Sansui AU-666 repair and restore thread

stopkidding said:
Thanks @stereofun and @AnalogueBen. I bought one from amazon. I am waiting for the new fuses to arrive. In the meantime, I lubed the pots and switch contacts and took these pictures.

Just love how easily accessible all the boards are. Amazing design for serviceability...
Click to expand...
If it is NOT dirt, the 68 ohms resistor next to TR805 looks burnt on the fourth picture. Also, it looks if the 8,2 k carbon composite resistor on the top left was hot. This only happens, if the DC voltage over it is very high on average, which does not happen if the resistor is okay AND the amp not defective having its output close to minus DC voltage..

If the amp is defective, my first bet would be that circuit card (and output transistors connected to it) or it may have been repaired in the past?
 
Last edited:
smurfer77 said:
The filter gave me good results when removed, but the speaker protection removal is one you should consider further. People seem hate this protection setup from a sonic point of view, but I didn't notice any sonic difference with it in/out, so if doing over again I would actually give some serious thought to leaving it in. Basically what it does is, if you have sufficient DC on the outputs the DC switches a transistor on, opening up a easy way for the outputs to blow the speaker fuse and protect your speaker from DC faults.
Click to expand...
Indeed, there is no technical reason to be found the impact of the protection circuit to have sonic influence.
The only thing it does is shorts the speaker output, hopefully blowing the speaker fuse, to protect the speaker, as the amp itself is deemed to be defective already if the protection circuit energizes.

I removed the circuit in both my AU666 and AU888, but the idea was not that bad. It can save speakers, indeed.
 
gslikker said:
If it is NOT dirt, the 68 ohms resistor next to TR805 looks burnt on the fourth picture. Also, it looks if the 8,2 k carbon composite resistor on the top left was hot. This only happens, if the DC voltage over it is very high on average, which does not happen if the resistor is okay AND the amp not defective having its output close to minus DC voltage..

If the amp is defective, my first bet would be that circuit card (and output transistors connected to it) or it may have been repaired in the past?
Click to expand...

Yes, I noticed that too. It definitely is the resistor that's burnt or blown. Nice eye!!
 
So I replaced all the fuses in the amp and turned it on and it immediately blew F006 and F004 quick acting fuses....now what? :idea:

Should I start with the power supply and start replacing components?

I guess I should be building a dim bulb tester?
 
Last edited:
Hi,

F004/F006 are the + and - supply for the right channel. It would appear that the right channel output transistors are shorted, but it may be something else in the right channel driver causing the fuses to blow.

Before you go any further you need to build a dim bulb tester.

Cheers

John

Edit: beaten by Kev!!
 
Thanks @skippy124 and @kevzep - I will work on the driver boards, replace everything and give it another go. The parts are on order. While I was waiting, I wanted to get help on getting the diode replacements sorted. In one of the older threads a similar issue was also reported as a busted diode that took away most of one channel.

What would be suitable modern diode replacements on the various boards?
Driver Block: F-1159A
D801, D803 - SM-150-01
Possible replacements? 512-1N4148 (mouser)

SP Protection Block - F-1334
D901 - 10DC-1R
D902 - 10DC-1N
D903, D904 - SR1FM-2
Possible replacements? 863-1N4004RLG; 625-RGP25M-E3 (mouser)

Bridge Rectifier on PS Caps.
I am thinking about up-rating the filter caps to 6800uF 63V or 80V
D001 5B2
Possible Replacements? 512-GBPC2504? (mouser)

Other Positions (no idea where?)
D905, D906, D906, D908 - SV-05

And Finally, In the Main-Amp Current adjustment alignment procedures, the manual says you need to test between points B1/B2 and 27? Are these just the ends of the fuse holders F003 and F004?
Testprocedure.JPG
 
Last edited:
It's a pity that the outputs my be lost, but all is not.. I used on semi mj21193g and 94g to replace mine. I guess you're keen to find out their state and the component tester is still in the mail, so here's a quick way to test them.
Remove your driver boards, place a little mark so you know which is R and L when they go back in, just be careful with that small bias transistor on the heatsink. Now you can easily get to the outputs, take a pick for reference to location (I find reliable pics faster than going to my schematic), now remove them. Set your multimeter to diode beep or just resistance. If they are good you will get no beep and a normal ohm reading on any of the case or pins you probe. If they are bad you'll hear a beep and they will have virtually zero resistance.
Also check your trimpots were not fiddled with as they might of been wound fully clockwise and caused the blow perhaps.
  • Yes I used 1N4148 on driver block.
  • Can't verify regarding protection block as mine is removed.
  • Yep I'd give those 6800's a go, I was a little modest and used 4700's in my case.
  • I ordered an RS804 rectifier from mouser but have not gone to the trouble of fitting as some people just don't with filter caps of under 10000uf..
  • D905-8 are located somewhere on the chassis according to the service manual, can you let me know also when you find them, seems I missed them.
  • Yes you hook up the + lead to 27 and your - lead to the - side of the fuse holder as shown in the pic, this is how it's laid out in the schematic. Make sure you connect the leads AFTER powering on, high chance of blowing a fuse in ya meter.
Oh and lastly, pleeeease build a DBT!!!
Good luck :)
 
AnalogueBen said:
  • Yes you hook up the + lead to 27 and your - lead to the - side of the fuse holder as shown in the pic, this is how it's laid out in the schematic. Make sure you connect the leads AFTER powering on, high chance of blowing a fuse in ya meter.
Oh and lastly, pleeeease build a DBT!!!
Good luck :)
Click to expand...

So terminal 27 is the positive end of the fuse holder?

Yes yes building that DBT :naughty:
 
AnalogueBen said:
Remove your driver boards, place a little mark so you know which is R and L when they go back in, just be careful with that small bias transistor on the heatsink. Now you can easily get to the outputs, take a pick for reference to location (I find reliable pics faster than going to my schematic), now remove them. Set your multimeter to diode beep or just resistance. If they are good you will get no beep and a normal ohm reading on any of the case or pins you probe. If they are bad you'll hear a beep and they will have virtually zero resistance.
Click to expand...

Yep Right outputs NEC 2SA627 and 2SD188 are both shorted.

ivandezande said:
Measure bias on the emitter resistors, not the fuses. Much easier and more reliable.
Click to expand...
Found it! Thanks!!
 
Bugger... the best advice is to replace the whole lot now.
I guess the question is why though? I'm sorry but I dont have the expertise to delve deeper into the issue, but when your tester comes it's going to put you in a position to test all associated components to see if a problem is elsewhere also.
 
AnalogueBen said:
Bugger... the best advice is to replace the whole lot now.
Click to expand...
Do the outputs need to be matched? Since each channel has complementary pairs, how do you match them? are the On Semi MJ21194G/MJ21193G the recommended ones because they are the closest match to originals or a better spec? What are the alternates? Like the absolute best...
 
This is a common question but you won't buy better new than those. They can handle more than the originals but on the negative side have a slightly lower frequency rating, apparently this is unlikely to even be heard.
Otherwise you search for originals, expect to pay 2-3 times what you will for the OnSemi and hope that they are usable, it's really your call.
No, matching isn't a big deal.
 
You will do absolutely fine with the onsemi 211 series - no need to match, Infact whats nice about the 21193/94 pair is that they were made as a complementary set, so they are much closer in HFE and FT than many of the original pairs. If memory serves, the Eight Deluxe has a pair where one is 15mhz and the other 5mhz.... as for speed, they will be faster than the 2sc1079 that the Eight sports. I would consider them an upgrade in terms of reliability - soundwise I doubt anyone could tell a difference.

P.s if you want to adjust bias over the emitters then put into ohms law the desired mA and the emitter resistor value, this will give you a mV. On a quasi amp you adjust for that as you probe across one of the emitter resistors, but on a complementary amp you have to divide that mV number by 2, otherwise you will double your bias. If you rather not divide by 2, then you need to probe across both emitter resistors - that is probe each resistors leg that connects to the emitter. The emitters resistors in a complementary amp such as the 666 are in series, unlike in a quasi amp such as the Eight or 555a.
 
Last edited:
Transitor hfe matching the zetex has been challenging. I bought the DROK tester from Amazon. When testing the zetex, the hfe value changes from test to test. I was able to "match" hfe for least 2 pairs each of the ztx795a and ztx694b, but the values are off by 2-3. Is it still okay to use them? The Fairchilds match head-on. How close do the matched pairs have to be .

Also looking for confirmation for which transitor positions need to be match and those that don't.

F-1159 - Driver block
TR801 XA495G (C,D) ZTX795A
TR803 XA495G (C,D) ZTX795A
TR805 8002-1(B,C) "ZTX694B/KSC2690AYS"
TR807 8002-1(B,C) "ZTX694B/KSC2690AYS"
TR809 2SC281(C)
TR811 2SC708(A,B,C) KSA1220
TR813 2SA537(A,B,C) KSA1220

Also with matched pairs, is it better to use pairs with higher hfe?

Sorry for all these questions, this is a bit new for me.
 
For F-1279, planning to use these. Do these need to be matched as well?

transistors onboard replacement
TR701 2SC871R(E,F) KSC1845FTA
TR702 2SC871R(E,F) KSC1845FTA
TR703 2SC871R(F) KSC1845FTA
TR704 2SC871R(F) KSC1845FTA
TR705 2SC871R(E,F) KSC1845FTA
TR706 2SC871R(E,F) KSC1845FTA
 
HEAD AMP BLOCK
transistors onboard replacement
TR601 2SC871R(E,F) KSC1845FTA
TR602 2SC871R(E,F) KSC1845FTA
TR603 2SC871R(F) KSC1845FTA
TR604 2SC871R(F) KSC1845FTA
 
You must log in or register to reply here.

Similar threads

S
Sansui AU-666 - fixing shorted output
Replies
4
Views
516
MyHeresyIsMe
M
Z
(Slightly) Noisy and (Slightly) dull AU-666
Replies
0
Views
223
ziggydave
Z
mmann68
Sansui QRX 9001 Restore/Repair?
Replies
3
Views
772
mmann68
mmann68
Duffinator
Sansui Devil Integrated....The AU 666
Replies
6
Views
2K
onepixel
onepixel
smurfer77
sansui au-666 refurb advice + one channel noisy
2
Replies
33
Views
8K
smurfer77
smurfer77
Back
Top Bottom