Au-777d overheating/right channel distorted

darai22jp

Active Member
Hello from Japan.
Finished recapping an au-777d but running with a couple of issues...
First the output transistor in the left closest to power transformer gets VERY hot after just minutes of powering on and the right channel sounds low and distorted.... I am thinking about a damaged transistor, I checked the power transistors and they were fine, so I will keep checking the others. My concern is with the overheating, do you guys think is maybe the power supply or maybe something wrong with board 1170? The thing is that only one gets super hot....
Any input will be really appreciated.
Thanks
 
It can be a multitude of issues - but some of the 777D's suffer from failure prone and noisy 2SC458 transistors....all 16 of them. The other thing that I've had issues with in the past are grossly out of spec trimmers. If you have no or very little experience with these old amps, I really think you should have the amp looked at by a good technician. They are precious.

Check the right side fuse regarding the low distorted channel - some Sansui's will sound low and distorted when the fuse is toast. If it is a bad fuse, then the question is why it blew in the first place. But it could be a start.

As for the very hot output transistor, remove the left fuse, set your DMM to measure mA, put the probes across the fuse holder, now have a small flat head screwdriver ready on the left side (driver board) VR803, turn on the amp with no signal and see what mA reading you get - if if continuously climbs past 30mA, try turn the VR803 counter clockwise to see if that reduces the mA reading, if it doesn't help and the number keeps rising, shut off immediately.
Other than that - a close visual inspection of each board might yield clues for hotspots or "blackened" areas on the PCB, burnt resistors or bulging caps. A careful visual inspection is very important and can often lead straight to the culprit. Post pictures if you find anything suspicious.

If you can identify the signal transistors as being the 2SC458 - then rather than to de-solder each of them to test and put them back, you are better off just replacing the bunch with KSA1845FTA's
 
Thank you for the reply. Well, found that fuse 002 was blown... replaced the fuse, turned on and it was blown again...
Any clue were to look....?
 
Yes the fuse will keep blowing until the problem is fixed.

Joint this site and download the service manual: https://www.hifiengine.com/manual_library/sansui/au-777.shtml

Build yourself a dim bulb tester - instructions on this site on how to do that - its pretty simple.

Look at page 17 of the manual - it refers to the bias setup procedure. Also look at Stereofun's advice above.

You have a problem with bias setup. You need a DMM and a small screwdriver to attempt to reset the bias - do this in both channels using the procedure in the manual. If it were me, I would do this with the bulb tester connected in series with the amp to minimise risk if you have something else wrong.
 
It may not be as simple as a bias setup problem, you very likely have a problem in the output stage of the affected channel, possibly the output transistors, or driver transistors, or even something as seemingly insignificant as a broken or oxidised bias trimmer. You could start by removing and testing the output transistors for the 'hot' channel, if they test OK do the same to the driver transistors.

Note that problems like this can be very difficult to trace for someone without sufficient electronics knowledge. These power amp stages are almost always DC coupled, and if you have a failed component and thus a wrong DC voltage at one point, its effect will be widespread and difficult to pin down. The circuitry is quite finely balanced, so unless you find and replace all the failed components, when you power on, the whole thing just melts down again destroying the new parts that you have put in. A DBT can help with this as it automatically limits the current which would otherwise destroy the new parts, but it isn't foolproof and damage can still occur, so you have to know what you are looking for in terms of failed or stressed components.
 
Last edited:
Hello. Removed all the c458's and output transistors.. checked them with diode method... checked fine... replaced 4 ceramic capacitors in f1194 (c603,604,609 and 610 ) and no more blown fuses, right channel doesn't get hot anymore... just a bit warm... but now sound in both channels sounds bad, the left sounds a little better....And also can't set bias... removed fuses and connected my tester. I get around 18mA in both channels but doesn't change with the volumes at all!!!
ovbiously I need to replace transistors.... any other clue regarding bias setting issue or sound dstortion will be really appreciated

Thanks a lot
 
This inexpensive transistor tester is dirt cheap and will give you more information than just a diode test. Once I tested 458's with this gadget and it showed that a number of them were way out of spec regarding hfe, but worse, turn on voltage that should have been .65 - .7 V, were in some cases 1.2 - 1.4 V. I believe you are better off replacing the 16 x 2sc458's because they are failure prone and approaching 50 years of age, and because they can mess up your diagnostics if you deem them working when they are in fact out of spec and not performing in the circuit. I would also replace the VR 801/803 trimmers as they can be erratic or as in my 777a, completely out of spec. With the service manual open, resistors are quick to test without removing, at least this will be a good enough indicator. Also do check the fuses that they are indeed 2A, quick acting - many times you will find that somebody put in whatever they had around. I assume you have already performed a thorough visual inspection of boards and components ? - again I cant tell you how many times I wish I'd just "looked"......and "seen" that right away.

Normally the best approach is to find what is wrong first before changing out parts left and right, but in this case with some known (and deceptive) culprits you might be better off just ruling out these areas and then go from there. This only works of course if you are very careful replacing transistors as to not create another issue by not keeping check on pin-outs and traces.
 
SO sorry for the delay but I've been pretty busy recently. At last had some time to work in this amp.
I was able to get original c458's ! my local shop had them on stock still! Install them and found out that the transistor in the back is overheating again, actually almost just instantly!!! So I suspected something wrong with the power supply. After a quick inspection found that the 470 uf capacitor was deformed, so I replaced it but still overheating. If I try to check bias using the tester connected to the fuse holder I get a very faint reading and then goes to 0... please any suggestions were to look/ check?
Thanks
 
I was able to get original c458's
The reason we advised to pull and replace these with modern equivalents is because this transistor is a real trouble maker and shoud always be removed. Are these the "barn" shaped boxy transistors with a hitachi logo and a sloping front bevelled edge?

Now you have installed NOS ones, Im not too sure if that would change things or not - certainly in a few years, these trannies are likely to give you problems again.

More reliable replacements for the 458s which will take actual or potential unreliability out of the equation are the KSC1845, however pay attention to the pinouts.

Depending on how you address this transistor issue - you really shoud be using the service manual to test some basic voltages in the amp. I would start with the power supply to see if it putting out the correct DC voltages to various parts of the amp.
 
As slimecity says, it's time to get the schematic out and start checking voltages. I have a very nice high resolution copy I can send you through email.

Check your multimeter too. When you get zero mA reading, the meter likely has a bad 500mA fuse. It happens a lot !

As for the bias pots, they can fail wide open, causing rapid heating in that sides outputs.
On the VR803's you can measure max ohm balue on the two outer legs and compare the R/L side pots just to see if they are even (on mine, one side was grossly out of spec causing adjustment issues) - you can also measure ohm on the outer and center leg to see actual setting and compare the overheating side value to the "good" side value. Measure on the same to legs, one of the outer and then the center.

Many times overheating issues or popping fuses are due to failed resistors. As mentioned, check values on the main board and two driver boards. It's fairly quick to do.
 
Thank you everyone for the inputs.
I suspected bad resistors as well. Found after a visual check that some of them were chipped ( maybe of overheating) the ceramic that is covering them was gone in places, so I replaced those and also replaced some ceramic capacitors that were deteriorated. No more blown fuses, BUT still sound terrible !! now no sound is the coming from the right and distorted sound from the left...:wtf:
I replaced the transistors with new 2sc458 So it can'f be transistors unless something really bad happened when the fuses blew...
I was suggested by slmecity to use ksc1835 instead... the thing any Ksc, 2n, Bc,net or on semi is nearly impossible to find here in Japan.. as far as I know ksc1845 and 2sc1845 are almost identiical so no prob there...
With the overheating... I think I found the problem. Seems that the to-66 housing transistors were not making contact very well so after bending the legs a bit and reinstalled them no more overheating.
I think I have no choice but to upgrade the transistors with 2sa1845..,
Any other theory about why the sound is distorted would be very nice helpful
Thank you
 
I applaud your eagerness to replace items and get the unit working, but also be systematic - we still don't know what bias you have across the fuse holders and if the trimmers are responding (if you get Zero, check your meter fuse)

Just check these items off your list and let us know where you are.

1) clean all controls, especially rotary selectors for speakers and source - those can easily drop or distort a channel after years of oxidation. You should also clean the VR801/803 trimmers, but set them back as they were (refer to my prior post on measuring)

2) Visual inspection including verification of working and CORRECT fuses before start up.

3) Basic settings such as bias and centre voltage (1/2 supply at output caps) - are the trimmers responding ?

4) voltage at large cap, out the B lines, At the output caps (see schematic for targets - I have high res copy if you need)

Once all the low hanging fruits are picked and checked off, then it is easier to go deeper as necessary or narrow in on specific boards (here you can often compare sides or even swap L/R boards and see if the error follows) - do post some detailed pics as well, this will give an impression on the overall condition and someone might even spot a wrong pin-out or other oddity.....plus we love pictures ;-)
 
Seems that the to-66 housing transistors were not making contact very well
What do you mean by this? Were these not screwed down to the chassis properly? These transistors need the transistor case connected by a screw to the electrical connection on the other side of the metal chassis to work properly, the transistor case is the "collector" connection of the transistor, the two pins are emitter & base. Are the screws making good contact with the transistor body?
 
Good catch slimecity, yes those drivers and outputs needs to be tightened down, not super tight, but snug. A quick test to make sure that the mica insulators are holding up is to continuity test between device casing (collector) and the heat sink. A breached insulator would cause continuity and certainly blow fuses. Even moving around a heavy amp without the stability of bottom plate and casing might induce enough torque to cause a somewhat loose device bolt to wiggle enough to make contact ... and puuff, seen that in an Eight ;-(
 
Hi. The thing is that the last owner replaced the to-66 transistors with to220, but he removed the sockets and soldered the wires to the transistors. I was able to find suitable replacements for them, but I had to make my own sockets so I could fit the transistors. I am not very happy with the result so I look around and I was able to find period correct sockets and they are on the mail at the time.
I connected the 777-d as power amp to my trusty ca-606 pre amp and is the same. (No sound from the right channel) so I thought the problem was in the f-11831 board or in the driver board, so I switched the driver boards but still the same, checked for failed transistors in the f-11831 board but no change. Also switched the 2sd152's but no change. I decided to try the pre amp of the 777-d so I connected it to my au-7500... and is no sound coming from the right channel as well! This is not making any sense to me... what do you guys think?
 
and is no sound coming from the right channel as well! This is not making any sense to me... what do you guys think?
Seems like you now have a fault in the right channel pre-amp and power amp, possibly something to do with the power supply - like a missing rail voltage for example? - Rule 1:- First check all power supply voltages.

It seems like this 777d is a bit of a basket case from your comments, especially post #16.
 
Hello. Well all problems solved!!!! I checked all the wires solder points in the F1194 board and that solved the issue in the pre amp side. Replaced resistors in the f11831board and that solved the issue in the power amp side. Now I have beautiful sound coming out of this unit from both channels. BUT, and seems that there is always something with this unit... tried to set bias using my meter connected to the fuse holders but I get around 22ma in both channels and it doesn't matter what I do the number doesn't change!! The pots are new. Also tried putting an 8ohm resistor in the speaker terminals with the meter connected and also get about 20ma...any ideas what could be happening? If somebody could guide me in another way to set bias would be great. Thanks
 
..And you are turning VR803 on the driver boards.....not VR801 on the main board. Alternatively you can measure mV across white ceramic resistor R851 located on the driver boards. For a 30mA setting with a .5W resistance you are looking for 15mV. You need to probe across the resistor. (you can also use R853 if access are easier - same number) Please be very careful, you should only do this with probes that can clip unto the resistor leads so you are hands free.
 
Back
Top Bottom