5 % is 0.4465 to 0.4935 ohms.
 
markthefixer said:
The resistors arrived, and I have completed the 4 wire ohm tests:

R183 end:infinite, center: 0.467 ........ R186 center 0.471, end: 0.467
R185 end: 0.470, center: 0.474 ........ R184 center: 0.482, end: 0.485

These have a 5 percent tolerance, so 0.47 ohm at 5 % is 0.4465 to 0.4935 ohms.
Click to expand...

SolutionRoom said:
So R183 is bad. What does that mean in the scheme of things for our work? Obviously it was bad, but what does it tell you?
Click to expand...

It is only of academic interest, since it and the output transistors were replaced. One could wonder if the other half sustained as much current, or if some of the current exited the center terminal and through the circuitry rather than flowing through the other half, thus allowing it to survive.

When a transistor capable of handling 12 to 15 amps pops because of current, it is not a gentle event.

It would explain some of the anomalous readings before it's replacement, and by not turning up the bias, probably saved frying the 100 and 200 ohm resistors as well as other mayhem as they popped.

Of course, had it been residing on the bench of markthefixer, I would have checked the resistors with the 4 wire method before applying power.

To anyone curious about ohmmeters, precision ohmmeters and what the heck a 4 wire ohmmeter is:

http://en.wikipedia.org/wiki/Ohmmeter

It's less than one full page, and only uses ohm's law ONCE.... an easy read.
 
So don't keep us in the dark! How does it sound???

For those of us who stayed with the continuing saga, the story got a lot convoluted over time. Could someone post the Cliff Notes version of what was actually finally determined to be the problem and what needed to be done to fix this one?
 
Greg B. said:
So don't keep us in the dark! How does it sound???

For those of us who stayed with the continuing saga, the story got a lot convoluted over time. Could someone post the Cliff Notes version of what was actually finally determined to be the problem and what needed to be done to fix this one?
Click to expand...

No intention of keeping anyone in the dark. The holidays have been consuming me right now. I will continue to post on this thread and will likely start a new one for the cleanup of the receiver rework.

The receiver functions and sounds stellar at this point. Of course, it is still in my garage on the work-bench and connected to a small pair of standard JBL low wattage speakers. The acoustics of that environment is less than ideal obviously and I am anxious to get it inside eventually to hear it in a treated room. Over headphones it sounds pristine.

I have been building a custom solid red oak hardwood case for it too, which will be assembled today and complete. AND there are a few other housekeeping issues to take care of, such as a few burned out lamps for indicators, which I am waiting on replacements to arrive. (coming from the east coast, where the winter weather has been brutal)

There were a couple of small nicks on the top left portion of the faceplate, which I have repaired, but still needs some final detail work to match up with the remaining sections of the faceplate.

Then there is wiring. There is some wiring that needs to be re-worked too, from having the entire unit essentially in pieces for 2 months.

I expect all of this to be done within the next couple of weeks at most and the receiver COMPLETELY done and ready for enjoyment.

Hopefully later today I will post pictures of the new case all assembled.

So at this point, I am very pleased with the results, but it is not complete.

Here are the things (generally speaking) that were done to this receiver since I received it about 2 months ago:

- Tested and checked it from out of the box, was dead, no power
- Found blown 10a fuse, replaced it, turned on for a second, re-blew the 10a fuse
- Pulled the bottom panel off, found one of the output transistor connectors was completely broken with wires shorting on the aluminum heatsink (doh)
- Moved those wires away from the heatsink re-installed another 10a fuse and it powered on, tuned stations, and the preamp all worked. Everything but the power amp portion.
- This is where the thread started
- The troubleshooting began and voltages were taken. Noticed damaged components on the Power Supply board, removed it and shotgun rebuilt the entire PS board.
- Then noticed some damaged transistors and burned resistors on the Power Amplifier board, removed it and shotgun rebuilt it. (To include 100 resistors)
- Replaced all the output transistors (damaging 1 in the troubleshooting process)
-Recapped every single electrolytic in the receiver on all boards, along with most of the ceramics (even though not required)
- And with continuous, relentless and tireless help from Markthefixer ultimately brought the receiver back to refreshed clean life...

There is more, but the above is the condensed version. It truly consumed me for over 2 months, and is still consuming me yet, but is close to being done. I will post a series of pictures once complete to show the progress along the way, either in this thread or another.

Thanks to MTF and AK, this receiver will live for many years to come when it likely was destined to be a boat anchor. And it has thus far triggered happy memories of 1980 when I had my very own brand new SX-3800.
 
Greg B. said:
So don't keep us in the dark! How does it sound???

For those of us who stayed with the continuing saga, the story got a lot convoluted over time. Could someone post the Cliff Notes version of what was actually finally determined to be the problem and what needed to be done to fix this one?
Click to expand...

At the end there, the busted output transistor socket blew up the output transistor and ALL the output transistor sockets were replaced, and the wrong polarity Q101 and Q102 transistors were corrected (ouch, sorry).

After that, the dim bulb tester was contributing to the uncertainty, due to the amount of power drawn by everything else BUT the output circuits, putting it into severe brownout conditions - strangling the 3900.

The danger was the amount of power available to those high power output circuits, and THAT'S where the dim bulb tester was needed, not strangling the lower power regulated voltages.

WE GOT LUCKY, the high power output circuits were run off of a separate toroidal transformer, so we were able to hook just THAT transformer to the dim bulb tester.

After that, there was an emitter voltage that was out of line with the rest of the emitter voltages, inordinately high with the bias current set to minimum.

Checking that emitter voltage, we found a bad emitter resistor (damaged by the original fault), so we replaced them all - cheap insurance considering the investment already in place.

After that the bias currents were set, and the emitter voltages rechecked, looking for differences that would indicate that some output transistors were carrying a lot more of the load and that others were loafing. A recipe for trouble at higher power operation.

Any differences actually found were negligible.

Clear as mud?
 
Thanks Mark for the technical response portion of the work...

Is there a way for me to test for THD figures relatively easily, or do I need a scope or some higher level of expertise to determine that spec? Would love to know where it currently stands as compared to the original 0.005% THD spec.

EDIT: One more thing... I keep reading about burnishing (polishing) the protection relay contacts. Is this something I should do prior to buttoning this thing up for good? Does the protection relay affect the sound quality if not making perfect contact, or is it one of those "it either works, or it doesn't, or is intermittent" things? If so, what is the best way to get the plastic case (cover) off the relay without breaking it?
 
You must log in or register to reply here.
Back
Top Bottom