Discussion in 'Exclusively Sansui' started by Hyperion, Dec 1, 2017.
Great job John - - what did you use as replacement bridge rectifiers?
I used 'KBL04-E4' rectifiers, their leads are slightly thicker than I realised, (so, for this reason were not ideal replacements), so I had to do some careful hole enlargement. Also I can't find SIP bridge rectifiers with the terminations in the same order as the originals i.e. "AC + - AC" - all I can find are "+ AC AC -" now - so 'lead crossing' was required as well. The reason for changing them was some reports that the originals (RB152) have somewhat worrying amounts of rectification noise on their output.
One removed heatsink assembly - note the transistor (left-front) remounted by me during initial fault finding of the unit because one of the screws was found to be arcing in its hole and shorting one of the supplies to ground, note two new screws fitted in the remounted transistor.
And now with the main body of the heatsink removed, to reveal the emitter resistors - and silicone grease on the mating surface of the transistor mounting plate.
I have just matched the new output transistors, they were very close anyway. I now have the best 2 pairs of each type set aside ready for mounting modification - adding gold plated pins etc, details of this will be shown in the following post(s).
And I have just run out of WD40 - so good for cleaning and removing old heatsink compound.
I just found out that the M3 16mm screws that are perfect for securing the TO-3 power transistors are too short by around 2mm for the 'Tab' end of the new OP transistors. They only just catch on the threaded contact strips in the TO-3 socket which isn't good enough. Can't get 18mm - ordered 20mm, I'll trim them if necessary.
Looks like care is needed on locating the emitter resistors to prevent any chance of shorting out against the heatsink, reminds me of the 919 in that regard.
Yes, there is actually plenty of room in there though.
Just a question what are those small things you use to cover the resistor leads(those mounted high) , i believe you have used multicomp metal fims as a replacement in above picture.
On a side note, i have sourced the M3 bolts for the X11, but i am not sure if the length will be sufficient after going through your restoration. The thread strength in the old screws are pretty weak, and i decided to replace them any way.
By the way nice work John !.
Keep it going !
The small white stand-offs are ceramic beads (fishspine beads) - useful for keeping hot components (one bead per leg) or bloody hot components (two beads) away from the PCB for a). Ventilation, and b). Reducing PCB damage if the worst happens.
The correct screws for the TO-3 transistors (on an X1 or X11 - and several others I am guessing) are M3 16mm, for TO-3P (i.e. TO-3 upgrades to TO-3P), this needs to be minimum M3 18mm - however, the screws I ordered which are M3 20mm turn out to be perfect for the BA-F1 as far as I can see, and should be OK for the others.
Here is one heatsink, upgraded with TO-3P transistors.
This took me some of the morning and most of the afternoon to complete from scratch, including cleaning. I should be able to do the second heatsink more quickly now, as I have worked out the technique needed to fabricate the transistors, which I will describe in a later post. I used sharpie to colour in the small voids (bare metal) in the anodising, seems like an original manufacturing defect. And lastly, keeping the heatsink compound looking neat is truly the devil's work!
Impressive job John!. Thanks for sharing the details. This thread has a lot of useful information beyond the BA-F1 model. I am looking forward to your description of how you adapted the pins of the TO-3P devices to the TO-3 socket may be I use your trick in the AU-D907 I have.
Please be patient - it's coming.
I began taking the pictures this afternoon for this, right after I found out that a previous 'tech' had installed two wrong value emitter resistors during a previous repair. This was found when I began work on the second of the two heatsink assemblies two x 0.27Ω used instead of 0.33Ω, these were tucked away (in the heatsink assembly), so not seen until now. - order placed, more waiting...
How to fabricate TO-3P transistors to fit into TO-3 sockets.
Here I have the 2 gold pins, a ring terminal, insulating bush, M3 washer, TO-3P Mica washer, Transistor, M3 16mm, & M3 20mm Stainless Steel screws.
Cut the E & B legs off the transistor (as shown) and shorten the C leg, use a small file to lightly round off the square ends of the E & B legs - for neatness when soldering.
Two gold pins shown, (left=uncut, right=cut) Cut the crimp buckets off the gold pins, use flush cutters (as shown) just leaving the insertion stop disk at the top of the pin. Use a small file to lightly smooth the top of the pin where it was cut. NB: Use safety glasses when cutting the pins, as the cut end of the gold pins really fly off when cut and could injure you.
Here is the cleaned heatsink, the sockets (underneath) have been removed, cleaned and then put back.
Insert the two gold pins into one of the sockets and tin the ends (as shown), the pins should be inserted just enough to grip in the socket. Cut a ~3mm strip off the bottom of the TO-3P mica washer so it doesn't interfere with the E & B legs, and position it as shown.
Position the transistor, insert the 20mm screw with washer, then tighten down the screw carefully making sure to keep the gold pins directly underneath the E & B tabs. When the screw is tight make sure the transistor and mica washer are positioned correctly, and also the gold pins - then solder the E & B tabs to the gold pins - leave a little blob of solder on the top of each tab to show that the new gold pin leg has been soldered and for neatness.
Now add the bush, and the ring terminal, slide the terminal onto the C leg and trim the leg as needed. Insert the M3 16mm screw, and tighten it up. Then solder the C leg to the terminal after making sure everything is lined up neatly.
This is one transistor complete, note that I have not added any heatsink compound yet as I think it advisable to complete all this work without getting the compound everywhere. I will complete all 4 transistors on this heatsink before removing them all and applying compound to complete the job.
Gold pins = TE Connectivity 1650155-1, Farnell P/No 2310222
Ring terminal = Molex 0.5-1.0mm M2.5, Farnell P/No 2295580 (I drilled it out to M3 )
TO-3P insulating set (x10) = Multicomp MK3305, Farnell P/No 520238
I hope this is helpful.
Excellent work! It is very helpful.Thanks for posting detailed pictures and the part numbers.
This should work for the X1 as well. I guess this solves the output transistor problem for those impossible-to-find high-speed outputs! Let's spread the word!!!!
Nice-beautiful work John. As always .
This has been known about for some time, it definitely isn't new. But I think it is the first time the creation of the modified transistors has been described in this level of detail.
I was going to say, full credit to Kale and others who have shown us this approach before, but John, thanks for showing extreme details (which many of us need!) - I've bookmarked this one! This may come in handy when I get around to my AU-X11
Plus the pin type and number part is priceless...
With the MJ TO-3 devices stock getting low, and very expensive NOS... I can see this will be a more common solution.
Very Nice and detailed how to.
Where were the Sanken's sourced from for future ref.. just in case..
Separate names with a comma.