Pioneer SX-1980 Regulator Modification

EchoWars

Hiding in Honduras
Sponsor
Everyone knows these get stupid hot. I suggested in another thread moving both the TO-220 pass transistors for the +/-80V supply to the big output transistor heatsinks.

So I did.

1980_PS_Full_Back_zpsa2688ab6.jpg~original


Closeup of one...

1980_Transistor_Heatsink_zps0aac8f6b.jpg


I used the MJE15032G and 33G for this experiment. I wanted to use the insulated MJF15030 and MJF15031, but no one has the '30. I have some ordered but delivery won't be till February or later. So, for insulation, Mouser has TO-220 shoulder washers and Bergquist thermal pads (or you can use mica and heatsink grease).

On the big 1980 heatsinks, there's some holes at the mounting location, but they are too large and too close to the edge of the heatsink. I drilled new holes in a location which allowed for proper mounting (#39 bit, 0.099") and used a standard 2.9mm x 13mm metric sheet metal screw and a 3mm flat washer with the shoulder washer to isolate the screw.

I wanted to be able to disconnect the transistors easily, so I used some Molex TO-220 sockets. These are seriously nice sockets and grip the transistor very well. Only complaint is their 'height', in that the leads of the transistor have to be bent up and then back to allow clearance (pic below). Used 20ga wire to wire 'em up, but the current is low and even 24ga would be fine.

Clearance from the TO-220 socket and the bottom panel is, well, more than tight. The bottom panel just misses the plastic connector by a mm or two, but all conductors are recessed so shorting is not an issue. But I may cover a small section of the bottom panel with a piece of insulating tape to make me feel better.

TO220_Lead_Bending_zps017a9448.jpg


Q203 (13.5V supply) and Q205 (8V and 5V) still get rather warm, but their heatsinks were changed to some 1" tall jobs that are much nicer than the originals. That, and with the 80V transistors located remotely, I can't think of a reason to leave the big heatsink in place. Q203 still gets damn hot, but it's an improvement and with the removal of the big heatsink airflow has to be better (poor still, I'm sure, but better than before). The original was held on with a screw from underneath the board, so I used a #4-40 x 1 3/8" machine screw and a few washers (on top and underneath) with a nut and lockwasher to secure the new heatsinks (bought locally, no part numbers).

Parts:
HS300-ND Digikey 25.4mm x 16mm TO-220 heatsink
532-7721-8PPS Mouser TO-220 shoulder washer (necessary if using collector tab TO-220's)
94997A275 McMaster-Carr 2.9mm x 13mm sheet metal screw #1 philips drive
93475A210 McMaster-Carr 3mm stainless flat washer
WM2550-ND Digikey Molex TO-220 socket.
22ga stranded hook-up wire (recommended)

The big output heatsinks show no perceptible rise in temp, and the two +/-80V pass transistors are cool as a cucumber.
 
Last edited:
I think I see why you mounted them to the bottom of the horizontal extension. A small angle drill would be need to bore a hole into the vertical section of the extension you chose. You could drill through both vertical extensions. This way a guy could use a drill press with the HS mounted securely to the bed. Then use the hole on the forward side of the HS extension. This would give more clearance by moving the transistor into the set a little more.

I like it. It really cleans up the PS board.
 
Well, the idea was to keep it simple and ensure that it was easily duplicated with basic hand tools and a minimum of special parts.

Yes, I like the look of that board much better with the big heatsink gone.

Next time I do this I may fill the existing hole with JB Weld and move the mounting hole out where I don't have to fiddle with the transistor leads.
 
The other two low voltage regulators actually get hotter than the two 80v on the duel HS. Finding another spot for them would complete the mod. There's a 1980 thread in progress where the 13.5v has failed. I wonder if this and the 8v regulator is the most frequent failures related to the PS. You more than most would know better.

I hoping to get the new larger HS' s installed this weekend but things are not moving that way.
 
As hot as the 13V reg gets, the most common failure is the +/-80V job, mainly through failed zeners or FET's. Once the 35V zener is replaced with a pair of 17V/18V ones and the FET's trashed for the current source modification the whole assembly is reasonably reliable. I've not had a 1980 come back with regulator issues after these mods.

BUT, heat is the enemy, so if we can get rid of the heat produced by the 80V pass transistors (which is significant) and use a much higher quality heatsink on the 13V and 8V pass transistors, I'd have to say we're moving in the right direction.

I wish there were more room to add additional heatsinking to the two remaining hot spots. But even as it is, in the above configuration, it's a pretty easy conclusion that things are much improved.
 
You are adding inspiration to my weekends plans. Looks like a limited sleep weekend!
 
Bottom cover? A millimeter or two between the plastic case of the TO-220 socket and the cover.
 
This is great. I may apply this info to fix my 13.5v problem, and i may relocate the 80v while im at it..
 
I see a lot of power supplies in this era of receivers that use linear regulators (or a generic transistor serving the same purpose as a series regulator along with some support components) to cut down voltage. In many cases these get so hot they desolder themselves from the board over time and cause the whole thing to shut down.

Have you given any thought to mods that make use of switching regulators, which generate far less heat for the power they regulate, or do you think they would generate too much electrical noise in close proximity to all the other circuitry? I have a few units that I've spent way too much time trying to move transistors around and upgrade heatsinks to handle the insane amounts of heat these generate, and I've been toying around on paper with ways of doing this, but haven't actually tried anything just yet.
 
Have you given any thought to mods that make use of switching regulators, which generate far less heat for the power they regulate, or do you think they would generate too much electrical noise in close proximity to all the other circuitry?
See first sentence of Post #3. Says it all.
 
This is great. I may apply this info to fix my 13.5v problem, and i may relocate the 80v while im at it..
I checked, for the hell of it. The pass transistor for the 13.5V supply has to drop about 10V at about 270mA. If you changed that 3.3 ohm resistor to about 18 ohms, looks like that would cut the power dissipation of the transistor by about 1/2. Would need a 2W resistor (3W preferred).

Edit: See post #20
 
Last edited:
Nice mod/upgrade to cool the unit down!!

On your 13.5V mod, (increase the V dropping R) just make sure that you have enough input/output differential (regulation) for a low line voltage, usually set at around (120V -10%) or ~108V.
 
In stock form the differential is about 10V. I think that's excessive. I have not done it, but was thinking of dropping it to 5V. And you are correct...need to verify with a variac that 5V (or any other chosen differential) would be sufficient.
 
Yes, 10V in/out is excessive, it can be brought down to take power away from the pass bjt. Puzzled why they chose such a high V, 26(24)VDC for the secondary winding tap? Will never really know why, only speculation.
 
Yes, 10V in/out is excessive, it can be brought down to take power away from the pass bjt. Puzzled why they chose such a high V, 26(24)VDC for the secondary winding tap?
Ugg...just realized that that last measurement was taken with the receiver on the DBT. Plugged directly into the AC line, the voltage on the high side of the 3.3 ohm resistor is about 27V. The only appropriate resistor I had was a 10 ohm 2W. Stuck it in and that dropped the voltage at the transistor by 2.7V (270mA, remember?). A 22 ohm 3W resistor would be even better (would drop about 6V and leave a differential of about 7V).
 
Back
Top Bottom