Sansui AU-517 DC-offset problem

[HansVenlo]

More than a year ago I restored a Sansui AU-517 with a lot of help from Sansui specialist here on AK. It is still working fine.

A few months ago, I was able to get one for myself, as an upgrade from my Sansui AU-217.

It played OK, but not super. This was expected since it is more than 45 years old.

I found that a previous owner replaced the relay with an Italian one, but he put it on the solder side of the board. The pins were too big for the board and the solder pads were damaged. I ordered a new OMRON MY02-24V DC drop in and managed to get it working, on the right side of the board. I hate it when it is not original. But yes, the Power supply is a real PITA, to work on.

I restored the whole unit following the guidelines of Leestereo in his AU-517 and his AU-717 thread. Great read, that thread. Also read almost all of Hyperion's advices.

This is what I've done so far:
- Replaced all electrolytics and upgraded some of them, with Nichicon Fine Gold, Kemet and Panasonic.
- Replaced all VD1212 varistors with two 1N4148 in series.

Driver boards:
- Replaced C12/13 with film cap 1uF.
- Installed new Bourns Cermet pots.
- Replaced fusistors with 150W metal film resistors.
- Replaced zenerdiodes with Vishay 2% zeners.
- Replaced other diodes with 1N4148.

Equalizer:
- Replaced fusistors on equalizer board with 82,5W metal film resistors.

Power supply:
- Replaced all zeners.
- Replaced power caps with Kemet 15.000uF and Panasonic bypass 2,2 uF.

It did not work very well, because I found that a new zener on the right driver board was installed the wrong way around. Yes I know, check beforehand and afterwards. But I was very confident.

After that I tried to set DC offset and Bias. DC offset was very bumpy on the right board. One of the 2-watt resistors (R46) was 40% lower then stock (6,2W) instead of 10W. So I corrected the resistor.

I went the full monty for the power supply: replaced all big rectifiers and other small diodes and replaced TR01/02 with KSC1845 matched!
After correcting that Voltages were OK and very stable (+46,48 and -/-46,46 for left and +46,51 and -/-46,54 for right channel.
Bias was set for both sides, but DC kept on jumping all over the place.
I followed Leestereo’s advice in the 717 thread and had changed the R16 resistors in the driver boards to 27,4kW. I realised that was a difference with the 517 schematics, so I put the old 22kW back, which solved the running and jumping DC.

Starting up I noticed that the DBT did not go down in a straight line. Halfway down it flickers up for a moment and then further down. I found out that it originates from F-2722 Right Channel board.

After testing it comes out of protection when I keep Bias low (< 10mV), it plays a while and goes back in protection. But it is working at least.

F-2721 Bias test point: minus to chassis is minus 26mV.
: plus, to chassis is minus 11mV
F-2722 Bias test point: minus to chassis is minus 103mV.
: plus, to chassis is minus 86mV
I cannot explain these differences.

DC offset pre relay : pin 21 minus 10 to minus 14 mV
: pin 22 minus 92mV to minus 100 mV
I cannot explain these differences.
Checked all orientations twice and measured all resistors on the driver boards and the power supply. Everything seems to be OK. I changed the Driver board connections. F-2721 is playing fine, also with the other connections to the power supply, ButF-2722 is not.

After setting BIAS (rather low) and DC offset again, both channels play, but it sounds harsh.

I also measured hfe on Right channel driver board power powertransistors:

TR11 A899 hfe 157
TR08 C1904 hfe 115
TR05 A899 hfe 178
TR07 C1904 hfe 166

Could the difference between TR08 and TR07 cause my problems?Or the difference between TR11 and TR07?

I’m stuck, ran out of options, please can someone help me in the right direction?

 

[HansVenlo]

In the above post, typos are introduced by copying the concept post from my Mac to AudioKarma. It looks like the symbol for Ohm was mistakenly converted to W. Looks very funny though, 27,4 kilowatt resistors… must be very big . But I guess you figured it out: it must read 27,4 kOhms etc. I apologise.

Anyways: some good news. Managed to finally get DC-offset and Bias ok. Unit comes out of protection and she sings again.
Still behaviour on DBT is funny. Bright-dimming to half-brightening to 3/4-and finally dimming to low. When I disconnect the connector on F-2721 (left driverboard) that runs to the pre-amp/power-amp switch on the back of the unit, it dims normally on power-up.
Any suggestions ?

 

[HansVenlo]

I changed all the transistors on the equalizer for new ones. Plays OK after readjusting DC-Offset.

Started checking the tone control board. I remembered a discussion on AK about the quality of the silk screens and the service manual. The advice was to follow the silkscreen.
When checking F-2673 I found a difference between the silk screen, the service manual and reality. Both silk screen and manual say that the base of TR05 is located towards the short outside of the board. My picture (yellow caps) showed that the SA750 transistor is mounted with the emitter towards the short outside, just like in the AU-517 of my son that I recapped a year ago (blue caps).
So be cautious: neither the silkscreen nor the manuals are free from errors. Following the printing on the board may lead to mistakes.

 

[HansVenlo]

Still my problem with starting the amplifier via DBT is not solved.
Should I renew all the transistors on the tone control board as wel, and also the power transistors on the Driver Board?
Or could it be just an improper ground connection? Anybody any suggestion?

 

[Hyperion]

If the only difference between bad DBT power-up and good DBT power up is disconnecting the input to the F-2721 driver board, I would suggest there may be some DC offset coming out of the F-2673 board because of a possible error or fault? - so I think you need to check your work on that board. You could try plugging the connector (if it is the input connector???) from the 'good' channel into the F-2721 driver board, if the unit powers up OK (via DBT) - then you know the error/fault is before the driver board.

 

[HansVenlo]

Thank you for responding Hyperion. It is perfectly well possible I made a mistake. I already started a systematical check bij changing plugs from the good channel to the "bad channel". Did not very well by feeding the right channel with double voltage ( e on 6 in my picture) The corresponding song is "Smoke gets in your eyes". Thanks to DBT it survived this abuse more or less.
After correcting that awful mistake it was still playing, but since I cannot find any fire marks I will have to check the right driver board thoroughly.
My unit is now playing, but with some noise in the right channel and still a very difficult DC setting. Also the unit has a volume dependant hum, which in my opinion could originate from improper grounding or bad shielding. (all shield are off att the moment) But I will deal with that later.
I will follow your advice and check the connectors again and F-2673 board and report afterwards.
Hans

 

[HansVenlo]

Hi Hyperion,
It took me a while to check things. Found the following:
Left driver F-2721 with its own connectors: Good DBT and playing OK.
Right driver F-2722 with its own connectors: amp stays in protection, Good DBT however, but no music.
Left driver with connecters from the right: Good DBT and playing OK.
Right driver with connectors from left: amp stays in protection, Good DBT however, but no music.
So I concluded that either Bias or DC-offset on the right channel were way of.
Bias set higher approaching 20mV and DC as good as possible near 0.
It is playing, but harsh.

I think because DC setting is so very difficult, and goes up and down.

I checked with connectors to pre-amp/separate switch disconnected: problem goes away
Cleaned and recapped the preamp-separate switch. Did not do any good.
Disconnected the connectors from the equalizer and the problem goes away!

So your suggestion that something might be wrong with the DC coming from theTone control board is probably true.
But I cannot find any mistakes on the tone control: I swapped the old electrolytics out for new ones, and changed the transistors
2SC1222 to KSC1845F and 2SA750 to KSA992F. All orientations correct. Checked twice.

So I went to recheck my powersupply and the output voltages. No boards connected, and no DBT.
I found the following:
Both transformers are putting out 38,94 - 38,97V (Service manual: 35,5V)
Output to the driver boards is 51,99 and -/-51,89 for right channel
and 51,84 and -/-51,9 for the left channel. (Service manual: 51V)

Now the very strange and disturbing thing: The outputs to the equalizer and the Tone control
should be Plus and Minus 31V. I measured the plusses to be 30,39 in the right channel and the minuses 33,62!
And for the left channel it was plus 30,56 and -33,32!
So the equalizer and the tone control are very unbalanced fed, with a 3V difference.

Does that make any sense to you wrt the unadjustable bias?

For you info: I followed Leestereos guidelines very close when restoring this amp
Hope you can push me in the right direction! I have absolutely no clue!

Hans

 

[HansVenlo]

Checked a few other things. I found out that the tone defeat switch is extremely noisy. I also found out that turning up the bass control by two clicks causes the protector to switch on. Turning bass control back wil recover the amp from the protector.
Does this mean that the tone control is at fault? Or just the bass volume circuit board, or a dirty bass pot?
I intend to take it out and check every comp[onent of the bass and treble volume circuit board.

 

[skippy124]

Hi Hans,

A couple of questions if you don’t mind -

- are you doing all this testing with the amp on the DBT ?
- what is the wattage of the DBT lamp?
- when you ah e stated in your last post that the unit goes into protection with a couple of clicks of the bass control. Are you playing music with speakers conntectc, and on DBT or plugged straight into mains power?

Your answers may be of assistance in determining what may be the problem with the protection issue.

Cheers
John

 

[HansVenlo]

Hi John,

Thank you for chiming in. I surely can use help with this beast.
To answer your questions:
- all testing is done on DBT.
- bulb is 75 Watt
- I keep it on DBT since I think there is ad least a bad connection somewhere.
- on DBT it goes into protection on click three of the bass control. This happens with and without a source connected, an even with the volumecontrol off.
When I switch on, I hear all kind of (not very loud) sound coming from the speakers. I use old cheap speakers for testing, I don’t want to risk my Dovedales or AR11
This noisy sputtering and sparkling sound is going softer after a few minutes. Music played via both channels, sounds ok.
I’m stuck in this problem,
However Hyperions suggestion to look somewhere in the tone control proved to be a good hint, it revealed this strange bass volume behaviour.
All you helpful people from AK: your help is much appreciated
Hans.

 

[Jefft91]

Hi John,

Thank you for chiming in. I surely can use help with this beast.
To answer your questions:
- all testing is done on DBT.
- bulb is 75 Watt
- I keep it on DBT since I think there is ad least a bad connection somewhere.
- on DBT it goes into protection on click three of the bass control. This happens with and without a source connected, an even with the volumecontrol off.
When I switch on, I hear all kind of (not very loud) sound coming from the speakers. I use old cheap speakers for testing, I don’t want to risk my Dovedales or AR11
This noisy sputtering and sparkling sound is going softer after a few minutes. Music played via both channels, sounds ok.
I’m stuck in this problem,
However Hyperions suggestion to look somewhere in the tone control proved to be a good hint, it revealed this strange bass volume behaviour.
All you helpful people from AK: your help is much appreciated
Hans.

Hans,

I've been watching this since you started. A few more questions that might help narrow things down:

1. Is the noise you describe in both channels, or just the bad channel?
2. When you swapped connectors from channel 1 to channel 2 in post #7, did that include the input signals?
3. How closely have you inspected the tone board? Did you look at every solder joint and every component you replaced to be sure they're correct?

 

[HansVenlo]

Hi John,
Thank you for al your help!
-The noise and cracking is in both channels.
- I only swapped the connectors in the amp.
I left the left and right input in place. So every connector from left went to their respective positions on the right channel.
- I checked the tone control, I changed TR05 for a fresh one because it caused noise and popping when prodding it. I reflowed most solder joints, but not all.
Today I started checking again, beginning with the PS. Except from the unbalance in the outputs to equalizer and tone control (negative -/-33,5V and positive + 30,2)the PS looks OK.
There is lots of background noise and cracking, and sometimes DBT goes higher and protection kicks in. When I use the tone control bass volume, 3 clicks up: protection kicks in and high DBT.
Could it be I damaged a few transistors earlier when I put positive Voltage on junction 4 of F-2722?
At this moment it plays, but the sound is not clean.

 

[HansVenlo]

I found and resolved at least one problem:
New transistor on position TR04 on the equalizer was not soldered at all. I corrected that an all popping and cracking and noise went away!
But DC is still jumping like cray on both channels (max-min is +3V to -2V. But it sounds better.
And my tone control triggers the protection at plus 3 clicks on the basspot.
is there anyone on AK with a suggestion?

 

[GrisPato]

Hi Hans,
I live relatively close to Venlo. If you would like I could visit and try to help. At least as another set of eyes.

 

[GrisPato]

I changed all the transistors on the equalizer for new ones. Plays OK after readjusting DC-Offset.

Started checking the tone control board. I remembered a discussion on AK about the quality of the silk screens and the service manual. The advice was to follow the silkscreen.
When checking F-2673 I found a difference between the silk screen, the service manual and reality. Both silk screen and manual say that the base of TR05 is located towards the short outside of the board. My picture (yellow caps) showed that the SA750 transistor is mounted with the emitter towards the short outside, just like in the AU-517 of my son that I recapped a year ago (blue caps).
So be cautious: neither the silkscreen nor the manuals are free from errors. Following the printing on the board may lead to mistakes.

I don't know if this is a picture of the board in present state, but the board with the Nichicon ufg is missing the VD1212 diodes. Forget that, I see them now!

 

[HansVenlo]

Checked a lot today. Openen and cleaned the pre-main switch on the back again. I found a few sloppy soldered joints. I corrected that. Playing quality gets better (cleaner) but the jumping DC keeps on going from plus to minus and back.
with all connectors attached the DBT goes to weak glowing in two stages, as before.
When playing music the protector comes in on heavier music, and restores to normal when sound is a bit weaker.
And the bass control on +3 triggers it as wel.
@GrisPato: yes, I could definitely use some extra eyes.

 

[Jefft91]

Hans,

I'd take up GrisPato's offer!

Regarding the mis-match of the +33V and -33V supplies, they would not likely cause the issues you are reporting. If you want them more closely matched, you need to adjust the resistor ratios of (R32 + R34) with R36.

As to the voltage jumping, are you referring to the output offset voltage, the power supply voltage, or something else? I'm assuming it's NOT the output offset voltage, because this would cause the amp to go into protect mode immediately (and could cause damage to speakers).

 

[spark1]

2-3V is unlikely to damage the speakers, but they really should be disconnected as you work to find/fix the underlying problem(s).

However, as @Jefft91 suggested, if those measurements are DC offset at the speaker terminals, you have a failure in the protection circuitry (to go along with your voltage imbalance from the power board and the problem triggering the unit into protection as volume/output voltage goes up).

 

[HansVenlo]

Hans,

I'd take up GrisPato's offer!

Regarding the mis-match of the +33V and -33V supplies, they would not likely cause the issues you are reporting. If you want them more closely matched, you need to adjust the resistor ratios of (R32 + R34) with R36.

As to the voltage jumping, are you referring to the output offset voltage, the power supply voltage, or something else? I'm assuming it's NOT the output offset voltage, because this would cause the amp to go into protect mode immediately (and could cause damage to speakers).

The mismatch is in the supply rails:it should be +31V and -31V but the negative rails are -33V. The voltage jumping is the DC on the speaker outputs, it jumps and it does not respond to VR01 and VR02. Hopefully with the help of GrisPato we will find the bug.

 

[HansVenlo]

2-3V is unlikely to damage the speakers, but they really should be disconnected as you work to find/fix the underlying problem(s).

However, as @Jefft91 suggested, if those measurements are DC offset at the speaker terminals, you have a failure in the protection circuitry (to go along with your voltage imbalance from the power board and the problem triggering the unit into protection as volume/output voltage goes up).

Thank you for your guidance. The imbalance is absolutely originating from the PS/Protector, but I can’t find any problem. If I measure the outputs of the PS I find good and stable almost 52V but the -31V is measured -33,3V,
Maybe the new zeners are not OK. I could put the old ones back in place?
Problem is that it is not one fault but several faults.
But we will fix it, with the help of all our AK friends.
Hans