SX-3700 Recap/Restore Thread

Strangely, the idle voltage on the Right channel is lower than spec. Should be 2.2mV.
TP48-49 (L Ch): 14.1mV
TP46-47 (R Ch): 2.0mV

Pins to Chassis:
TP46: 29mV
TP47: 34mV
TP48: -7.0mV
TP49: -20.8mV
 
Ok, more mystery. I realize this isn’t perfectly kosher, and I hope AK can forgive me, but on a hunch I took the unit off the DBT and put it on AC. The relay engages just fine. Now back on the DBT and.... the relay won’t engage again. I’m afraid to try it again, but maybe this sheds some light here?
 
First time posting, short time lurking. Found a lot of very helpful info to repair and recap my SX-3700. This all started when I noticed the speaker protection relay loudly clicking, and eventually failing completely to push power to the speakers. I thought replacing that relay would have done the trick, but if I had read these forums more closely, I would have found out my problem likely lies elsewhere.

That brings me here, where I'm going to attempt to recap the Power Amp, document my process, and hopefully learn a few things in the process. I haven't seen a ton of SX-3700 recap threads on AK, so I hope this might be helpful to others in the future. I've read the 3700 Service Manual and the Tuning Fork guide on Pioneer caps.

First, here's a draft of my cap list. I'd appreciate any feedback folks have here. I'm a bit unclear about the difference between Nichicon caps that end with 1TD,1TA vs no suffix after M[x]D.

2x ACH-215; 8200uF/50VDC; 647-UKW1H103MRD
2x ACG-004; 0.01uF/150VDC;
2x CEA R47M 50L; 0.47uF/50VDC; 647-UKW1HR47MDD
3x CEA 010M 50L; 1uF/50VDC; 647-UPW1H010MDD1TD
1x CEA 470M 10L; 47uF/10VDC; 647-UPW1A470MDD
1x CEA 470M 25L; 47uF/25VDC; 647-UPW1E470MDD
3x CEA 470M 50L; 47uF/50VDC; 647-UPW1H470MED
3x CEA 101M 50L; 100uF/50VDC; 647-UPW1H101MPD
1x CEA 221M 35L; 220uF/35VDC; 647-UPW1V221MPD
1x CEA 221M 50L; 220uF/50VDC; 647-UPW1H221MPD
1x CEA 331M 10L; 330uF/10VDC; 647-UPW1A331MPD
2x CEANL 100M 50; 10uF/50VDC; 647-UKL1H100MDDANA
2x CEANL 2R2M 50; 2.2uF/50VDC; 647-UKL1H2R2MDDANA
2x CEANL 4R7M 50; 4.7uF/50VDC; 647-UKL1H4R7MDDANA
2x CEA 4R7M 50L; 4.7uF/50VDC; 647-UPW1H4R7MDD
2x CEA 100M 50L; 10uF/50VDC; 647-UPW1H100MDD
2x CEA 220M 25L; 22uF/25VDC; 647-UPW1E220MDD
2x CEA 101M 35L; 100uF/35VDC; 647-UPW1V101MPD
 
Ok, more mystery. I realize this isn’t perfectly kosher, and I hope AK can forgive me, but on a hunch I took the unit off the DBT and put it on AC. The relay engages just fine. Now back on the DBT and.... the relay won’t engage again. I’m afraid to try it again, but maybe this sheds some light here?

ya, that can happen - the unit needs enough current to pass safety and turn on relay.

I didn't go back and read, but if the bulb goes dim, just stop using it - it's meant to protect the unit from a serious issue, one that pulls high current and fries things. IF your bulb stays bright, then DO NOT bypass.
 
Like ohgoshjosh I am a short time lurker first time poster. I have started the process of recapping my brothers SX3700 which has been a constant companion to him in his garage; he loves it and wants it restored. My experience is with old radios and point to point wiring although I have had some limited experience with rebuilding circuit boards. Unlike the 3700 being discussed here I am lucky - there are no problems with the receiver except three lamps that need to be replaced. There is no doubt that this unit is completely original and if the information on this board is correct with respect to determining its age, it came off the line in November 1980.

On looking at the boards I am amazed nothing shorted out. It is covered in a thick layer of dust and grime some of which looks like it might be metal. Dodged a bullet I guess. I would appreciate some advice at this stage. The power supply board is now cleaned up a little and I wonder if the two large Nichicon caps should be replaced. I think they are 4700/50v. Some forms suggest they almost never go bad' but my view is if I am already here and they are 38 years old, they should go even if they test out ok.

I also wonder about the power transistors that are bedded in a white compound and screwed into the cooling fins. Its still soft but I think that a proper restore says that they should come off be cleaned up and re-bedded. If so what is the collective opinion on what should be used to do that properly. I don't know if the gel that is used for bedding computer chips would be acceptable.

Thank you in advance for your help.
 
It sounds like you diving in.
The SX-3700 is one of a few NSA design amps. Pioneer used them on several receivers and separates.
They can be tricky to repair due to the high frequency transistors they use. They are difficult to source and nothing modern really works reliably.
The trick is to keep them running before they fail. Your plan is a good one.
There are many threads concerning the 3700, 3800 and 3900.
The power supply runs hot so look too resolder all the connections.
Recap it all of it.
Maybe use the modern Sil-pads instead of the paste.
 
It sounds like you diving in.
The SX-3700 is one of a few NSA design amps. Pioneer used them on several receivers and separates.
They can be tricky to repair due to the high frequency transistors they use. They are difficult to source and nothing modern really works reliably.
The trick is to keep them running before they fail. Your plan is a good one.
There are many threads concerning the 3700, 3800 and 3900.
The power supply runs hot so look too resolder all the connections.
Recap it all of it.
Maybe use the modern Sil-pads instead of the paste.

Thanks for the info and advice. I had not heard of Sil-pads so that is helpful and I will look those up. I was nervous about touching any of the transistors. Since is working well I will stay away from them, recap and see how it sounds.
 
I am hoping ohgoshjosh can point me in the right direction because I am stuck. The question is what is the best and safest way to remove the circuit boards from my unit to start recapping. Do you take all of the screws and disassemble the metal case to avoid damaging them or do you attempt to remove the board screws and release the from the top. Must admit this is different from my point to point wiring in my old radios. I have noticed how brittle everything is after almost 40 years and the hot and cold this old thing has endured over the years in my brother's garage. I appreciate your advice.
 
I’m a newbie here, but for this project I removed the bottom panel and both of the front panels. I’m able to access just about everything I need this way. As others have pointed out, it’s a tricky amp to work on for many reasons. One of which is that not all of the components are easily accessible. I’ve found though that I’m able to access what I need, including the power amp.

In the way of updates on my own sleuthing for the source of the offset on the right channel, I started testing all of the components (offline) leading up to the output transistors (in reverse order). I got some interesting readings on what I think are the driver transistors (2SA913 and 2SC1913 pair). I performed the EW 6-way test on those drivers, and found the right channel to be somewhat different than the left. There’s no short here (if there was, I’m sure I’d be having different problems altogher), but I wonder if those drivers have just been running at a saturated level for so long. I’m going to test all the resistors next, one by one, until I see something that looks off. Taking my time with this project, going one step at a time.
 
Last edited:
I am hoping ohgoshjosh can point me in the right direction because I am stuck. The question is what is the best and safest way to remove the circuit boards from my unit to start recapping. Do you take all of the screws and disassemble the metal case to avoid damaging them or do you attempt to remove the board screws and release the from the top. Must admit this is different from my point to point wiring in my old radios. I have noticed how brittle everything is after almost 40 years and the hot and cold this old thing has endured over the years in my brother's garage. I appreciate your advice.

One other thing I’d recommend (you may have already done this), is to give the board a good thorough cleaning before attempting repairs. Clean everything: the PCB, wires, IC’s, etc. 91% isopropyl and an acid brush. Go slowly. If I had done as thorough a job as I should have in the beginning, I may not have run into the issues I did earlier in this thread (like when I didn’t notice a simple blown fuse) :oops:
 
Last edited:
One step at a time, that's great - you've got the right attitude.

Whew, testing all resistors is a chore, I wouldn't do that. Rarely do I find resistors are an issue.

Good luck on quick resolution! :)
 
One other thing I’d recommend (you may have already done this), is to give the board a good thorough cleaning before attempting repairs. Clean everything: the PCB, wires, IC’s, etc. 91% isopropyl and an acid brush. Go slowly. If I had done as thorough a job as I should have in the beginning, I may not have run into the issues I did earlier in this thread (like when I didn’t notice a simple blown fuse) :oops:

Thank you for your advice and sharing your experience. I am in the process of cleaning the boards up and they need it after years sitting in the garage. As you have suggested the I will remove the front cover and see what kind of access I have. Started by clipping all of the cable connectors to ease up on the movement of the boards. Three of the lamps are gone but I think I going to replace all of the them. Have not identified what the small lamps are in the lower front push buttons but will get to that today. Everything is so brittle. Gently removed two of the wedge lamps and their connectors and snapped off one of the small clips. I am going to find and replace all of the sockets if I can find them. Thanks again for your advice.
 
Those lamp sockets are really a bugger, aren’t they? Very brittle and fragile after decades of heat from those 300mA bulbs. The tabs on mine broke right off almost immediately. I probably won’t replace the sockets themselves since they seem to be tough to find (I’ll figure out a way to mount them), but I did order some replacement LED’s for the bulbs. There’s a guy on eBay who sells a nice kit of replacement LED’s for this unit. Good luck!
 
Those lamp sockets are really a bugger, aren’t they? Very brittle and fragile after decades of heat from those 300mA bulbs. The tabs on mine broke right off almost immediately. I probably won’t replace the sockets themselves since they seem to be tough to find (I’ll figure out a way to mount them), but I did order some replacement LED’s for the bulbs. There’s a guy on eBay who sells a nice kit of replacement LED’s for this unit. Good luck!

Thanks for the lead on the replacement bulbs. You're right the lamp sockets are a bugger, makes me nervous about the condition of the boards. Don't want to crack one of those. My old point to point wiring in radios is sure looking good to me. I emailed the maker of Sil-pads to ask them to identify which of their products would be suitable for power transistors in a high heat environment. Haven't had a response yet but will post it if anyone would like to consider using these in their rebuilds. Thanks again for the lead on the lamps. Much appreciated.
 
Just an update on where I am in this restore of my brothers beloved SX3700.

As I noted before I got lucky here in that everything in this old receiver (November 1980) is working and the two primary concerns were a thorough cleaning of the boards and recapping. The boards had perhaps an 1/8” of grime, dust and generalized goo on them from years in his garage. It took some time with an acid brush, high grade alcohol and Q-tips to get down to the original boards. Everything including the wiring was cleaned.

I take note of the member who advised that my cutting of the wire ties was a no-no. Duly noted. Thank you for the advice. The others will remain intact.

I have ordered all new LEDs for the front panel and push-buttons along with 3 new sockets. The others broke when I took them out and no..I wasn’t rough with them. They just disintegrated. I am being very careful with the boards as a result of this experience.

It is now time to begin the recapping process and I have some concerns.

Transistors in this units schematic show that some are tied together at their base suggesting to me that these are matched sets with near identical bias. My concern is that by swapping out the capacitors that this biasing may be thrown out. I don’t want to do such a thorough restoration job that it doesn’t work.

I am going to replace all of the electrolytic caps including the bipolars as a start. These would include the large 8200 or 8700 caps in the power supply ( can’t recall exactly what they are) . The others I will check and perhaps leave in place especially those that are used for coupling. Any comments would be appreciated. I am going to use Nichicon where I can do so, or those from a recognized source.

The 4 large power transistors have been unscrewed from the cooling fin and it has been removed. I wanted to thoroughly clean the thermal grease or whatever it is off both, clean the sink well and re-seat using new compound or sil-pads. I did contact the company that makes these pads to ask what they would recommend in this situation but no one has responded. If they do I will post it if anyone else wants to go this route. Someone suggested Arctic Ice. Probably any good quality paste would work well if applied properly.

I am going slowly not having a lot of experience on restoring boards. My experience is with old radio and tube amps which use point to point. Thank you to everyone who has offered their advice so far.
 
I wonder if someone can offer some advice regarding my 3700 restore.

I was looking at the Pioneer Fork relating to transistors and read the following:

T0-3 type transistors can be used for the output states of power amplifiers. The metal section, with holes with screws acts as the collector.

I have removed the large heat sink from the board, cleaned the thermal compound from both the transistors and the sink and removed the think plastic sheets that were on the power transistors. Given the above, what type of thermal grease should I be using and does it matter if it is conductive or non-conductive. With the plastic insert on the front of the transistor, between it and the heat sink, there should have been no electrical contact between the transistor and the heat sink which appears to be well grounded. However the Pioneer info seems to suggest otherwise.

Any advice would be appreciated.
 
Just a follow up here and, I suppose, a debriefing now that the unit is back in working order. The fix for my initial issue (the relay was not engaging) had to do with the protection circuit. After I replaced the caps and transistors in this circuit, I was getting output to the speakers. The issue with the flouroscan display was resolved by replacing fuse F4, which was blown (likely due to my own careless probing). I used a small glass fuse puller to get that one out. It took a while to order the fuse puller, but it was 100% worth it. No need to wreck your project by trying to pry out a glass fuse and having it break.

As for the offset issue, well, I decided to leave well-enough alone. Offset readings are at around 40-50mV for the right channel, which is at the upper end of the "acceptable" scale mentioned in other threads. Left channel is fine. I tested the transistors for the right channel in circuit, which didn't reveal any glaring issues like a short. I also tested the resistors R332, R335, R336, but didn't see any obvious issues. I had suspected the driver transistors (Q14: 2SC1913, Q16: 2SA913) might be the culprit, since the readings I got were a bit different from the L pair. I might have replaced them, but I didn't find a crystal clear answer on the forums for a suitable replacement. It seems like Toshiba may have made some replacements for a while, but no longer. The issue could just as well be with the output transistors, but I didn't want to pull them off the heatsink to test them. I'm interested if folks have thoughts about this.

I replaced the lamps with LED's as I mentioned above, and they really look stellar. As usual, the brittle tabs had broken off all my lamps, so I applied a very small amount of super glue and some electrical tape to put them back in. They seem pretty secure for now. I'm aware that there are some suggested replacements for those lamp sockets from mouser in other threads, but I decided to go the lo-fi route. I may replace them soon so that I don't create a situation where the lamps fall out and cause a short. I also broke the original dial string on reassembly, so I ended up replacing that with a new nylon cord that I bought on eBay (for future reference: the .80mm cord works perfectly for this tuner). The service manual has very good, clear instructions on how to replace that dial string, and it wasn't nearly as difficult as I thought it would be to restring it. I also used deoxit on the pots, switches, and inputs. No noise at all from those pots now. Here's a photo (can anyone spot the mistake I made on reassembly?)

As this was my first time attempting any repair like this, I thought I would at least document a few things I learned that I would have told my past self when I started this project. First, past self, don't ignore the very sage advice to build yourself a DBT. I built one, and it saved me at least once. Second, do a good thorough cleaning of the board before getting started. Use 91% isopropyl and an acid brush (take care to ensure good ventilation when working with isopropyl). Take your time, and inspect every component. Use it as a chance to familiarize yourself with the layout of the board, and pay attention to which transistors are in which spot.

Finally, past self, take your time. I know that you really want your trusty old receiver to work again, but there's no need to rush. It's nearly 40 years old, so taking another week to properly diagnose an issue is well worth it.

Thanks to AK for all your help with this project! I'm already looking forward to my next project: rehabbing my mid 60's era Victrola-branded RCA console. I'd be much obliged if anyone could direct me to the service manuals for those units.
 
Congratulations on successfully completing the recap and reassembly. Is the dial pointer in the wrong place on reassembly. The unit looks great and I hope my turns out as well. I have run into a bit of a problem that I hope someone can help with. I am about half way through my recap and completed the power supply except for the two 8200 caps. When I tried to fit them onto the board the lead sizing was wrong. It appears to be about 16mm and the caps I bought are quite a bit smaller than that. Something I should have looked at earlier. I tried to find the correct lead spacing for these caps on Digikey and others but wasn't able to find the correct size. If someone could point me in the right direction it would be appreciated.
 
Back
Top Bottom