Sansui AU-111 rebuild

Well breaking with my tradition of chronological threads, here is a sneak peak at the final result. The sides are very much hammer finish silver-grey in person. The hand holds were not something I wanted to attempt to make myself. I fold that aluminium myself and it took some planning to get thing right in terms of fit. I toyed with more holes, but there is method to the position and airflow so I didn't play with this. i will post more details later, when I get through the electronics in this thread.

7diPhqc.jpg
 
Very nicely done. If I didn't own one I would assume it was stock. Particularly impressive that you folded it yourself!

Thanks. i was really trying to keep it original as possible since it is such an iconic amp which is becoming rarer. I thought about attempting to recreate the hand holds but it was out of the question without investing in a more serious metal sheet break (folding device), as the thickness of the metal I used is not really sufficient to pick the amp up by... that unit is heavy.
 
A thing of beauty. I've stared at mine for hours.DSC01929.JPG

With the bottom cover removed, what initially looks like a rats nest, becomes a thing of beauty if you stare at it long enough.
This is a 'before' picture, taken over a year ago. I've hauled this thing around for 35+ years -- time to fix it -- especially after I've read so many good things about it. I had never heard its full glory.
Before repair it had very noisy pots and switches and no bass on the right channel. There's no evidence of any prior service -- all of the original soldering inspectors marks (red or blue ink) on the solder joints are intact -- they disappear if the joint is resoldered.
I have given the inspectors names -- Inspector Red and Inspector Blue.
(Note: A very long time ago, during breaks from school, I worked in a factory on Long Island, building power supplies. Solderers dressed their connections out and placed components so that inspectors did not have to move stuff around in order to check them. -- Explains the way this chassis is wired -- all of the components and connections are visible without moving stuff around. Also, in our factory, Inspectors had X-ray vision and could see through lead/tin solder and reject a connection for broken strands in a wire even though it was not visible to us mortals.)
1-DSC01389.JPG

When I had rescued the amp from the curb I was with a co-worker, who insisted that I stop and for this "big black box with silver knobs all over it". I was an 'on the road' service tech at the time & if I stopped for every pile of junk I passed I'd never get any work done. So, heavy -- it even has handles and wheels...
Too heavy for me to carry it like it was a case of Bud long necks. (How do you spell sciatica?)
Where there's a will, there's a way... Here's an extra handle -- I only used one on the variac-isolation transformer-current limiter thing that I built. I probably should have it welded...
DSC02039.JPG

Ok -- all kidding aside(maybe) I've replaced all of the oil filled caps in the power supply and bottom of the amp - except the 2 ceramic ones that protect the power switch contacts(wired to the outlets on the back panel) and the one on the voltage selector that , I think, keeps the fuse from blowing due to transformer remanence during power up. ( or maybe it keeps my curtains from catching fire when I yank the power cord out of the wall socket.) I think there were 49 caps on the bottom, four big ones on top, 10 more to go on the front panel.
DSC02017.JPG

After the power supply caps, I wasn't sure whether I should troubleshoot and find the problem, or just go for broke and start replacing caps -- I can't find my scope, haven't seen it since before my last move. I knew the tubes were good. I had isolated the problem to the pre-amp. I changed the 4 power tube coupling caps first, they were all no good - One was double the value it should have been, one was open, and two measured at a couple hundred picofarads. Surprised that a had good noise from the left and anything at all from the right.
So, from there I started with the last 12ax7 and worked backwards replacing caps one at a time, then power up and test, and then replace its pair and test again. Each time verifying that it did not get worse. Changing C56 resolved the issue with the right channel, initially sounding edgy and raspy -- I played this --
-- and the miracles began -- sounded horrible at first, but after a few minutes I was in heaven.
Sitting curbside for this music should punch you in the chest, but for enjoying it on youtube - hey, the clowns with the basses are in the fifth or sixth row - need the loudness and presence switches turned on, the bass filter off, bass tone switch set @ 500 c/s, treble @ 2.5k c/s and tone controls at 1 o'clock. I'm driving a set of Wharfedale W70E's.
I think my bass filter switch(Sw6) is wired upside down -- no marks from Inspector Blue or Inspector Red on the front panel. Found this vid on youtube --
-- He goes through the switch functions. At 1:36 he flips the bass filter switch on, but the filter goes off, then at 1:41 he turns the switch off, but the filter is on. Mine was like that, seemed wrong, so I stared at the schematic for too long and then changed it, so it is now consistent with the action of the treble filter switch.
I continued to work backwards through the pre-amp, and then the pre/middle amp and bias sections. Powering up after each pair of caps to make sure I didn't screw it up, each time playing the same music. Lots of improvement in volume and quality as I went forward. I've let some of the caps age in for several hours --same music, and this one --
-- with a bit more bass reverberating off the buildings - do my caps need a workout?? Not sure if they'll continue to mellow(gradually??)
I'd probably not want to use the loudness or presence features at all for properly recorded music -- what those switch settings do, seems excessive unless you are playing your 78rpm blues records and need to pump 'em up.
I'm done with the bottom for a while, so i put the bottom cover on. Who knew there were holes to adjust the hum pots?DSC02014.JPG

I powered it up -- all is well -- remarkable that I can hear no 'on' noise -- you know that presence from the speakers that lets you know the tubes have warmed up -- this amp is absolutely quiet - if there is no input, there is absolutely no noise at all.
Same music, same switch positions & things started to mess up -- the 2 watt resistors in the middle section moved when they got hot and touched bottom, intermittently. A few scary moments, but easily resolved -- and let that be a lesson to me. (Wires get longer when heated)
There are some clear and present dangers in this amp - it clearly would not have received UL approval, even by the standards of its day. The 2 watt resistors in the power section move when'd they get hot and can rub against each other -- potential smoke release? My resistors test ok but the glaze has deteriorated to the point where the outside surface is conductive and if they touch, bad things could happen.
Also, insulation cutting, abrasion & breakdown is another clear danger. Wiring in the amp is routed through holes and around corners in the sheet metal without grommets or anything to prevent the sharp metal from abrading through the insulation.
The wiring should be laced into a wiring harness they way Johan P. did it or some other means of preventing strain, abrasion or cutting of the insulation.
1-DSC01389-001.JPGAU111_005.jpg

OK -- enough for now -- I have a thumb thing from screwdrivering too many screws -- must rest now... more thoughts later.
 
A thing of beauty. I've stared at mine for hours.View attachment 1166139

With the bottom cover removed, what initially looks like a rats nest, becomes a thing of beauty if you stare at it long enough.
This is a 'before' picture, taken over a year ago. I've hauled this thing around for 35+ years -- time to fix it -- especially after I've read so many good things about it. I had never heard its full glory.
Before repair it had very noisy pots and switches and no bass on the right channel. There's no evidence of any prior service -- all of the original soldering inspectors marks (red or blue ink) on the solder joints are intact -- they disappear if the joint is resoldered.
I have given the inspectors names -- Inspector Red and Inspector Blue.
(Note: A very long time ago, during breaks from school, I worked in a factory on Long Island, building power supplies. Solderers dressed their connections out and placed components so that inspectors did not have to move stuff around in order to check them. -- Explains the way this chassis is wired -- all of the components and connections are visible without moving stuff around. Also, in our factory, Inspectors had X-ray vision and could see through lead/tin solder and reject a connection for broken strands in a wire even though it was not visible to us mortals.)
View attachment 1166147

When I had rescued the amp from the curb I was with a co-worker, who insisted that I stop and for this "big black box with silver knobs all over it". I was an 'on the road' service tech at the time & if I stopped for every pile of junk I passed I'd never get any work done. So, heavy -- it even has handles and wheels...
Too heavy for me to carry it like it was a case of Bud long necks. (How do you spell sciatica?)
Where there's a will, there's a way... Here's an extra handle -- I only used one on the variac-isolation transformer-current limiter thing that I built. I probably should have it welded...
View attachment 1166169

Ok -- all kidding aside(maybe) I've replaced all of the oil filled caps in the power supply and bottom of the amp - except the 2 ceramic ones that protect the power switch contacts(wired to the outlets on the back panel) and the one on the voltage selector that , I think, keeps the fuse from blowing due to transformer remanence during power up. ( or maybe it keeps my curtains from catching fire when I yank the power cord out of the wall socket.) I think there were 49 caps on the bottom, four big ones on top, 10 more to go on the front panel.
View attachment 1166173

After the power supply caps, I wasn't sure whether I should troubleshoot and find the problem, or just go for broke and start replacing caps -- I can't find my scope, haven't seen it since before my last move. I knew the tubes were good. I had isolated the problem to the pre-amp. I changed the 4 power tube coupling caps first, they were all no good - One was double the value it should have been, one was open, and two measured at a couple hundred picofarads. Surprised that a had good noise from the left and anything at all from the right.
So, from there I started with the last 12ax7 and worked backwards replacing caps one at a time, then power up and test, and then replace its pair and test again. Each time verifying that it did not get worse. Changing C56 resolved the issue with the right channel, initially sounding edgy and raspy -- I played this --
-- and the miracles began -- sounded horrible at first, but after a few minutes I was in heaven.
Sitting curbside for this music should punch you in the chest, but for enjoying it on youtube - hey, the clowns with the basses are in the fifth or sixth row - need the loudness and presence switches turned on, the bass filter off, bass tone switch set @ 500 c/s, treble @ 2.5k c/s and tone controls at 1 o'clock. I'm driving a set of Wharfedale W70E's.
I think my bass filter switch(Sw6) is wired upside down -- no marks from Inspector Blue or Inspector Red on the front panel. Found this vid on youtube --
-- He goes through the switch functions. At 1:36 he flips the bass filter switch on, but the filter goes off, then at 1:41 he turns the switch off, but the filter is on. Mine was like that, seemed wrong, so I stared at the schematic for too long and then changed it, so it is now consistent with the action of the treble filter switch.
I continued to work backwards through the pre-amp, and then the pre/middle amp and bias sections. Powering up after each pair of caps to make sure I didn't screw it up, each time playing the same music. Lots of improvement in volume and quality as I went forward. I've let some of the caps age in for several hours --same music, and this one --
-- with a bit more bass reverberating off the buildings - do my caps need a workout?? Not sure if they'll continue to mellow(gradually??)
I'd probably not want to use the loudness or presence features at all for properly recorded music -- what those switch settings do, seems excessive unless you are playing your 78rpm blues records and need to pump 'em up.
I'm done with the bottom for a while, so i put the bottom cover on. Who knew there were holes to adjust the hum pots?View attachment 1166294

I powered it up -- all is well -- remarkable that I can hear no 'on' noise -- you know that presence from the speakers that lets you know the tubes have warmed up -- this amp is absolutely quiet - if there is no input, there is absolutely no noise at all.
Same music, same switch positions & things started to mess up -- the 2 watt resistors in the middle section moved when they got hot and touched bottom, intermittently. A few scary moments, but easily resolved -- and let that be a lesson to me. (Wires get longer when heated)
There are some clear and present dangers in this amp - it clearly would not have received UL approval, even by the standards of its day. The 2 watt resistors in the power section move when'd they get hot and can rub against each other -- potential smoke release? My resistors test ok but the glaze has deteriorated to the point where the outside surface is conductive and if they touch, bad things could happen.
Also, insulation cutting, abrasion & breakdown is another clear danger. Wiring in the amp is routed through holes and around corners in the sheet metal without grommets or anything to prevent the sharp metal from abrading through the insulation.
The wiring should be laced into a wiring harness they way Johan P. did it or some other means of preventing strain, abrasion or cutting of the insulation.
View attachment 1166329View attachment 1166330

OK -- enough for now -- I have a thumb thing from screwdrivering too many screws -- must rest now... more thoughts later.

Thanks for the interesting story. Those coupling caps like C56 are common issues. What is lucky is that your failing output coupling caps didn't go short circuit :).

I agree there are some safety issues here, and I did a fair amount of rerouting as I did my recap. I spent a lot of time thinking about what was too close to the chassis or could move to the chassis... it's a bit scary in there. I repositioned quite a few things and also added some insulator in a couple of places. I also had the issue with the slightly cooked resistors in power supply with resistors losing there coating. I also found several drifted resistors in other sections....

In terms of dead caps the main issue I had was a leaky mica in preamp that was easy to find as I had a 'pop' when turning tone defeat on and off. I found those brown rectangular mica caps to be problematic in other amps from this era.

Will do a full write-up when i get time.
 
Happy to see that you start recap your tube amp.

I get one last year that I fully recap with Russian caps and orange drop. The sound is amazing !!! I will post some pics of my amp before and after recaping.

Do you know someone who is selling the big Knobs ? I miss two of them. I can't find remanufacture, I just redesigned and print it on a 3d printer and paint it in gray !!.
 
If these are not the 2 tone knob, there was a chap in South Africa re-manufactured it and was available for purchase. I think he made an announcement on this forum some years ago about this.


ps.. sent you a pM with the contact details.
 
Last edited:
Here is some pics of my AU-111 before I start the cleaning and recapping :





During my first test, One oil caps explosed you can see on the pics below. The sound was ok but not as expected.


So I start the complete recapping without doing intermediary test like Sniffer, I was too afraid that an other one explosed again.
 
Here is the caps I used for the recapping :

The orange drop and the electro :

The Russians caps (paper in oil and PTFE)

The 4 biggest green caps have been replaced by PTFE caps because there were to big to fit to the old one
 
Sorry but I realized I just don't have time to do my usual detail thread so I will give the summary in a couple of posts.

In the first post I said my unit was 2-pot, but I was confusing this with my AU-70 which was originally 2-pot. My AU-111 was 4-pot variety, but with some parts of circuit following 2-pot schematic.

First thing I did was replace the output coupling caps. And added 1ohm sensing resistors on output cathodes and checked bias.
oAfYrVb.jpg

Then fired her up slowly. Found she made music, nice music, with a slight balance issue and a pop. The pop on flipping one of the preamp switches was quickly traced to DC leaking through a brown rectangular mica cap.

A before shot:
Here is what went on underneath.
mKvX23S.jpg

OhOMV07.jpg

Here is what came from the phono board alone (lower left in above picture):
5uaGeXQ.jpg



There is also some work to do at the tone controls etc. Here is what came out of the front section of the amp near the faceplate.
fT7LtaV.jpg

Where I have the orange caps, they were all oil caps. You may spot the black silver mica caps... they were those brown rectangular mica caps before.
VvFlXe4.jpg


Volume pot was poorly balanced (something I find common in the older units) and I replaced it with same spec (250kOhm with centre loudness tap) from another unit. Before I did this the balance was never close to right, even after I fixed up all of preamp and checked/replaced tubes etc. In the pic below you can see the new multisection cap cans (off-the-shelf) and the two big brown caps are in the voltage doubler and I used extra beefy united chemi-con; you absolutely can not skimp on those voltage doubler caps.

eAqVupj.jpg

ixDJ4LQ.jpg
 
Last edited:
Now, if you don't scope carefully and do full frequency response analysis and different power you may never notice, but even after the volume pot is nicely balanced, your AU-111 will likely (a) not have balanced output and (b) have different frequency response on left and right channels.

This is because the preamp feedback loop doesn't use well-matched feedback resistors, but actually uses the treble pots. As we have discussed above, pots of that era were not that well matched and even small differences in value between left and right can make a big difference because this part (feedback loop) of the circuit is resonant. My feedback resistance on one channel was 112K and on other 116k. This little difference there was enough to have me chasing my tail for quite some time.

This is easily fixed with an additional resistor on one side channel which I placed up at the pot hidden away so it can't really be seen. Before this fix the balance was really mind boggling, varying with volume and frequency. Afterwards....as you would expect, at all levels and frequencies - see pic below
o7h8FeU.jpg


seriously, before finding this one I was going a bit crazy, bypassing this and jumpering that to rule things out:
TOgX2He.jpg
 
Last edited:
In the end pretty much all caps were replaced. My experience with old Sansui amps of this era is that electroytics are not the only concern and I've had most trouble from those brown rectangular mic caps in this AU-111, my AU-70 and also older Sansui amps.


KEY SUMMARY:
- 2 output tubes could not be biased appropriate. 1 Ohm cathode resistors added for safety and shouldn't affect sound (if you don't have a 4-wire ohmeter you might go for higher resistor values but you may start to affect sound).

- All replaced with JJ (after much research with regard to tubes that can take high plate voltage JJ is great in many tubes), except 6AQ8 which is RCA NOS.

- Did try to keep some of the original tubes which appeared okay but they all ended up having issues, even if just balance.

- I found ALL the big electroytics to be in good shape (all had higher value than spec, ESR was ok) but replaced anyway, including the multicap cans. I upgraded Capacitance were appropriate throughout and considered the role of each capacitor when choosing replacements - no skimping done here!

- the two big single caps that sit on top do the work in the voltage doubler and they must not be skimped on and should be reasonably well matched. Beefy chemi-cons were used.

- several oil caps were swollen, a couple were leaky.

- multiple brown rectangular mica caps were leaking and where the cause of multiple initial issues observed so I decided to replace all caps at that point. Similar story on my AU-70.

- some resistors had to be replaced in power supply as well as some plate resistors (this is common when leaky caps throw bias off and plate resistors run hot).

- a couple of phase splitter resistors had drifted high. This was obvious from the waveform shape and clipping behaviour. They needed to be replaced with quality matched components and this took some fiddling as it's really sensitive to value if you look at the output in scope and analyze distortion.

- My unit is a multivoltage unit, but beware, like my AU-70, these earlier units can run way to close to the limit on modern line voltage so I use a variac for all. - 120V in gives 520V at output plates.

- I set the 6L6GC to 32 mA @ line voltage of 121 V (~525 V at plate) (spec of tubes is 28 mA - 40mA at 525V)

- the screws for the two bottom feet need to be shorter. one of them reach through dangerously close to the power supply section. I replaced the screws with shorter ones.

- preamp feedback goes through half of the treble pots 250k VR3 and VR4. balancing by adding resistor is required for good L/R level and EQ match.

- volume pot VR7/8 (250k with centre loudness tap) can be very unbalanced (40% different). Replaced mine with AU-555A. Some other models have some setup.

- when operating preamp sites at 33C, OPTs at 38C and PS at 48C (ambient temperature 20C).

- I get about 42 W, both channels driven into 8 Ohm at 1kHz, up from 38 W before repairs. What is impressive is how well the power output holds from 20Hz - 20 kHz.
 
I was lucky, I haven't so much problem like you. I have red on a french forum that the guy had to change the mica caps because they caused noise in the amp section.

For the mica caps, is it important to keep the exact value ?

I have seen in you previous post that you change the transistor 2sc650. Which substitute to you used ?


Here is the pics after recapping :




This pics is not up to date. I change all the trimm pot because I wasn't able to set the correct bias. During my first test afer recaping, one tube became red. And then I directly changed the trimm pots.



After cleaning :


While running :

Like you I set the bias at 32mA for each tubes.




I keep the power tubes because there are good (aroung 68mA@250V), I change the 6AQ8 because it was too weak by a NOS telefunken and all the 12AX7 by NOS EI

With all my system :

 
3D printer for Knobs... Is the plastic strong enough to handle the grunt needed for function selector switch ?. Sent you a PM, found couple of Knobs - not perfect condition. Please also advise on the 3D printer used.. Intriguing..
 
3D printer for Knobs... Is the plastic strong enough to handle the grunt needed for function selector switch ?. Sent you a PM, found couple of Knobs - not perfect condition. Please also advise on the 3D printer used.. Intriguing..

Mano, forgive me for responding on behalf of someone else but here is my view as i've considered 3D printing for knobs. The part will definitely be strong enough, I've used it for bicycle parts. But the finish will not be satisfactory unless you have a lot of skill with finishing such products. I'm going to buy a 3D printer anyway for reproducing some parts, but for the metal knobs CNC is the way to go, at least until 3D printing of metal comes a long way.
 
I redesign the knob with freecad and I use a basic 3D printer : an ANET A6 customized.
The materiaI I use is PETG, it is very strong and easier to print than ABS.

I had make many tests before the knobs fit in the shaft.

Here is the knob printed with a black PETG

CmFoO2K.jpg


TgbOYZ5.jpg



Then I painted the knob in gray. I made a test with the painting I had, in order to be sure that the painting sticks to the PETG

Since now every thing is OK.

8IiQD2N.jpg


iIYCPgg.jpg


I also print the 2 rca socket to change the connector in PETG black
0J7rQzF.png
 
I redesign the knob with freecad and I use a basic 3D printer : an ANET A6 customized.
The materiaI I use is PETG, it is very strong and easier to print than ABS.

I had make many tests before the knobs fit in the shaft.

Here is the knob printed with a black PETG

CmFoO2K.jpg


TgbOYZ5.jpg



Then I painted the knob in gray. I made a test with the painting I had, in order to be sure that the painting sticks to the PETG

Since now every thing is OK.

8IiQD2N.jpg


iIYCPgg.jpg


I also print the 2 rca socket to change the connector in PETG black
0J7rQzF.png

Obviously not perfect but seriously, great effort! I think I will get a 3D printer when metal printing advances further. But maybe with some more experiments with surface finishing (solvents) and painting this plastic one could be quite good. well done sir.
 
3D printing seems to be the answer more and more... rather than using a jigsaw etc... on a piece of plastic.
 
I was lucky, I haven't so much problem like you. I have red on a french forum that the guy had to change the mica caps because they caused noise in the amp section.

For the mica caps, is it important to keep the exact value ?

I have seen in you previous post that you change the transistor 2sc650. Which substitute to you used ?

- Regarding the MICA caps, I recommend replacing ALL based on my experience with this era of Sansui. About capacitance value, the usual rules apply (some must stay the same as they affect tone/EQ), but some can be increased as they are just signal coupling caps (but be careful with coupling caps, as some are actually in feedback loops and changing value can change EQ and even stability). So it's case-by-case. As you can see, all caps went to film or silver-mica - no more electrolytic in the signal path. I might have used a couple of C0G ceremics on the back.

If unsure about any of that, stick to original values.

- headamp (aka phono) transistors: typical low-noise small signal transistor will do (dont use ZTX here, they seem to do better in slightly higher power jobs. I used KSC1815). as mentioned earlier, note that schematic shows a PNP (like the AU-70) but it's definitely NPN setup. Here are some pics of phono board before and after (note there are a few caps underneath which I also replaced).
GxGhKMq.jpg

pzZOklB.jpg


UdjiCIy.jpg
 
Last edited:
Back
Top Bottom