Sansui AU-X11 Restoration

The different gain of the 2ST is not likely to be a problem. Someone please correct me if I'm wrong in this particular case (I don't have schematics in front of me now) but in this case the global gain is clamped the the circuit and the feedback, and is not dictated by the output transistors alone, and some variation in output transistor gain will be okay (provided the gain of outputs is not less than original part). Consider that from the preamp out (lets say 2V) to the speaker outlets (lets say 20V) the overall gain is only on the order of 10... it's always this way, to reduce distortion etc. That is oversimplifying but you get the idea.
 
Updates:

1. Driver Board:
All the carbon films in the driver board section have been replaced with Metal films from MultiComp(Thanks Hyperion) rated at 0.5 watts each, and a tolerance rating of 0.1%, better temp co-efficient.
Some things i noted and wanted to share, the MultiComp Metal Films resistors are a tad bit bigger than the 0.33 watt Carbon Film resistors, you might want to adjust the spacing of the resistors, so you atleast get around 0.5mm to 1mm spacing between resistors placed beneath each other or the sides .

I did check each and every resistor before replacing with the multimeter to check for their rated values before soldering them in.
Also verified for trace continuity after the solder work was done.

What is left are couple of Carbon Films in the protector section, which i had missed to place an order for, and the potentiometer from Bourns, the Bourns potentiometers that i have, have the second wiper pin closer to the pins 1 and 2 , i can adjust the pin so they fit in the holes properly, but i am going to check if i can find a potentiometer that has a wider spacing for the middle wiper pin.

Inspired by work done by Hyperion, i have sourced the Fish spine beads to be used for non-flammable resistors, so i wait for them to arrive.

Ronito6 is a knowledge bank, and am very very thankful to him for the support rendered.

Driver Board: Front
Driver Board Front .jpg

Driver Board Rear:
Driver Board Rear.jpg
Only Carbon Films in driver board replaced, and couple of other carbon films in protector section which i had already.

The small transistors mounted in heat sinks have been re-greased and soldered back.
I noticed that one of the A968 transistors, had been leaky and had very low Hfe gain compared to the other A968 transistor, i have replaced them with the below transistors, from Fujitsu i suppose, these measured well and equal to the Toshiba A968 transistor, so i replaced them both with the ones from Fujitsu.
A968B.jpg

I would want to know, if you guys have any inputs with this replacement used, in case i am wrong, i will reconsider my choices.

Heat Sink(Transistor Mount Assembly):
heat sink.jpg

I have replaced the emitter cement resistors, with those from Vishay Dale LVR series if remember correctly, and the base resistors with non-flammable ones from Vishay, i have to admit , i am not happy done with the work done on Base Resistors, the key is to bend them properly, in such that you can remove the transistor mount screws easily, i had only 8 pieces of these resistors, and was not able to do this job properly.
I might add some more solder to the emitter resistors, the base resistors are just mounted and not soldered for now, as the base wires feed in there, and i will solder each and every one of them during this work.

Side Note: The transistors mounts were removed prior to this work being done, the emitter and base contacts of the TO-3 transistors were cleaned, with a dose of Deoxit, followed by IPA wash completely, as i found some heat sink paste inside the holes where the emitter and base leads feed in.
If one removes the driver board completely isolating it from the heat sink, its very easy to get this job done.

Lesson Learnt (Mistakes Were Made But Never Again) !:
After desoldering the components from the driver board, i rubbed it with a dose of IPA, to clean the contacts before soldering, little did i notice that one of the solder pad, used to mount transistor A968B(collector) pin, had come off, and the another solder pad(Preamp Positive), was becoming a bit loose.
upload_2018-1-16_11-58-18.png
You can see the removed pad in above picture, i have temporarily soldered the transistor lead to the solder trace in PCB, i might use a small bit of wire, and then connect the two collectors together, to be on the safer side.

upload_2018-1-16_11-59-35.png

The Pad on extreme left has become a bit loose, i believe half of it still clings to the adhesive on the PCB, can you guys recommend a Epoxy , that i can use to stick this pad back.
Any inputs will be appreciated.

I immediately stopped the work, and inspected the boards for further damage.

The rubber adhesive on this board, in certain places, are quite weak, and i should have dosed, it with IPA after the solder work and not before, IPA also loosens up the solder pads easily if too much heat was applied, earlier, it is safe to use a lesser or very very minimal dose of IPA on areas on the PCB, where heat marks are present.

Eyelets, are a good way to go forward, but i do not have the tools required for using the eyelets, i will in near future when i have necessary funds, will do a proper work on the pads removed(count:1) /loosened(count:1) in above picture. !

Can you guys recommend/suggest me a way to fix the other pad that is a bit loose i do not want to go ahead with soldering on the loosened pad, as this might just make things worse.

Awaiting inputs from the Guru's of the forum.
Thanks in advance for all help and support rendered. !
 
Update 2:
While i wait for the fish spine beads to arrive, i decided to inspect the other boards.
Removed board F-3451 L-Eq:
F-3451 Front.jpg
Never judge a book by it's cover, let's look at the rear of the board.
L-EQ F-3451 Rear.jpg
Probably some one gave the amp a dose of Coke, there were spots on the PCB, where the contacts were all full of grime and dust, trust me the board was in a worse condition earlier, the above is after 1 round of cleaning.

Round 2 Cleaning:
Round 2.jpg

Round 3 Cleaning:
Round 3.jpg
These amps deserve a lot of respect !
I was literally disgusted and saddened when i removed and inspected the under side of the PCB.


Parital Re-Cap(the work will be over once the potentiometers are replaced):
I believe a bipolar mod can be done here, but i just replaced the electrolytics for now.
Re-capped.jpg
 
Nice clean work. Looks really good!

giphy.gif
 
Update 3:

F-3449(MC Left Channel) and F-3450(MC Right Channel).
I had some free time and decided to inspect these boards, these boards were a pain to look, all the flux had oxidised and oozed over the entire PCB Sections.
The flux was piled up in certain areas, and had to be cleaned, i did not restore these boards, but probably a better to post the details here.
I have also have a reference of the boards in stock condition, it was such a painful thing to watch these boards in this condition, i decided to clean them and put them back.
Due to lack of necessary spares, i did not restore them, but i would neither power on the amp in the condition the boards were.

F-3450 (Current Condition)
1. IMG-5473.JPG


F-3450 ( A closer image )
2. IMG-5474.JPG

F-3450 After treatment:
3. F-3450 Cleaned.JPG

F-3449 and F-3450
4. Both Board IMG-5477.JPG


F-3449 After Treatment:
5. F-3449 Cleaned IMG-5480.JPG

There's a small dab of cotton in above pic, i have it removed. !
 
Update 4: Driver Board

The older version of the driver board which is used in my X11, has an extra resistor in the rear(pictures below).
This resistor is supposed to fit in slot R28(100 ohm 5% Carbon Film), on the driver board, but the slots are sealed.

Slots for R28, are boxed in blue in the below picture:
R28 empty slot.jpg
Do not add the resistor R28, in the slot, if you use an older version of the driver board, move it back to the rear just like how it was earlier.

The traces, are cut in the older versions of the driver board, compared to the newer version:
Cut Traces.jpg
The newer version of the driver boards, have the places boxed in yellow (traces joined together), and not cut like i do have in mine.

Basically one could join the traces, and then put the resistor back in slot R28, recommended, but i feel it's quite risky, and have left mine as it is .
So yes, there will be one empty slot in the driver board with no resistor in it, for the older version of the driver boards.

Moving on to progress, yes the driver board is done !.
Finished Driver Board.jpg

Wima Bypass Capacitors:
Wima Bypass.jpg

Sad to say, but i could not use the ceramic, spacers or beeds, as the height was insufficient.

Mounted the potentiometers, and did a quick test to see for shorts of components, when adjusting bias.
The potentiometer has been slightly tilted to avoid shorting the small transistor beside it .
The rear of the potentiometer hits the cylindrical Polystyrene capacitor, but i have adjusted it to the max possible extent to have a good spacing.
Bourns Trim Pots.jpg
I used a slightly larger flat head screwdriver, to check for issues for short or touching of other components during adjustment, i do have a smaller one to get this done.
A ceramic screw driver is best suited for this purpose, when the pots are replaced with Bourns multi turn, the metal ones are just too risky as you can see in above picture.
The older stock pots, are quite easy and you an easily work your way through without touching/harming other components, this ones a bit tricky.

The driver board has been attached to the heat sink mount, the two bases wires connected.
And yes of course, the power supply blue and red wires connected and soldered onto the driver board.
The next step was to add, the extra components on the driver board back, pictured below.
Driver Board Rear.jpg

The On Semi MJ21193G and MJ21194G transistors mounted on the heat sink with Kapton Insulators.
Heat Sink On Semi MJ.jpg

The work is not over yet, and i am still not ready for mounting the driver board back in the amplifier.

To-Do Checklist:

1. Check for shorts across traces in driver board and check trace continuity in places where solder rework is done
2. Re-Check component values, resistors used for replacement
3. Re-Check heat sink assembly to driver board connection
4. Re-Check transistor polarity mounted on the heat sinks
5. Make sure potentiometers are adjusted to null/zero position (i don't want to start the amp and have it run at max bias)

I am so eager to put it back and test but no will only do it after cross verifiying multiple times.

I will work on these points over the weekdays, and probably mount it the coming weekend.
Thanks all for guiding/supporting me in this build.!
 
Last edited:
Preliminary checks are completed, since i had some more time this weekend, i decided to check the small signal transistors.
And thankful to Ron, for helping me with the Bias Transistors, and relays, i did see his commendable work in another thread, where he has provided the instructions to re-wire the Omron G5-V2 relays.

I also want to know if the rust in copper chasis of the AU-X11/AU-X1, can be removed, is there a easy method to do it ?
I have rubbing it for couple of hours, and my elbows started aching, any easy method to go about this ?

The driver board will mostly be put back the coming weekend, since i decided to spend more time on re-wiring checks and also checking the small signal transistors i wasn't able to do it this weekend.
 
Meanwhile, i managed to spend some time comparing output transistors.
The below transistors were compared with the original PNP outputs transistors NMA1718.
1. MJ21193J
2. 2ST2121
3. NMA1718

Just a note of caution about your conclusions regarding matching o/p transistors - the collector currents you are measuring at are far too low to make reasonable judgements. The hFE on many power transistors varies enormously with collector current, and you really need to be looking at currents in the 0.5 - 5 amp range.

If you look at the curves attached for MJ21193, ST2121, and NJW0302 (I know not a TO-3 but included to illustrate the point) you will see that the gain of the MJ21193 droops at a lower current than the ST, while the NJW is much more linear. The effect is that if hFE droops with current the feedback available diminishes at increased power, and so the distortion rises. Another parameter is fT (more relevantly switching time but data on fT is more often available), which varies even more dramatically with collector current. Faster switching 'generally' means improved cross-over distortion.

And I can heartily endorse using metal film resistors to replace carbon composition having done that to my AU-D11 II last September. Yes it will reduce warmth, but the improvement in clarity is enormous. Up to individual taste I guess :cool:
 

Attachments

  • MJ21193.jpg
    MJ21193.jpg
    112.6 KB · Views: 10
  • 2st2121.jpg
    2st2121.jpg
    103.2 KB · Views: 10
  • NJW0302.jpg
    NJW0302.jpg
    120 KB · Views: 11
I also want to know if the rust in copper chasis of the AU-X11/AU-X1, can be removed, is there a easy method to do it ?
I have rubbing it for couple of hours, and my elbows started aching, any easy method to go about this ?
As a chemist, I can offer some ideas. The chassis appears to be iron electroplated with copper. If you are carefull and confident, you may try to test-clean a small spot by gently rubbing the corroded zone with a q-tip soaked in a rust remover. Rust-Oleum has one but it is quite strong and it needs to be rinsed with water after applying it. There are products in some car parts shops under the phosphate rust converters. They convert the iron oxide in iron phosphate which inhibits further corrosion. Chloride containing mixtures (muriatic = hydrochloric acid) dissolve the iron oxide and copper oxide but do not dissolve copper. However, they are to be avoided because they will promote corrosion and make the problems worse. If you use these try locally and as a last resource. But they must be washed very carefully with deionized water. Whatever you do please try first on a non-visible little spot and observe the results.
Actually, if you want this thing done fully correctly, you need to disassemble the whole chassis, strip the oxide with a dilute acid and then electroplate it again with copper (from a solution of copper sulfate in dilute sulfuric acid). It is quite involved but it will look like new.
Copper spontaneously deposits onto iron from copper sulfate solutions. But the coating does not adhere very strongly. Try dipping a clean nail in a solution of copper sulfate with dilute sulfuric acid and you will get a copper-plated nail. May be good enough for touch-up using a q-tip saturated in the copper sulfate solution.
 
Update 5:
Long story short( the amp is up and running with the restored driver board).

Hassles And Niggles:
I have hooked up the driver board, re-wired everything and powered on the amplifier.
The DBT started to glow, I had to wait to 3 seconds, and see if the glow starts to fade away, but then no luck.
The power LED turned on for a second and then went off completely, the amp refuses to power on. !
The bias was reset to zero in driver board, transistor orientations are correct and the re-wiring done on the driver board was also correct, including all solder and wiring connections.
In fact I re-checked with the already existing untouched driver board, for any changes to wiring or mistakes which i would have done, to my dissapointment there wasn't a mistake which i could identify .

I was patient and went through the following diagnosis. !
Powered the amp again with DBT, the power LED would not glow, turned it off.
1. Checked fuses all of them were perfect, removed them and closely inspected all the fuses, no issues with them also did a quick continuity test with the multi meter, and the connections were perfect, so i could rule out the fuses for now.
The fear slowly started to build inside me, since the fuses were all fine and for some reason the amp would not turn on at all.
Check switched power supply output and unswitched outputs, ruled out power switch, as i had 100v output at the switched supply with power switch on and 0v at power switch off.

2. Decided to check the power supply board for any issues, and yes i found a open soft start resistor, even though the resistor looks good and there were no burn marks on it.
IMG-6179.JPG
IMG-6180.JPG
I wanted to analyze the reason for why the soft start resistor would fail just like that, it can't happen all of a sudden, few reasons i could think of was, an initial high ampere surge, or a short to ground some where.

In order to rule this out, i replaced the soft start with a temporary resistor which i have the same value but a slightly higher wattage.
I powered on the amp and the power led comes,on but the DBT still glows.
Turned it off immediately, as i knew it might cause some damage if turned on for a longer period of time.
All i could confirm was that there is a short some where in the amplifier.

3. I decided to cut off supply to the output transistors, hence i removed the 10 Ampere Fuse(pF1) it supplies the AC voltage to the primary winding of the cylindrical transformer.
Now when i started the amp the power led would come on, the DBT glows for 2 seconds and slowly starts fading away, voila !

4. With the supply to output transistors cut completely, there were few reasons i could think of , either a short in the secondary windings, or a problem with the wires leading from the power supply board, to rail supply capacitors.
Removed all primary and secondary winding's of the cylindrical transformer and did a quick resistance check, it turned out to be ok.

5. I then traced the wires of the secondary winding output from the power supply board to the driver boards and look what i found, this blue wire supplies power to the 10,000uf rail supply capacitors.
unnamed.jpg
I cannot make out the reason for why this wire has been damaged and was slightly touching the center chassis ground, with part of its insulation cut off, it was weird and absurd, as i believe no work was done on the protector board earlier and this wire comes from the protector board, hard to conclude, as this wire is left intact when you remove or install the driver board back. And there are no ways this wire could or would have been damaged, during driver board removal or re-mounting procedure.

6. I have re-taped the wire, in the minimalistic gap that i could manage to work on , i would replace it later on but i wanted to see if everything else is ok .

7. Powered on the amplifier again, there was no output from restored channel, but it was not in protection mode either, the output would work when i use the X11 as a external power amp, narrowed it down to the power supply board of the preamp which i failed to plug in the molex connector for the right channel alone.

There is output from the right channel, now and i have the amplifier running for couple of hours. !
The amplifier has been re-biased and DC Offset calibrated to 0mV, the multi turn potentiometers are a real boon and am extremely satisfied with the results.
 
Unfortunately, the head phone output in the X11, is not working, but the speakers outputs do work.
I regret checking the amplifier with headphones before working on the rebuild.

I checked the 220ohm 5 watt cement resistors and they are in spec, i then checked the reed relay and the center two pins have the preamp output, when i vary the preamp knob the center two pins(boxed in yellow) carry the varied voltage, whereas i have no outputs at the other end of the reed relay(boxed in red).
unnamed (1).jpg
The other two pins on the sides of the yellow colour boxed area carry a voltage of around 22volts, so the relay should be energized but is not engaging properly ?
Any pointers to other things that i should also check ?

Can any one help me confirm if i going in the right direction, and also confirm that the reed relay is not working as how it is opposed to be .

I would not have made this far without the help of you guys,and am extremely thankful to fellow forum members.

I will post my listening impressions after giving it from break in time, but my initial impressions with the On Semi MJ21193's and MJ21194's are that they sound a bit warm and looks like this highs are a bit rolled off compared to the other channel running the NMA1718 and NMC1718. The On Semi's/restored board seem to be more transparent, and changes to source quality is so easily audible, i tried it with two sources one was a direct headphone out from my laptop, and the other was the output from Musical Fidelity M1A Dac, the source quality change was so clearly audible in the restored channel, the low end response is a also more precise. I am guessing it sounds more cleaner that the other untouched channel.

There are too many new components in the driver board, and i have decided to give it some time to break in before i start listening or comparing them again.
 
Make sure the speaker switch is clean, then check the operation of TR606 on the F-2781 Power Supply & Protection PCB. (monitor its collector voltage when you operate the speaker select switch).

(Circuit references above are from the AU-X1 SM - I don't have the AU-X11 SM).
 
Last edited:
Re--- that nicked blue wire. I have seen that a couple of times over the years. No way to know if it was done when the amp was first assembled or later on in life, normally not a problem unless wires are moved about during repair. Usually caused by an errant wave of a hot soldering iron, nicely sorted out.
Curious as to what you used to wash down those circuit boards as the results are spectacular.
-Lee
 
Re--- that nicked blue wire. I have seen that a couple of times over the years. No way to know if it was done when the amp was first assembled or later on in life, normally not a problem unless wires are moved about during repair. Usually caused by an errant wave of a hot soldering iron, nicely sorted out.
Curious as to what you used to wash down those circuit boards as the results are spectacular.
-Lee

I just use regular dose of ipa for cleaning , clean them a couple of times . The other boards in the AU-X11 are in pretty good shape apart from the boards exposed to the sides of the wooden panels .
Regarding the nicked blue wire I saw excess solder build up in the chassis center ground located in the middle of the rail supply capacitors , the excess solder build up was in the shape of a spike probably it got nicked due to that . I have also removed the excess solder .
 
Make sure the speaker switch is clean, then check the operation of TR606 on the F-2781 Power Supply & Protection PCB. (monitor its collector voltage when you operate the speaker select switch).

(Circuit references above are from the AU-X1 SM - I don't have the AU-X11 SM).
I have 22volts going into the reed relay, i believe the voltage coming to the reed relay is from the collector of TR606(as per AU-X1 schematics).
To be clear, 22volts in two pins of the reed relay( one of the pins is ground), and the other 2 pins have the varied voltage, when i adjust the preamp/gain potentiometers.
The last two pins in the end of the reed relay have no output at all, maybe the reed relay is not being engergised with a 22volt supply, i am not sure to be frank.
But i will again test it from the transistors terminals, and keep you posted, with varied voltage results with different speaker selector changed to off, A,B and A+B.

Thanks Hyperion.!
 
Make sure the speaker switch is clean, then check the operation of TR606 on the F-2781 Power Supply & Protection PCB. (monitor its collector voltage when you operate the speaker select switch).

(Circuit references above are from the AU-X1 SM - I don't have the AU-X11 SM).

I have just checked this now , and the voltage across collector is 23 volts no matter what speaker selector i choose, it remains 23volts for Speakers set to Off/A/B or A+B.
I see the same voltage at the reed relay too, and the voltages across the relay are 23volts for every combination of speaker selector.
 
I just realised the speaker select switch should be at position 1 = speakers off, which doesn't use any of its contacts. :)

Both sets of speaker terminals work ok ? - and you get a 'flashing' protector light - and then 'solid' yes?

And I presume this is off DBT - on full mains?
 
I just realised the speaker select switch should be at position 1 = speakers off, which doesn't use any of its contacts. :)

Both sets of speaker terminals work ok ? - and you get a 'flashing' protector light - and then 'solid' yes?

I have set the speaker selector to off and still no output in the headphone jack .
Yes I get a flashing protector to be precise 6 blinks and then a solid light .
The A/B/A+B work fine with no problems, but then when I set the speaker selector to off, the headphone output is absent .
 
OK, well I am working from an AU-X1 schematic, so I'm hoping there are no differences with the AU-X11 here, but from what you say it appears the relay is faulty.
 
Back
Top Bottom