Pioneer SX-7 very hot running TO-220s due to poor design or component failure?

Based on the schematic posted below, would these transistors be expected to run this hot? None of these are heat sinked. This unit is from the early 1980's.

What is the cause of this excess heat?

Q26 95° C (note that this one is a Darlington, if that matters)
Q27 103° C
Q28 77° C

SX-7 regulator TO-220 temps.jpg
 
Last edited:
That’s one of the reasons regulator circuits like this, in amplifiers like this, like to fail. Make sure the solder joints haven‘t failed or oxidized, and add some small heat sinks if possible. A little scrap of aluminum angle works wonders.
 
  • Like
Reactions: 1 user
It looks like Q27 drops more than 30 volts. Anyway the power in each transistor depends on the product of current times voltage. I don't see any info on the schematic snippet to tell what the current is. You can measure the voltage drop across R116 and R117 to get an idea of some of the current loads.
 
  • Like
Reactions: 1 user
@Bill_P Thank you. Yes, I am able to calculate current as you describe. And with that I can calculate power dropped in the transistors.

But with that information, will we be able to know whether the current being drawn by circuits downstream is correct? Or if there is a fault or a failure downstream?

I am trying to make a decision on whether to heatsink and re-solder these transistors, or to abort the project. This unit was free to me from a recycling facility. Unfortunately, I do not have time to do extensive troubleshooting. :( :geek:



@anyone Is it typical to see 100°C on series pass TO-220s in equipment of this type?

.
 
Last edited:
So far I see nothing definitive indicating a fault. Does the unit work? Are the voltages you measure in agreement with those on the schematic? Case temperatures at 100C are about as high as should be allowed. It would be beneficial to add small heatsinks if the unit works well otherwise and you plan to use it. Do a few measurements so the transistor power can be calculated. A serious fault would probably pull down the regulator output and really drive up the transistor temperature beyond ratings (150C).
 
  • Like
Reactions: 1 user
Moderator
Joined 2011
If it works normally, just use it as-is. Probably there's no real fault causing the high temperatures.
Clean any dust or dirt off the device tab. This receiver is designed for a 120VAC line.

In case you want to clean it up and give it to someone else, I'd resolder the hot devices.
If there is room, adding small heat sinks might help a little, but don't expect much.
 
  • Like
Reactions: 1 user
Thanks everyone! (y)

Yes, this receiver works...and surprisingly, it sounds pretty good to me. It carries a bass line well for a 60 watt, 45 year old, all-original consumer piece. It is also, I believe, from the first post-vintage series from Pioneer. It boasts "computer controlled" functions including volume and balance control, and has several good sized ICs, all which does not inspire confidence IMO. And I found it near the top of a pile of BPC in a giant box at the recycling facility. Kind of like an ugly dog that followed me home, and now I am starting to have feelings for it, lol.

As for the approximate 100°C temps I read on the tabs of those transistors, that is with the case cover removed, obviously. But also removed is a metal shield (flat sheetmetal) that covers 1/4 of the unit, directly above where these transistors live. They have no direct access to air (for convection) either above or below. It is amazing the thing still works at all. I can only guess how much higher those temperatures might be if readings were taken in its fully enclosed state, after a couple hours of playing music on a hot day. o_O

It may be a day or two, but I will get the voltage readings as suggested and report back here.
 
Last edited:
2SD507 NPN transistor, (Your Q27 !), " 250A maximum collector current, 150 C maximum operating junction operating temperature."....Attaching a nice substantial heatsink would be rather straight-forward. PS, does your SX-7 have the elegant curved heatsink that goes full length front to back?....or do you have the plain-vanilla two plates of aluminium?


-----------------------------------------------------------------------------------------------------------------------------------------------Rick...
 
  • Like
Reactions: 1 user
Thanks for joining in @Richard Ellis Are you familiar with the SX-7? If so, feel free to contribute anything you may know about it.

Yes it sounds like the D507 is pretty robust. And yes, attaching a heat sink would be easy, and I have some on hand. But the real point of this thread was making a decision as to whether something else might be wrong in the amplifier section that would preclude investing any more time in the project.

The main heat sink, choosing from your two options I'll say it is the more elegant one. It does have the curved fins. That heat sink gets decently warm too, at approximately 49°C at its top at idle. This is my first time to work on a vintage era Pioneer. Some things they did very nicely. Others, like thermal management, seemed to get very little attention. For example that pretty, curved-fin heatsink is solid on the bottom, meaning that air can not flow bottom-to-top through it as you would expect. Further, along its sides there is very little open space. On the left side it overlaps the PCB and is raised only about 3mm above the board. Not much of a gap. Along the right side of the heat sink is about a 5mm slot where there is a clear shot for air to rise from below and out the top, past the heat sink but not necessarily through it. There are also two 3-terminal regulators in a separate regulator circuit that also run hot (approx 70°C).

But all that said, it is still running, and sounding good, after all these years.
 
Last edited:
Not a 2SD507, it’s a 2SB507 - the complement to the 2SD313. They are TIP31/32 class devices, but a bit faster. 2 watts of dissipation with NO heat sink will run them that hot. It’s actually safe for the transistor, but can oxidize the solder joints and turn paper-based PCBs to carbon slowly over time. And that’s what fails.
 
  • Like
Reactions: 1 user
...For example that pretty, curved-fin heatsink is solid on the bottom, meaning that air can not flow bottom-to-top through it as you would expect. Further, along its sides there is very little open space. On the left side it overlaps the PCB and is raised only about 3mm above the board. Not much of a gap. Along the right side of the heat sink is about a 5mm slot where there is a clear shot for air to rise from below and out the top, past the heat sink but not necessarily through it.

Correction to this portion of my previous post. Along the length of the right side (only) of the main heat sink, the approximately 5mm slot between the solid base of the heat sink and the board IS situated such that air can flow upwards through the heat sink fins. So I was wrong about that (going from memory).

But, I have not seen any other amp/receiver where the entire main heat sink did not have unrestricted air flow across its length and width, from both above and below. In this receiver, the width of the heat sink is 50mm of which 40mm is solid on the bottom, and 5mm on the left side overlaps the board, leaving only 5mm, or 10%, where convection can occur unimpeded.

And I am putting my thread off-topic by complaining about the main heat sink :rolleyes:
 
Last edited:
For the B507, I actually have a couple KSB596Y and also MJE15033G.
Would either of those be a "no-worries, drop-in swap" (I don't have the knowledge to test) ? Or should I even bother with replacing the original B507?

For the D836A Darlington, I have no nationality-type preference. Are there non-Jap-type options?

I looked at the 2sd2014, which is available at Digikey. It is a TO-220FP, and looking at that case style my first thought is that it might possibly send more heat down the leads into the PCB (last thing this amp needs) through the CU clips ?? This (generic) data sheet reads: "CU clip interconnect for better heat dissipation efficiency". Is that a potential concern?
https://c44f5d406df450f4a66b-1b94a8...ssl.cf2.rackcdn.com/2018/02/TO220FP_DS610.pdf
Also, TO-220FP is coated so there could be no metal-to-metal contact with a heat sink?
Also, is gain matching with the original device important, especially with devices having gain over 1000hFE?

Obviously, transistor swapping is far beyond my knowledge level...
Perhaps regulators are not finicky about series pass transistors, I don't know?
 
Last edited:
Don’t need to swap them at all if they are working. And if you do address the heat problem by tacking on a small piece of metal you can be assured they will continue to work. The 596 is a drop in for all applications (including small amp output stages and drivers). The MJE’s are a high speed device. They also work fine as regulators - many DIYers design them in.

The insulated TO-220’s are my choice for this sort of thing. Why? Because just using one in one position means you can bolt two to the same piece of metal. This makes a more rigid structure. PCB layout may or may not favor this, though. If you dont add heat sinks to pull heat out the top the uninsulated ones are better, so it doesnt send more heat down the leads.

”CU (copper) chip interconnect” has lower internal temperature rise - junction to case. But that’s not where the bottleneck is here. At low dissipation, it would only make one or two degrees of difference.

The “nationality” issue is more about getting a closer match to the original in terms of fT and the values of the internal base-emitter resistors. Neither are critical here, but the better repair shops (expensive) would at least attempt to get close to originals. About the only ON or ST option in medium power TO-220 darlington would be TIP types. Some people are allergic to them.
 
Not a 2SD507, it’s a 2SB507 - the complement to the 2SD313. They are TIP31/32 class devices, but a bit faster. 2 watts of dissipation with NO heat sink will run them that hot. It’s actually safe for the transistor, but can oxidize the solder joints and turn paper-based PCBs to carbon slowly over time. And that’s what fails.
My bad, the 250? amps worth should have had me looking it up over & over...but the 150C temperature still holds true. Flail me with a three-foot strand of solder for my mistake...

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------Rick...
 
  • Like
Reactions: 1 user