Pioneer SX-1980 Transistor Modification

OK...so whatta do? You got a SX-1980 with blown output transistors. You can still find the PNP's around, and even Pioneer shows them available on their website (for about $50 each). But what the hell good is that if the NPN's are gone? Well, the answer is 'not much'.

The original 2SB706A's and 2SD746A's were pretty unique devices. But it has now been some time since you could order some up from Pioneer (or anyone else). Thing is, there are good transistors out there, but making them work is going to be tricky.

What I want to do is to replace ALL the output devices on a SX-1980 with new On-Semi MJL4281A's and MJL4302A's.

Let's compare specs:

An entire set of 12 new On-Semi transistors can be bought for well under $100. Looks like a real bargan if you can make 'em work. The only real concern here is the much higher ft of the new transistors. The possibility of oscillation exists, but that can be addressed if necessary. First things first...find a way to get the things mounted.

Here's the stock setup as delivered to me. The last clown went crazy cutting wires, which in reality is not a big deal, since this whole mess will have to be rewired for the new transistors. Every transistor is blown. I'm not sure about the other side, but both sides will get new transistors anyway.

Another problem is that the wire end connectors are burnt...a result of someone installing the connector in the wrong place. The pic doesn't show it too well, but two of the ends are vaporized. This won't do, and I've found replacements. In the second pic are a couple of Amp pins designed for a totally different style of plastic connector, but they do appear as if they will fit in the stock Pioneer connector. Notice in my last post, in the pic the left plastic connector is removed. Not easy, but it can be done.

The pins that these fit on are 1.5mm diameter. Well, these female pins are designed for 0.062" pins. Do the math. Nice huh?

The silver pin at the top of the pic is a simple tin plated one designed to be crimped. The one at the bottom is identical, except it is gold plating over nickel and is a solder tab design where the tab extends out beyond the body of the plastic plug. I'll likely use the second style, as I don't think I will be able to get the top of the pin body into the Pioneer plastic connector far enough to make reliable contact.

FWIW, the part numbers for these socket pins (made by the Tyco division of Amp) is A1323-ND for a tin crimp type, and A1343-ND for a gold crimp type. For the solder tab, A1668-ND for a gold one, and A1631-ND for tin. The numbers are Digikey numbers, and the crimp type are for 24 to 20ga wire. Seem small? Well, those big wires that Pioneer used are nothing but 20ga wire with big, thick insulation.

Here's what the mounting surface looks like (pic 1). In pic 2, you can see how the new transistor will sit on the heatsink. I plan on drilling and tapping a hole for the top screw on the transistor...not to hold it on, but to install a short M3 set screw to act only as a locator. I could have used a screw here to hold it on, but experiance shows that, when tightened, the end of the transistor opposite the screw 'lifts up', and contact with the heatsink is ruined. What I want to do is to make a 'bar' that will sit on top of the transistor and use the two screw holes on the left and right to tighten the bar down onto the top of the transistor. This will make for wonderful contact of the transistor to the heatsink. Only drawback is that I'l likely have to farm out the manufacture of the transistor hold-down bars to a machine shop. I also have to use a material that will not bend, so it will be fairly thick.

Stay tuned...

EW, I'll fab those clamps for you, I'll even do it for free, Just to see you pull this one off since it's one of the things that keeps me from buying a sx-1980: unobtainum outputs!.(well, that and the $2000.00)

You need 24 of these right?

Material will be 6061 alum.

EW;

As a owner of a SX-1980 I have worried many times about the unobtainum outputs transistors for this TOTL receiver, even to the point of considering selling it while it's still in good operating condition.

If you pull this upgrade off to your total satisfaction, including identifiying and solving any high freq oscillation, I will donate $15 to the AK site. Hope other AK members that own this receiver consider doing the same.

Lefty

PS: About the hold down bars, looking at the first picture couldn't the original hold down bars be flipped over and reused as a clamp for the new devices?

Lefty said:

EW;


If you pull this upgrade off to your total satisfaction, including identifiying and solving any high freq oscillation, I will donate $15 to the AK site. Hope other AK members that own this receiver consider doing the same.

Lefty

Click to expand...

Count me in. I've sold every SX-1980 I've had (3 of 'em) and this is one of the main reasons. Lefty, I'll meet your $15 and raise you $15. GO EW!! :thmbsp:

Have you looked into a modification using the original clamps? It would seem to me that you should be able to modify the existing clamps.
I’ll just toss out an idea or two. Since the original transistors are themselves providing additional support for the bracket because of their width you
would need to use a spacer on the screw itself. This is so that the bracket would have something to rest against when you tighten it down to avoid bending it. The spacers could be rectangular which would help keep the new transistor from shifting side to side. Plus they would require little in the way of fabrication. If the thickness of the originals is more
than the new ones you could use a shim to make up the difference.

It's all just a thought
Maybe you can do something with it.

EW,

Any chance the original trans are darlingtons? Looking at the spec, count the number of legs, and the old technology I think they have to be. The replacement beta is a bit lower at minium spec, you may need to do a discrete darlington. However, it may still be ok with one tran if the driver stage can handle the higher loading.

No...the originals aren't Darlingtons, they are ring-emitter transistors. They do have two connections for the collector, but B-E diode drop is a standard .55V, so biasing will be the same as the stock units. Oh...and that minimum beta of 50 on the On-Semi transistors? That's at 8 amps. It's considerably higher at lower currents.

The stock hold-down is thin sheet metal, and worthless for securing the new transistors.

Technut...nice drawing, but all that is really needed is a flat piece of .250 thick metal about 44mm long and 19mm wide with two holes on the ends to 'clamp' the transistor to the heatsink. You could even skip adding the M3 locator setscrew, but if I'm gonna do this, I want it my way. :yes:

Edit: What I mean is that your design would call for clamping the transistor with only one screw, and require a lot more work than necessary. Sure looks cool, but one of the criteria that I am going for is to have this easily duplicated, and with a piece such as you show...nice as it is...would mean difficult duplication. It is a fact that one of the best ways to ensure good contact to the heatsink with a TO-3P device (such as the new transistors will be) is to use a hold-down 'bar' such as I am proposing.

EW, I am looking forward to your results on this modification. Certainly worthwhile to restore those units that have seen abuse in the past.

I plan to keep my SX-1980 indefinitely. It's my daily driver. I don't abuse it. I play it at reasonable levels with 8 ohm nominal rated speakers, keep it adjusted for DC offset and bias, and provide ample ventilation.

AFAIK, there's no known "aging" process in output transistors that would cause their demise, even if the unit is working properly with a safe loudspeaker load.

EW, great pics. The wire Pioneer used was certainly top quality, "Star Quad" configuration so they boasted on the SX-1980 brochure!

EchoWars said:

No...the originals aren't Darlingtons, they are ring-emitter transistors. They do have two connections for the collector, but B-E diode drop is a standard .55V, so biasing will be the same as the stock units. Oh...and that minimum beta of 50 on the On-Semi transistors? That's at 8 amps. It's considerably higher at lower currents.

The stock hold-down is thin sheet metal, and worthless for securing the new transistors.

Technut...nice drawing, but all that is really needed is a flat piece of .250 thick metal about 44mm long and 19mm wide with two holes on the ends to 'clamp' the transistor to the heatsink. You could even skip adding the M3 locator setscrew, but if I'm gonna do this, I want it my way. :yes:

Edit: What I mean is that your design would call for clamping the transistor with only one screw, and require a lot more work than necessary. Sure looks cool, but one of the criteria that I am going for is to have this easily duplicated, and with a piece such as you show...nice as it is...would mean difficult duplication. It is a fact that one of the best ways to ensure good contact to the heatsink with a TO-3P device (such as the new transistors will be) is to use a hold-down 'bar' such as I am proposing.

Click to expand...

Okay, how about this one (see attachment)? This is 44mm X 19mm X 6.35mm. The holes are 3.5 mm to allow the screw some clearance, I am assuming that the screws thread into the heatsink right? All I need from you is the distance between the center of the holes or the distance from the edge to the center of the hole.

By the way, on the first one I meant for you to use 2 per transistor (one on each side).

technut said:

Okay, how about this one (see attachment)? This is 44mm X 19mm X 6.35mm. The holes are 3.5 mm to allow the screw some clearance, I am assuming that the screws thread into the heatsink right? All I need from you is the distance between the center of the holes or the distance from the edge to the center of the hole.

Click to expand...

OK..here's a drawing. I reduced the width to 43mm for clearance.

By the way, on the first one I meant for you to use 2 per transistor (one on each side).

Click to expand...

Ohhh....!! (light bulb above head...) I see. Still, if you went that way, the thickness of the machined piece is too critical. The transistors are not uniformly manufactured, being largely plastic and all. I'd be afraid of the transistor being held too tightly or too loosely. The 'bar across the top' method is tried and true, as long as the bar material is flat as possible, and will not bend when tightened down, as I'm sure .250 stainless will not.

OK...I've got two machine-shop guys wanting to help out here...englissa and technut. Do we have two 1980's that need outputs? I really apprecate the offers...perhaps a second set is not such a bad idea, since once an owner of a dead 1980 sees this thread, I'm gonna get an email.

EchoWars said:

OK..here's a drawing. I reduced the width to 43mm for clearance.Ohhh....!! (light bulb above head...) I see. Still, if you went that way, the thickness of the machined piece is too critical. The transistors are not uniformly manufactured, being largely plastic and all. I'd be afraid of the transistor being held too tightly or too loosely. The 'bar across the top' method is tried and true, as long as the bar material is flat as possible, and will not bend when tightened down, as I'm sure .250 stainless will not.

Click to expand...

Done, How many do you want?

Send me a PM with your shipping address and will get them out to you next week.

Email sent.

Twelve are needed per receiver. If both you and englissa get them done, I'll be set up for two 1980's. If this works, and I can make these output devices work, I'm positive it won't be long till SX-1980 #2 comes along.

I can't thank you two enough...I didn't expect this much assistance, but it is greatly appreciated. :thmbsp: :yes:

Hey EW, please keep all of us inform of your progress!

Rgr that.

Pm

If this works count me in for the donations to AK.

Is there anything that should be done to prevent the outputs from going bad?
( other than common sense, cooloing 8 ohm load etc...)
Mandatory cap replacement ?
Will this fix apply to a 1280 ?
Thanks Again
EC

EW: Your clamps are done, I will put them in the mail tonight.

The piece of stainless we had was .210 thick instead of .250 (1/4) thick. But if you can bend this stuff with a 3mm screw without stripping out the threads in the heatsink, I'll eat your hat.

Seriously though, it should work just fine.

Edit: I just thought about the holiday=no mail. I will send them tommorow.

Whatever1 said:

If this works count me in for the donations to AK.

Is there anything that should be done to prevent the outputs from going bad?
( other than common sense, cooloing 8 ohm load etc...)
Mandatory cap replacement ?
Will this fix apply to a 1280 ?
Thanks Again
EC

Click to expand...

The 1280 is more complex, as it uses a circuit board to mount the emitter resistors remotely with the output transistors. I have a drawing that I made to use a perfboard with inline traces for replacing the outputs on the SX-1280, as AK had a dead 1280 a little while back. However, before I got to fabricating the parts, Grumpy found some original parts from a 1280 that was being parted out. So yeah, this could be done on the 1280, but it would be more work.