AU-4900 Stereo separation issue

Mr. Nick

Active Member
I've got this great amplifier.
I have fully recapped it with Silmic II caps and replaced some polyesters with polypropylenes on preamp etc.
After recapping I put the amp into use for say 9 months or so. Mostly used on weekends only. It sounds wonderful, but lately I've become bothered by almost non-existent stereo separation. Overlooking this issue the amp works perfectly.

When I have the normal setup balance knob works accordingly, there is separation, but almost non-existent.
I have measured resistance on stereo/mono -switch and it appears to work well.
Also I have deoxited all the controls as best as possible. I have also measured a great bunch of resistors from the main amp and they match between the channels. 2sa798 IC been replaced with matched KSA992 sets. Other transistors and outputs are originals.

DC offset is 13mv and 8mv.
Bias is set and checked.

I don't know if I should even look for anything to fix or is it just nature of the beast?

http://audiokarma.org/forums/index.php?threads/c28-lack-of-stereo-separation.800914/
I did this test (post #2):

Try disconnecting the left channel from a source component, then play the source with the balance turned fully right. (Stereo mode selected)

Separation is excellent if there is barely an audible signal on the right channel.

Disconnecting the other source input (CD) and other channel is barely audible when balanced to the other side. Both sides work, but the problem is there when I center the balance knob I hear music from both speakers, even if the one of the input RCA connectors remains disconnected. This is normal for this amp?

I have tried to look for similar posts on AK, but nothing stands out.

WIth every other amp I have worked on the stereo separation have always been much better, more distinctive.
Speakers and CD-source are in good working order.

Any help or info is appreciated.
 
this is not normal for any amp, when you say you have "1 rca disconnected, and sound is coming out of both channels". The questions I would ask are
did you have this separation issue before the recap??

It seems as if both left and right signals have been bridged together before getting to the balance/volume section.
I would look to see if you accidently bridged a connection near or at the volume control or balance control.
Also look for a disconnected shield or ground wire, near the tone amps or control amps. Without the shield (ground ) in place, the left and right channel can just "float together"
Also look and make sure the audio connections into the power amp stage are wired properly (should have 2 signal wires, and their shields, or just 2 signal wires and a black ground wire.
Double check the stereo mono switch making sure its connections are ok, and that its not dirty.
Verify that there is a ground connection on both sections of the balance control. Typically it would be on the left tab of one section, and the right tab of the other section. With power off, center the control and use an ohm meter to verify ground in present on both sections.
Verify that there is a ground connection on both sections of the volume control. Typically this ground would be on the same tab position of both sections. With power off, center the control and use an ohm meter to verify ground in present on both sections (same tab)
Using the ohm meter you can also check that both the volume control, and balance control are working properly. (verify center tab to left, and center tsab to right on each section. The value should vary from the full rating of the control (like 250K ohms) to 0 ohms as you turn CCW or CW. Never should the control show "open".
Lastly, look for bad connections, or possible cracks in circuit board etch runs.
 
Thanks for the detailed input Tom.

The amp as I bought it was in a project condition. The volume level was stuck at 10 o'clock and I was able to mute it by turning balance all the way to the left side. Balancing to the right side didn't make any difference.
So I thought it needed a good deoxit and that worked then for a while. I have brought it back to the bench once after I put it into use for more deoxit shots to the volume control and that settled it for then as I recall it.
One of the wires from muting switch, entering the volume circuit board was almost hanging loose, but I resoldered it back on to the board.

I don't exactly remember how it was straight after recap with the stereo separation, but "float together" as you said makes a lot of sense.

I measured the resistance again for the balance control, black probe under to the ground screw behind the amp and red for measuring solder points. Balance control seems like in order, but I got very odd readings for volume control.
Balance control resistance is only 23K when turned max to one side. About 16K when centered, ground is good for every position for both channels. I should remove it from circuit to get full 100K then?

Volume control (250K) however first gave normal readings, ground seems it was good, but after playing a while, panning between min and max for while. Then at 11 o'clock resistance measured 16K, but at 12 o'clock 12K then at 1 o'clock 16-17K.
After these measurements I brought it very near to min level and now the "inner" section of the potentiometer measured for like 45K and outer section maybe 2K when these were to be at roughly same resistance.
I've used PRF 7-78 for deoxiting, but appears it's not working for this potentiometer anymore.

Can this type of potentiometer be opened carefully or I might just need to hunt down new potentiometer?
That might be a challenge, since there's no volume pcb layout on the service manual at all, so should pay exreme attention to reconnecting the wires. That's why I've been hesitant to do it until now, but it looks like there's no other option left for me.

Here are the pics of the pot and schema part for one channel of the area.
 

Attachments

  • au4900volume1.jpg
    au4900volume1.jpg
    53 KB · Views: 27
  • au4900volume2.jpg
    au4900volume2.jpg
    69.7 KB · Views: 23
  • au4900volume-wiring.jpg
    au4900volume-wiring.jpg
    43.8 KB · Views: 22
Thanks for the detailed input Tom.

The amp as I bought it was in a project condition. The volume level was stuck at 10 o'clock and I was able to mute it by turning balance all the way to the left side. Balancing to the right side didn't make any difference.
So I thought it needed a good deoxit and that worked then for a while. I have brought it back to the bench once after I put it into use for more deoxit shots to the volume control and that settled it for then as I recall it.
One of the wires from muting switch, entering the volume circuit board was almost hanging loose, but I resoldered it back on to the board.

I don't exactly remember how it was straight after recap with the stereo separation, but "float together" as you said makes a lot of sense.

I measured the resistance again for the balance control, black probe under to the ground screw behind the amp and red for measuring solder points. Balance control seems like in order, but I got very odd readings for volume control.
Balance control resistance is only 23K when turned max to one side. About 16K when centered, ground is good for every position for both channels. I should remove it from circuit to get full 100K then?

Volume control (250K) however first gave normal readings, ground seems it was good, but after playing a while, panning between min and max for while. Then at 11 o'clock resistance measured 16K, but at 12 o'clock 12K then at 1 o'clock 16-17K.
After these measurements I brought it very near to min level and now the "inner" section of the potentiometer measured for like 45K and outer section maybe 2K when these were to be at roughly same resistance.
I've used PRF 7-78 for deoxiting, but appears it's not working for this potentiometer anymore.

Can this type of potentiometer be opened carefully or I might just need to hunt down new potentiometer?
That might be a challenge, since there's no volume pcb layout on the service manual at all, so should pay exreme attention to reconnecting the wires. That's why I've been hesitant to do it until now, but it looks like there's no other option left for me.

Here are the pics of the pot and schema part for one channel of the area.

As far as the difference in resistance, many time the wires need to be disconnected from the control (or the control contacts unsoldered from the board) to be able to read it true value, and verify t is going from 0 ohms to 100k for balance, and 250k for volume. I use solderwic to unsolder 2 of the 3 tabs of each control section when I am verifying volume or balance controls) . Remember that volume control has a loudnes tap, so there are 4 tabs per section. 2 tabs are for the 250k resistive pad, 1 tab is the fixed loudness tap (probably at 50k, or 100k), and one tab is the wiper contact. Usually the loudness tap is on the far left or far right.

I am a bit concerned about that cleaner you used for deoxiting. I only use Caig "Faderlube" for rotary or slide controls that have a carbon pad, or strip, because some other "cleaners can damage volume, balance, tone controls etc. Other cleaners like deoxit contact cleaners, D100, D5 are best used for switches that are strictly metal contacts to metal contacts. Faderlube has a nice lubrication included with its cleaning capability, so best for volume and balance and tone controls.
I wouldn't take the control apart because most they can be difficult to take apart, and sometimes they don't wont reassemble because of having to cut or file the shaft to allow you to separate the sections.
I also notice that the connections of volume control seems to have been redone. Iyou should check wire for wire to make sure they are going to the right location.
Also confirm that the wires connected to the tone board near the balance control 5,6,7,8,9,10 are all in the right location by verifying each wire.
 
Hi,

That volume pot in your photo is not the original part, there is no small pcb - the wires are directly terminated to the pot.

I can post some pics of the original pot if that would help.

Cheers

John
 
Thanks for your responses guys!!
I've been little bit busy lately, but my plan is to take the Sansui volume control pot apart on the next week or on a week after next.

Certainly it would help, if you have pictures of the original pot / wiring setup. I believe the pot on my Sansui is original, but of course it might be a replacement anyway.
The board that I have there has small traces of "J" -shape, so it's pretty straightforward to understand.

I think I will let the balance volume pot intact ( almost no harm done ), until I sort out the issue with volume pot, because most certainly there's issue with the volume pot.


By the way... if you have any opinion about this (post #6 with pictures)?
http://audiokarma.org/forums/index.php?threads/sansui-au-317-volume-pot-replacement.642528/
It's not my first plan, but what I got to lose, if the deoxiting won't help my case......

I must say AU 4900 amp is hell of a beast!
I'm really tempted to buy AU 7700 / AU 7900 after hearing this amp out.
Just If I find one in good shape with sensible cost.
 
Hi,

That volume pot in your photo is not the original part, there is no small pcb - the wires are directly terminated to the pot.

I can post some pics of the original pot if that would help.

Cheers

John
not original pot makes sense. whoever swapped it may have miswired it.
 
Hey thanks for the pot photos!
I have a replacement pot then.

So now I had some time to conduct more tests.
Realized that both volume and balance pots seem fine, since I took some AC readings along the signal path as CD was playing through AUX. Left channel seem to work alright with only one input RCA connected. Then same test for right channel and there's the culprit. Right channel bleeds into left channel.

Here are two pics attached and I was able to confirm that everything works good until the signal enters the preamp PCB from volume pot at connection point 11 (schematic). However on the other side of R55/R57 there's AC present, while there should be none as one of the RCAs remains disconnected.

So I figure TR09 might be faulty. It's a suspect, but this is where the trail ends and these are my findings so far.
Voltages seem more or less fine, but it might be leaky I think.

UPDATE:
TR09 might be just fine, since there's some tiny bit of AC present also present on the left channel corresponding location RCAs reversed.
 

Attachments

  • Au4900_ac1.jpg
    Au4900_ac1.jpg
    60.8 KB · Views: 4
  • Au4900_ac2.jpg
    Au4900_ac2.jpg
    89.1 KB · Views: 4
Last edited:
To be more precise about my earlier post I have taken these AC readings with only one RCA input jack connected at a time. Left channel works perfectly with one input jack connected, just as you would expect it.
And now I re-measured the AC voltages today, there's none present with only left RCA connected. With only right RCA connected I see AC "leak" just as in those "#9" pictures on left channel.
Don't this effectively rule out the potentiometer problems regarding volume and balance controls, since this AC leak problem appears in the signal path after these controls?

Yesterday I made conclusions hastily, since I think with the potentiometer and likewise the AC readings may have been a result of bad contact from my DMM probe to solder pin.
At preamp exit points 13 and 14 for each channel I have one third of the left AC present that what's on the right when only right channel input RCA is connected.

Tom, before posting this thread I tried to resolve this issue with all the existing posts here on AK and took many readings from various locations and made sure it's not say Stereo/mono switch issue.
Also pretty confident the traces are not bridged either. I have measured many resistors and they all seem up to spec.
Also I have tested the functions now with better care and the controls work, volume among balance controls works, but you can't detect it properly unless you play a disc with distinctive channel separation.
Some discs don't have very distinctive separation at all. So I don't believe it's miswiring from the previous owner or repair service. Faderlube is probably a very good product, but it's not easy to find that in Europe.
Many have reported PRF 7-78 is excellent for potentiometer and general electrician contact spray, but not like basic CRC.

What you think could it be an issue regarding right channel TR08 or TR10?
I've measured the voltages and compared them to corresponding left ones and they're roughly at same voltage, so the problem is not significant.
 
Last edited:
Ok, I've been playing around. Measuring AC voltages from various locations.
Before I get into any details more, I discovered the replacement volume potentiometer is actually only 100K pot, not 250K as per manual.
Is it now self-explanatory that the missing resistance will not create separation needed between channels even, if the pots are working just fine?
 
Thank you all for the help.
I have just successfully installed a replacement pot from a vintage Marantz (dual 250K w/ loudness).
Lucky find from the bay.
Tested it and channels are now separated, also stereo/mono works better, though it does not mute the other channel all the way when panned all the way to left or right.

Now the Sansui rocks the house like never before.
What a substantial improvement to the curve of loudness function!!
Little guy is really pushin hard those bass notes.
House shakin' all over :biggrin:
 
Back
Top Bottom