My restored SX-1980

Monday, June 24, 2019

Pioneer SPEC-2 Gold Tier Restoration

Here is a restoration I did a while ago and is a piece from my personal collection, a SPEC-2. Some background, I got this from a guy who I restored a SPEC-1 and SPEC-2 for a few years ago. Long story short, he had a second one in the basement that had problems and gave it to me for free! When tested it would not come out of protection. Upon inspection of the boards I found the caps and transistors had been replaced on both power amp boards with a weird mix of good and bad replacements and a couple of questionable mods.
 
Let us start with the left power amp board, as it has some interesting things done.
The left channel STV diodes had homemade replacements made from three 1N4007 diodes in series for the STV-3H and four 1N4007 diodes in series for the STV-4H in a weird plastic case screwed to the heatsink.


Some close ups with the plastic cases removed:



An interesting way to replace the STV diodes, however not ideal for good thermal tracking. I tracked down an original STV-3H diode from a seller in Germany and a STV-4H from a parts receiver that I had here. The 4H diodes are not easy to find, so I was thankful I had a parts unit with one in it.
 Board before (note I have already replaced the STV diodes)


Molex pins (note I cleaned the right 2 to show the differences):


Parts ready to be used!


A cool thing I noted when I was looking for a polypropylene replacement for the input cap, the board actually provides pads for bigger capacitors than the little tantalum being used currently. It allows for, in addition to the 5mm lead spacing of the tantalum, 10mm, 15mm and 22.5mm lead spaces via the trace here:
 
And after! In total 8 caps, 14 transistors, 6 diodes, 3 variable resistors and the bias relay were replaced. Nichicon UPW for the 3 electrolytic caps, Panasonic and Wima caps for the rest of it. Transistors are Fairchild and On-Semi.

Molex pins:

STV-4H:


 STV-3H:

Close up of the new input cap and transistors, I beta matched the differential pair to 1% and then heatshrinked them face to face for better thermal stability.


The outputs were also replaced, but they do not look good, some look like somebody hit them with a punch and they are dented pretty good in the middle of the case. Not going to trust them.


And after, the micas were also replaced with new:


Now for the right channel. The STV diodes were all original on this board, but it still had questionable cap and transistor replacements. Some of the transistors were replaced with the correct type however.


And after! All the molex pins were cleaned and other than the STV diodes, all the other replacements parts are the same.


The right channel outputs are good originals, but in the interest of symmetry I am replacing them with the On-Semi MJ outputs as well.


And after! Micas were replaced here as well.


Next, the power supply. Pretty simple job here:



And after! Just 2 caps and 2 relays here. Nichicon PW for the caps and the relays were replaced with Omron low noise relays.


Last, we have the meter amp board, this also supplies some of the regulated DC rails as well. This also has had some caps and transistors replaced, none of them however are good quality replacements.


The primary surge relay on this board is pretty easily replaced, the case can be transplanted onto the new relay with no mods needed.


Old vs new:


New relay, old case!


I added spade connectors to the wires for easier removal later if necessary.



One unique thing on this board is the 10uF 160v bi-polar cap they use to pass current until the relay closes. It is pretty much impossible to find a electrolytic replacement, but you can find a polyester film with that capacity. The only issue is that the poly cap is much bigger and so is the lead spacing, however you can use the lead hole that's circled and it allows you to fit a bigger cap.


And after!


All done! 18 caps, 17 transistors, 2 trim pots and the relay were replaced. Nichicon PW for the electrolytic caps, Panasonic and Wima everywhere else. Transistors are all Fairchild.


 On to the filter caps! Oddly they only used 15000uF x 4 instead of the 22000uF x 4 like the SX-1250 and SX-1980. Since this amp has three soft start/surge current circuits, one for the primary and one for each of the secondarys, we can safely up the filter cap size to 22000uF with no problems. Note one had been removed already before I took the photo.



Old vs new.


And after!



And it's ready to fire up! Fired it up on a DBT, checked bias to make sure it set properly, everything passed so I hooked it up on line power. Set the bias and meters to spec, connected my speakers and source, hit play and voila!


Words fail to express how good this thing sounds, but safe to say it sounds better than my 1980 by a good margin. However, when I recapped the 1980 I didn't upgrade any of the signal path caps to poly films, so it's not exactly surprising. Upgrading signal path caps to polyester or polypropylene film where possible is the single biggest upgrade to the sound you can do by far. Replacing the tired old electrolytic caps is a big step in itself, but polys in the signal path is just that much better. Clearer and more transparent highs and midrange, tighter bass and a much more defined soundstage are just some of the benefits. Not as noticeable audibly but can be noted on the test equipment is reduced THD and better S/N ratio as well.

Sunday, April 15, 2018

Exciting new developments!


At long last I now offer full performance evaluations of any restorations I do along with full tuner alignments. Performance reports include THD+N at rated power, max power @ 1% THD+N, signal to noise ratio, frequency response and more! Next restoration will be a SPEC-2, stay tuned!

Friday, January 12, 2018

Pioneer SX-1280 Gold Tier Restoration


Up today is a customer's SX-1280, AK member suitekids. This receiver is one of his babies and he asked me to give it the best, so here are the results!

This 1280 has been serviced before, the previous tech did a dubious mod to the protection circuit which I'll be showing here a little later. We'll start with the preamp boards.


First up, the flat amp board, a 2 stage DC circuit employing a FET input stage for higher input impedance. It also utilizes NFB for a lower output impedance.


And after. 10 caps and 2 transistors were replaced. Panasonic FM and ECW poly caps were used here. The 4.7uF input caps were replaced with a polypropylene as space allowed it. Transistor replacements are low noise Fairchild equivalents.


Here are some photos showing how much the original input caps have deteriorated.

At 120Hz:


At 1KHz:


Way out of tolerance and high loss as well. A near perfect cap's loss tangent should be around -90°.

Moving on, we have the tone amp board. This is also a 2 stage NFB circuit.


And after. This board has a lot of parts that are replaced, 36 caps and 4 transistors were replaced here. Replacements are Nichicon KL for the orange low noise caps and tantalums, Rifa and Wima polypropylenes for the caps under 1uF and Panasonic FM for everything else. Transistor replacements are low noise Fairchild equivalents.


Next, we have the function board.


And after. Just two caps to replace here, using Panasonic FM and FC here:


Now, we have the filter amp board. A couple caps have been replaced already on here, though they are not the ideal replacements.


And after! 8 caps and 2 transistors were replaced, Nichicon KL for the caps above 1uF and Wima polypropylene for the caps under 1uF. Transistor replacements are low noise Fairchild equivalents.
 

The meter amp board is sometimes overlooked, but it does have a few caps that need replaced.


And after! Nichicon PW caps were used here.

Next, the phono amp. This is the last board on the bottom of the receiver that needs to be done.


And after! 14 caps and 10 transistors were replaced, Panasonic FM used for the blue filters, Nichicon KL for the orange low noise caps and Nichicon ES were used for the 4.7uF output bipolar caps. The mylar films were replaced with Panasonic ECQ and Wima FKP caps.


Moving on to the power amp boards, here is the right power amp board:


An unpleasant surprise greeted me when I flipped the board over, the previous tech destroyed part of a trace leading to the DC-offset adjustment pot when they replaced it, and the jumper they put in is a little remiss. Not hard to repair thankfully.


I bypassed the broken trace completely and soldered some small gauge solid core wire to a solder joint further up the trace. Makes for a much cleaner look.


The 2SA979 dual transistor I replaced with a 1% beta matched pair of KSA992 transistors put face to face and heatshrinked together for better thermal tracking and stability. I've found that the 979 tends to become leaky, causing the offset to jump around a lot and make it hard to dial in, even with a multi-turn pot. The 1uF input cap was also upgraded with a Panasonic polypropylene.
  

And after! 9 caps, 3 transistors and both trim pots were replaced. The molex pins were also cleaned to a nice shiny surface.


The left amp board before:


And after! Same as the right amp board, except this one didn't need any traces repaired!


Big one here, the regulated power supply board. This was already recapped by the previous tech, however the cap replacements used were less than ideal and most of the them were not low impedance. Amazingly the soft start relay was replaced with the proper Omron replacement.


While I was replacing the relay (just so everything is new), I noticed this. That is pin 1, which is the 120V AC in to the soft start circuit. This thing did work when I got it, so it must have been just barely hanging on and me moving the board to get to the back broke it loose.





Much better!



And after! 16 caps, 7 transistors, 7 diodes and the relay were replaced. An extra heatsink was added to Q5 as well. Nichicon PW along with Nichicon TVX caps were used here. Diodes and transistors were all On-Semi/Fairchild, with several transistors getting upgraded to a TO-92L case style for better heat dissipation and longer life.


Up now is the protection board. One of the concerns brought up by suitekids when he sent this to me was the fact that there was no protection delay at all and would thump his speakers pretty hard. Anybody see what's wrong in this photo?


This is a good mod, right? I mean, it allows the relay to engage, so it's obviously a good thing! 😒


And after, looks a lot better! The protection relay was replaced with the proper replacement, but for the sake of everything being new I replaced it again. 6 caps, 2 transistors and the relay were replaced, Nichicon PW was used here and relay driver was upgraded to a TO-126 case style for better heat dissipation and longer life.


And finally, the tuner board.


And after! 26 caps were replaced, Nichicon KL were used for the low noise and tantalum caps above 1uF, Wima and Kemet stacked film polys used for caps under 1uF and Panasonic FM and FC for everything else. The tuning capacitor was also cleaned.


Filter caps were also replaced and upgraded. Before:



And after! Filter caps were upgraded from 15000uF to 22000uF, this is safe to do as the 1280 has a soft start circuit. I also cleaned and polished the chassis and transformer, came out pretty nice!

 


Fired her up, set the offset and bias to spec, let it play for an hour to settle everything in, then reset the offset and bias adjustments. Got a video of it playing, this thing sounds phenomenal!


And now some beauty shots! Protect anything damaged by drool, you have been warned! :-P




And that completes this restoration! Feel free to leave comments or questions. Up next will be a Pioneer SPEC-2, so stay tuned!
An unpleasant surprise greeted me when I flipped the board over, the previous tech destroyed part of a trace leading to the offset pot when they replaced it, and the jumper they put in is a little remiss. Not too difficult to solve, just annoying.
An unpleasant surprise greeted me when I flipped the board over, the previous tech destroyed part of a trace leading to the offset pot when they replaced it, and the jumper they put in is a little remiss. Not too difficult to solve, just annoying.