Sansui AU-X11 repair and restoration

[pete_mac]

Here is one more pic from Mr Amp8 below. Looks like he kept the 1 uF caps but added some film caps of ~10uF value, as well as some electrolytics of value in the range of a hundred or so uF. Overkill? I know the community is a bit divided on the bypass caps, but 1uF paralleling 20,000 uF... seems like it could use some help from something a tad bigger. Thoughts?

Of interest, many of the Japanese domestic market Sansuis feature cascaded film caps and/or electrolytic bypass caps, which were not seen on the international models eg.

Top to bottom - AU-707 (AU-717 equivalent), B-2302, AU-D707 x 2 (AU-819 equivalent)

 

[stereofun]

From Elliott sound products ( http://sound.whsites.net/power-supplies.htm) :

"Use of a small (e.g. 1µF) polyester, polycarbonate or polypropylene capacitor across the DC output is a common practice. [ - ]. I have never found the necessity to add a bypass cap across the electros to maintain amp stability, but it cannot hurt. Some amplifiers will oscillate if the supply impedance is allowed to exceed some specific (low) impedance at high frequencies. Essentially it is a good idea, and in the greater scheme of things, is inexpensive. Should you choose not to include a film bypass, it is unlikely that anything 'bad' will happen - the impedance of the large electrolytic will usually remain much lower than that of the film cap at any frequency below 1MHz or so. [ - ] The proper place for film bypass capacitors is on the amplifier board itself - not directly in parallel with the filter capacitors. You can do both, but only the caps on the power amp board will have any useful effect."
---------------
Seems like an inexpensive insurance policy against oscillation in high speeds amps with no bearing on sonic - but keep the leads as short as possible.

 

[Willy6]

Ronito, I also don't see your pics.

Yes, on corner of the Left channel output heat sink is not screwed to the chassis. As you say it looks to be like that from factory. I don't know it. I can push the heat sink and move it around a bit with one finger. Here is what Amp8 did and I may try something along those lines:

I have a question for the AU-X1 owners. X1 manual says 25 mV across test points for setting bias, but manual shows 0.33 Ohm emitter resistors in schematic and part list, whereas I think most of you (judging by pics from various threads) have 0.47 Ohm emitter resistors installed. Are you assuming the 25 mV is correct for the 0.33 Ohm or 0.47 Ohm? I.e. Did you assume the goal is ~75 mA or ~50 mA? Sorry if I'm being pedantic and it doesn't make much difference, but I won't be happy until I get your thoughts

This came up partly because I saw in another thread that some folk are using the 25 mV from the AU-X1 manual to set their bias in the AU-XII, but the AU-XII has 0.22 Ohm emitter resistors on the power transistors so one would more likely want to use ~11 mV (assuming ~50mA idle is the target, or maybe 16 mV or so if you shoot for 75 mA.). For what it is worth I can tell you is that on my good (Right) channel, the voltage was sitting on about 10 mV before I touched anything. Also, I confirmed on Mr Amp8's pics that he also has 0.22 Ohm emitters in his AU-XII units.)

Regarding the little caps across the big supply caps, I can confirm they are 1 uF value across each 20,000uF pair, same as in the AU-X1. Here is a pic of the present/stock situation:

Here is one more pic from Mr Amp8 below. Looks like he kept the 1 uF caps but added some film caps of ~10uF value, as well as some electrolytics of value in the range of a hundred or so uF. Overkill? I know the community is a bit divided on the bypass caps, but 1uF paralleling 20,000 uF... seems like it could use some help from something a tad bigger. Thoughts?

This explains why the dbt has a faint glow if i set the bias on my AU X-1 to 25mv and if I turn it down around 16mv the bulb go's completely out! Mine has the 0.47 ohm emitters and the service manual schematic says they are 0.33 ohm...

 

[Hyperion]

I have 0.47ohm emitter resistors in my X1, and confusingly in the SM (for the X1) it says either 25mV (bias setting instructions) or 20mV (schematic) - but this is across 2 emitter resistors which yields:

2 x 0.33ohm @ 20mV = 30mA.
2 x 0.47ohm @ 20mV = 21mA.
2 x 0.47ohm @ 25mV = 26mA - this is what I consider to be correct for the X1.

And in the absence of an SM for the X11 I believe it is a good starting point.

2 x 0.22ohm @ 11.4mV = 26mA

 

[Ronito6]

My X11 driver voltages.

I managed to do the voltage check earlier than I anticipated.

Truth be told, I am a little hesitant to issue these results as they show how much work my amp needs.

BUT as there is some chance that the info that I provide here might be helpful...here are the results.

I think the results here suggest that 25mv is likely the optimal voltage for the X11 drivers.

This is where you check that driver voltage. (After a bit of poking about I sussed it.)

A close-up for detail.

Hope this tidbit helps!

I suppose the flat amp and the regulated power supply board for theX11 would follow the same procedures and targets laid out in the service manual for the X1

The equalizer boards would follow procedures and targets laid out in the document in the pic below. This pic of a Sansui service bulletin provided by Nicowico and posted at http://audiokarma.org/forums/index.php?attachments/img_3759-jpg.635616/

All together I believe these are nearly all the necessary voltage check pionts for an X11.

The only question left is what are the procedures and targets for the "head amp" phono boards.

I suppose the X1 service manual with the service bulletin above can help point the way forward for those.

 

[smurfer77]

John, thanks for the AU-X1 bias info. I didn't realize the test points were placed over two emitters. I've checked and it is the same on my AU-XII - bias test points are over two emitter resistors. So I was right to shoot for 11mV (but wrong on my target current by a factor of 2x! And 25mA or so makes me more comfortable than 50mA or so). Thanks!

 

[smurfer77]

I think the results here suggest that 25mv is likely the optimal voltage for the X11 drivers.

Ronito, thanks so much for pulling together the info into your last post.

Regarding the bias voltage across those test points, as discussed in the other posts here, the emitter resistors are just under half the value of those in the AU-XII. Those test points just follow the PCB to the other/output side and then wires go across the emitter resistors of the inner pair of NPN/PNP power transistors. (I will post a annotated pic of the AU-XII PCB to make things clearer shortly). It's exactly the same setup as the AU-XI in this regard, except the emitter resistor values. So we need to shoot for a voltage at those test points just under half that of the AU-X1 manual, assuming Sansui used same bias current in both models. This means about 11.4mV (giving ~26mA) over the bias test points. You are probably fine with the values you have, but you could easily give a turn to the little variable resistor (not the one at the edge of the board, that is for DC adjustment, but the other VR near the middle of the board - again, will annotate my driver board pic to make that clearer too).

Cheers
David

EDIT: PS. In an earlier post I noted that I recorded the test point voltage for bias at about 10 mV on my good channel, before tinkering.

 

[smurfer77]

Since the driver board is the one that supposedly differs the most from the AU-X1 I will focus on that a bit more (since we have a schematic for the AU-X1 but not for the AU-X11)

AU-X1 F-2775 driver board for reference (from Amp8's website):

AU-XII 3460 driver board (from my unit):

So there are a few obvious differences:
- you get an extra filter cap on each rail (note the 6 large caps on the X11, vs 4 on the X1) (this PS is for front end of driver board. Back end is supplied by power transistor supply, already filtered by the big caps if I'm not mistaken.)
- the AU-X1 has those big caps bypassed with 0.22uF caps. My AU-X11 has empty slots for bypass caps (just above the caps you can just see one, marked C16 under the red wire).
- the dual FET at the input is a different package.
- You can two entirely new sections marked "Protector (l)" at top right and bottom left on the AU-X11 board, along with some additional connectors to those.
- mounting of a few transistors is quite different. The bias transistor no longer has that metal mounting plate and the other transistors that were also on that get there own heat sinks now. The thermal feedback to the bias transistor is via a 'leg' on the power transistor heatsink that presses onto the little transistor in the bias circuit, with a little bit of thermal paste between the two. (I assume the AU-X1 has the same leg on the heatsink but it made contact with the metal plate that the bias transistor is mounting onto in that case.) I like the bias transistor getting being thermally coupled only to the power transistor heat sink, makes sense to me. However, I don't think I like the new thermal connection between the power transistor and the bias transistor... looks fairly easy for that thermal paste to unsettle over time or mechanical strain (remember that the power transistor heatsink actually does move around a bit due to a missing screw in the mounting scheme, mentioned above. So I now think adding that extra screw somehow is essential). And, in my cast, the thermal compound has dried out and a lot of it escaped (you can see some of it on some caps actually!). So if you have an AU-X11 redoing the thermal paste in this position is probably important as it seems easy for a few reasons for the thermal feedback to be lost. There is another problem. When I lift the transistor from the sponge backing I can see that the sponge has permanently deformed over time and is no longer pressing the transistor back up onto the thermal coupling 'leg', at least, not as it was before.... another way the connection could be broken and the thermal paste can escape more easily as it crumbles.

Here is a pic of the 'leg' from the power transistor heat sink that presses onto the bias transistor. You can see the leg is meant to press directly only the transistor, which is pushed into sponge on the board.

But the circuit does look very similar to the AU-X1 otherwise. Including the 6 black flag caps (although values and positions might be a bit different).

Following is some AU-X1 reference material:
AU-X1 PCB layout:

AU-X1 part list:

AU-X1 driver board & power outputs:

AU-X1 driver board voltages:

 

[smurfer77]

Ronito, Here is the annotated driver board I mentioned. NOTE: Use with caution (I was tired when I made this, and still have a couple of pins/connectors to identify). Click the image to zoom in. Note the board is grounded to several different ways. A couple of the components in the 'protector circuit' have doubled up on component names from the driver board so I marked those instances on the protection circuit with an "*". For example, there are two C8 and two FT1 on the board.

 

[smurfer77]

And one more version, to show a bit more clearly the added "Protector" sections.

 

[Hyperion]

Those pictures are really interesting to me, noting the differences between the 2 models, thanks for posting.

(you have a random '2071B' label next to C9)

 

[Ronito6]

Amp forensics. Super detail. Incredible

 

[Ronito6]

Here is a stickler to ponder.

The X11 drivers use a 2sk129.

Amp8 knows of a substitute, but alas refuses to share what it might be. (At least with me...so far.)

If it needs replacing...any ideas?

 

[Hyperion]

See post #8 here

http://www.audiokarma.org/forums/index.php?threads/au-919-dc-offset-transistor-advice.264949/

Also Littlediode has 2SK170 in stock, (use 2) but origin probably unknown.

 

[Ronito6]

See post #8 here

http://www.audiokarma.org/forums/index.php?threads/au-919-dc-offset-transistor-advice.264949/

Also Littlediode has 2SK170 in stock, (use 2) but origin probably unknown.

Well that's sorted.

 

[Hyperion]

One of these days I'll try to find a SM dual JFET device.

 

[Ronito6]

One of these days I'll try to find a SM dual JFET device.

Seems the LSK389 mentioned in that post is 2sk389 but in a more compact package.

2sk389 looks strongly like one of the substitutions used by Amp8 for the 2sk129.

 

[Hyperion]

Very interesting, there is even a UK distributor not that I need any right now, but I might buy some anyway. (they are a bit spendy at £5.94 each - nearly $8)

 

[GPS16]

Hi guys,

just read the comments about 1uF caps across 20,000uF. The 1uF caps will be there to reduce HF noise. The 20000 jobs will be a rolled foil construction which will have a large stray inductance at high frequencies. Any HF noise may not get past the stray inductance to see the large capacitance within. I have seen 20000uF, 1uf and 0.01uF caps in parallel before now!!!

GPS16

 

[Ronito6]

Hi guys,

just read the comments about 1uF caps across 20,000uF. The 1uF caps will be there to reduce HF noise. The 20000 jobs will be a rolled foil construction which will have a large stray inductance at high frequencies. Any HF noise may not get past the stray inductance to see the large capacitance within. I have seen 20000uF, 1uf and 0.01uF caps in parallel before now!!!

GPS16

Very good info. (Especially to my amateur ears.)

What might the difference like on an oscilloscope I wonder. (The difference between adding the extra cap vs going without one.)

Well...time to do a bit of Google-ing to find out.