SX-850 In The House! Recap/Restoration Ahead!!

[MarkyM]

Hi, I know there are a lot of SX-850 restoration threads but this is my first Pioneer ever! Back in the day, I bought a Marantz 2275 instead I have been looking for one of these for quite a while that was not a complete ripoff, was in pretty good shape and close enough to me that the shipping wasn't so much. And I finally found one! This is a seller photo:

Cosmetically it is very good, especially the front panel.

It's an "HG" model with the rear panel voltage selector. Serial number is XE so it is from May 1977.

And very clean inside as well, no tobacco smell and not much dust!. So far it all looks original! All 8 of the outputs appear to be the original Toshiba transistors with the red lettering.

It passed the DBT and passes audio well..until the loud pops & crackles start From reading the other threads, I'm pretty sure that's those renowned 2SA725's and 2SC1313's in there. Replacements are on order along with all electrolytic caps, plus all diodes and transistors on the power supply board...which was literally hanging on by one of those plastic standoffs. The others had crumbled into little pieces in shipping. How dumb that they mounted this board upside-down like this without much ventilation. But the board appears to be in very good shape.

I'm really looking forward to this project, it is an amazingly solid piece! They sure don't make'em like this anymore!!!

--Mark--

 

[Super Noob]

Marky Mark…. The 850 is a great rig and you will love it when restored. I prefer 850 over 2275 / 2270 - all restored.

 

[MarkyM]

Marky Mark…. The 850 is a great rig and you will love it when restored. I prefer 850 over 2275 / 2270 - all restored.

Cool, I’m really excited for this project!
And thank you so much for your 850 thread. It is phenomenal!! It has already been a huge help.

 

[MarkyM]

Question about the main filter caps…Is 22000uF too much? The originals are 15000uF/50V. Those 18000’s are no longer available.

 

[EastPoint]

22k is fine.

 

[t4t3r]

She's a beauty. I know they made a lot of the xx50 series but can't be too many floating around these days in great condition and untouched like that. I have a 950 that I still need to go through at some point, but even though it cleaned up nice, it came to me in a much worse condition than this one.

Look forward to reading your progress if you decide to document it here (highly encouraged!).

 

[MarkyM]

22k is fine.

Thanks.

 

[MarkyM]

She's a beauty. I know they made a lot of the xx50 series but can't be too many floating around these days in great condition and untouched like that. I have a 950 that I still need to go through at some point, but even though it cleaned up nice, it came to me in a much worse condition than this one.

Look forward to reading your progress if you decide to document it here (highly encouraged!).

Thanks, will do!

 

[MarkyM]

Starting on the power amp board. Will remove each transistor, clean off the old thermal compound, inspect the mica insulators and apply fresh thermal compound.

I'll do the tuner board next to finish the topside first. I pulled the dial string wheel off and used some blue-tak to keep the dial string in place. The tuner board lifts up from that side after removing the 5 wire pins that have the wires going off to the right. The other wires all go to the left of the board.

 

[sruchris]

Another NE Ohio guy! Woo hoo!

 

[MarkyM]

Another NE Ohio guy! Woo hoo!

Yep! Nice to meet you neighbor!

 

[cuda440]

I restored on of these 3 years ago, but mine did not have sockets for the power transistor heat sink bank like what you show here. I have a 1050 as my daily driver and it doesn't have anything like that either.

What does the receiving end of those connectors look like?

 

[MarkyM]

I restored on of these 3 years ago, but mine did not have sockets for the power transistor heat sink bank like what you show here. I have a 1050 as my daily driver and it doesn't have anything like that either.

What does the receiving end of those connectors look like?

They aren’t actually sockets. They’re just solder pin blocks. I think the 1050 has molex connectors.

 

[cuda440]

Well whatever they are, the 850 I did didn't have them. And my 1050 doesn't have them or molex connectors either, though the 1250 may have molex.

I wish they all had some kind of quick disconnect mechanism for the power transistor assembly. Would make servicing much easier.

 

[markthefixer]

The ones in the picture are NOT quick disconnect, those were desoldered to liberate them from the PC board.
He did a great job keeping those clear plastic "headers" intact.

Quick disconnects add another point of long term potential failure - like on the SX-1250 (and SX-1980), and having to shine up the pins. The SX-1280 went to soldered connections and greater long term reliability.

Then they had to go and socket the output transistors in the SX-3900. Between the sockets and skimpy wave soldering, they really compromised the future of those models.

 

[MarkyM]

Another NE Ohio guy! Woo hoo!

The ones in the picture are NOT quick disconnect, those were desoldered to liberate them from the PC board.
He did a great job keeping those clear plastic "headers" intact.

Quick disconnects add another point of long term potential failure - like on the SX-1250 (and SX-1980), and having to shine up the pins. The SX-1280 went to soldered connections and greater long term reliability.

Then they had to go and socket the output transistors in the SX-3900. Between the sockets and skimpy wave soldering, they really compromised the future of those models.

Thanks!

Ahh, makes sense. These really weren't a problem to desolder. Just a little extra time, which is worth it for long term reliability IMHO.

Question...I'm ready to start matching transistors for Q1-Q2 & Q3-Q4 and I'm a first-timer for this. I have one of these testers:

As I understand it, hFE is the most important followed by Vbe correct? What is the +/- tolerance I should look for and is Ie important at all?

Thanks!

--Mark--

 

[Watthour]

There are varying opinions, but I've had good success matching simply on current gain measurements. I try to get "pairs" by mating devices within 1½% of measured gain. If you can get within ± a couple of digits between devices for both current gain and forward voltage drop, you win some kind of prize, I think.

I have also found that devices packaged together (I usually get them on tapes) can vary up to 30% from one device to the next, and 20% is pretty common. Thus, avoid the mistake of assuming that any given lot of devices packaged together would have a similar measurement. They don't, even though they may have been manufactured in sequence.

 

[MarkyM]

There are varying opinions, but I've had good success matching simply on current gain measurements. I try to get "pairs" by mating devices within 1½% of measured gain. If you can get within ± a couple of digits between devices for both current gain and forward voltage drop, you win some kind of prize, I think.

I have also found that devices packaged together (I usually get them on tapes) can vary up to 30% from one device to the next, and 20% is pretty common. Thus, avoid the mistake of assuming that any given lot of devices packaged together would have a similar measurement. They don't, even though they may have been manufactured in sequence.

Thanks!

 

[painttoad]

Question...I'm ready to start matching transistors for Q1-Q2 & Q3-Q4 and I'm a first-timer for this.

since it's your first time you may or may not know: don't handle them with your fingers. bodyheat will affect hfe. i use hemostats and handle them by the legs.

i like the look of your 850 better than my 1010. way better.

 

[MarkyM]

Thanks for the tip!