SAE 2400L Restoration

sabrefencer

sabrefencer

AK Subscriber
Subscriber
I'm getting ready to restore a very dirty and improperly repaired SAE 2400L Power Amp. I have it ripped apart and have started to clean it, and have inventoried the parts on each board.

Before I place my parts order I have a few questions about the following components:
1) I'm looking for a good replacement for the TO-220 transistor on the Relay board. This is an NPN type S1375, which is rated at 80v. I found that MJE182G is similar (or higher) on all specs. Is this there a "go-to" for replacing the S1375, or is the MJE182G a good choice?
2) I have not found any source for the relays. They are 2 pole with 48v coil. The pin-out is 6 pins in two rows, pretty evenly spaced - see photos. They work, but are in rough shape and missing the covers. Any info on part numbers and sources for replacements?
3) The main (output) board looks like it might not have been messed with and has no obvious issues. I tested the original Toshiba 2SB554/2SD424 and they test fine. I am planning to re-install these with new mica and paste. If there is a good reason not to do this and to install new outputs, let me know.
4) For the TO-220s on the driver board I am planning to use MJE032/33 for the VAS, and the same for the drivers. I have seen MJE028/29s spec'd for the drivers, but I have the 32/33s. I assume that the higher voltage rating on the 32/33s can't hurt. Again, let me know if this is a bad idea.

@hamrules and @llwhtt, you guys were very helpful on my recent 2200 rebuild. I've done lots of research, and there is a bit more info available on the 2400, so hopefully I won't need as much help with this one, but I
definitely look forward to your guidance along the way.

Thanks,
Gary
0113232001.jpg 0113232315a.jpg 0113232315.jpg 0130231238a_HDR.jpg 0131230202.jpg
 
Register to hide this ad
Relay: LY2-0-DC48 should do the trick.
S-1375/1376: Last ones I’ve used were CENU07/57,
same pin out too.
Outputs: I’d keep if they all test good.
TO-220’s: Those MJE15032/33’s are my fav’s, even though they are overkill.
 
2) For the relay drive transistor I'm sure you can just use a TIP31C or whatever. That MJ is too small.
3) I would use silpad's over mica. The only good thing about replacing outputs with MJ21195/6 is that the ratings are much better, and I have seen blown 2SD424's.
4) Perfect.
 
Do you have the Service Bulletin about the wrong polarity of the 220uF/16V capacitors? The silkscreening of the capacitor on the right side of the board is wrong, the -dot should be on top NOT the bottom. The left side is on the bottom and is correct. I have a 2400L here with one board silkscreened correctly and the capacitor was still installed wrong and the other board has incorrect silkscreening AND the capacitor installed wrong. The amp must have been built on Monday morning or Friday afternoon. Looks like both of your boards are silkscreen incorrectly, can't tell if the caps are correct though.

MJE15034/35 work great as the VAS and MJE15028/29 for the drivers. The VAS doesn't need the 8A of the other MJExxxxx.

Craig
 
llwhtt said:
Do you have the Service Bulletin about the wrong polarity of the 220uF/16V capacitors? The silkscreening of the capacitor on the right side of the board is wrong, the -dot should be on top NOT the bottom. The left side is on the bottom and is correct. I have a 2400L here with one board silkscreened correctly and the capacitor was still installed wrong and the other board has incorrect silkscreening AND the capacitor installed wrong. The amp must have been built on Monday morning or Friday afternoon. Looks like both of your boards are silkscreen incorrectly, can't tell if the caps are correct though.

MJE15034/35 work great as the VAS and MJE15028/29 for the drivers. The VAS doesn't need the 8A of the other MJExxxxx.

Craig
Click to expand...
Craig,
Yes, I have the SB -- you sent it to me ;). You are correct that my board has both + dots at the bottom of the board and both caps are installed with (-) at the dot. On one board they look like they might be factory (gray Elna 220uF/16v caps), on the other board, one cap was replaced with a 220uF/16v, and the other one was replaced with a 470uF/16v. This is on side with the incorrect polarity marking and it is installed to match the (incorrect) polarity marking. On top of that, on the same board the 100uF/50v has been replaced with a 330uF/63v. This board was definitely hacked.
 
ivandezande said:
2) For the relay drive transistor I'm sure you can just use a TIP31C or whatever. That MJ is too small.
3) I would use silpad's over mica. The only good thing about replacing outputs with MJ21195/6 is that the ratings are much better, and I have seen blown 2SD424's.
4) Perfect.
Click to expand...
2: I’ve used 2SC4883A, MJ’s should work too.
3: Silpad’s sure do save the mess. 93/94 95/96 would do fine.
 
Because the heatsinks are so narrow it's hard to get the heatsink clean and then try to be neat with mica/grease is a PITA. I've switched to Silpads, K52 by Laird is what I use.

Those SanKens along with some other current SanKens will work great also. This version of the 2400L circuit board has Metric studs so the Fullpacks will fit. The earlier boards will need the regular metal tab transistors.

Using a 330uF capacitor in that spot may cause the front end voltages not to stabilize fast enough and may cause problems. Ham knows the electrolytics, C26 and C27, were replaced with ceramics in the SUMO "THE POWER" Current Sources. I'm thinking the Current Sources were not stabilizing fast enough with the electrolytics. I DO bump the voltage on that capacitor up to 63V since having a 50V capacitor across a 51V Zener didn't make sense to me. It's only 1 volt though and may not make a difference since they've lasted this long.

Craig
 
Last edited:
llwhtt said:
Because the heatsinks are so narrow it's hard to get the heatsink clean and then try to be neat with mica/grease is a PITA. I've switched to Silpads, K52 by Laird is what I use.

Those SanKens along with some other current SanKens will work great also. This version of the 2400L circuit board has Metric studs so the Fullpacks will fit. The earlier boards will need the regular metal tabs transistors.

Using a 330uF capacitor in that spot may cause the front end voltages not to stabilize fast enough and may cause problems. Ham knows the electrolytics, C26 and C27, were replaced with ceramics in the SUMO "THE POWER" Current Sources. I'm thinking the Current Sources were not stabilizing fast enough with the electrolytics. I DO bump the voltage on that capacitor up to 63V since having a 50V capacitor across a 51V Zener didn't make sense to me. It's only 1 volt though and may not make a difference since they've lasted this long.

Craig
Click to expand...
What are these Silpads you mentioned. I used to use Laird but the ones I had I can't get any longer, they were great !
 
Mouser, 739-A15036008 for TO-3 and A15037005 for TO-220. There were other ones I used to use but they went from under a buck to over 6 bucks now, Bergquist SPK10-0.006-00-05. at Digikey. Then the king daddy of them all are the beryllium insulators I save those for very special occasions.
 
Craig, Ham, and Ivan this is great input.
I'm definitely going to use silpads on the outputs for all the reasons given.
The S1375, has an EBC pin-out. After comparing datasheet for the suggestions above, the MJE182G seems to be a good match and has the right pin-out:
S1375 Replacement.JPG

Ham, thanks for the tip on the relays, after looking at the datasheet for these, they are DPDT with 8 pins. It looks like two of the pins need to be removed to make it a 6 pin DPSP relay which will then drop in.

Getting back to Craig's post #4, My driver boards are version SAE 17-0310 Rev A, and have six 100uF/10v caps. I have seen photos of drivers with eight 100uF/50s, and the schematic I have for the MK24 has 8 100uF/10v caps. I have noticed a few differences between my boards and the 2400L schematic, and between the some component markings on my board and the component installed. Some of these differences align with the MK24 schematic, and others with the 2400L schematic. It appears that my boards are a bit of a hybrid between the two.

Here are the differences I have found so far:
A) The Zener (CR3) is installed in parallel with the 100uF/50v cap (C5) which matches the MK24 schematic but does not match the 2400L schematic - and I will definitely upgrade to a 63v cap. The Zener installed on my boards is in a TO-92 package. I’m thinking of replacing it but I’m not sure if I should leave it in, or if it is a good idea to replace it with a Zener in a modern DO-35-2 package. I was planning to use the same Zener that I used in my 2200, which is a 1N5262.
B) R19 is shown as 68k on the 2400L schematic, but both of my boards are marked 100k. On one board, R19 was clearly replaced with a 68k, and on the other board it looks to be a factory installed 100k (matching the marking). I think R19 in the 2400L corresponds to R121 on the MK24 schematic, which is 100k. My plan is to put new 100k on both boards.
C) C47 is shown on the 2400L schematic as a 470p (in parallel with 1.8k R1). On both of my boards there are ceramic 470p caps installed, but the boards are marked "001uF". I don’t see a similar cap on the schematic for the MK24, and can't figure out why the boards are marked "001uF". My plan is to replace these with new 470pF multilayer ceramics.
 
llwhtt said:
Mouser, 739-A15036008 for TO-3 and A15037005 for TO-220. There were other ones I used to use but they went from under a buck to over 6 bucks now, Bergquist SPK10-0.006-00-05. at Digikey. Then the king daddy of them all are the beryllium insulators I save those for very special occasions.
Click to expand...
Wow those beryllium are costly, yikes.
 
That S-1375 isn’t even a 1amp device. Many choices depending if you want to twist them or not.
 
I've never encountered a TO-92 diode, and was wondering if there was any issue to replace it with an axial package. I'm planning on replacing everything that is why I 'm asking.
Regarding the S-1375, I'm just not that good at interpreting transistor data sheets. Is there something about the MJE182 that makes it a bad choice?
 
I’ve done those TO92 diodes with axial ones, no problems.
I think it’s easier in the long run just to replace it all or almost all anyway.
I’ve never tried the MJ182, can’t say it wouldn’t work.
 
Last edited:
I have the same boards in front of me now with pretty much the same parts you described. Rebuild yours the same as they are, whether you use the 68K or 100k will depend on the adjust range for the VAS current, if you go with 100K on both boards there should be enough range on the trimmer. The biggest difference between the MK2400 and the later 2400L is VI limiter transistor type and where it's connected in the circuit. The VI limiter xsistor is the second one from the top on the driver heatsinks. The MK2400 uses a TO-66 or TO-220, Q111/112, and it's collector is tied to the emitter of the VAS transistor. The 2400L uses a TO-92, Q9/10, and it's collector is tied to the collector of the VAS transistor thru CR10/11. I never really understood why they did the MK2400 that way, Ampzilla I also, when the limiter is activated it's pretty much tying the power supply to ground thru the 62 Ohm resistors. Note that the limiters are activated when the amplifier is trying to drive very high loads or the output is shorted this takes the output of the amplifier to basically ground level. In the 2400L the VI limiter xsistor is tied to a much lower voltage and more typical of most VI limiter circuits.

Another difference between the various versions is sometimes that 100uF/50V capacitor is floating, no gnd reference, and sometimes it's tied to ground.
 
Thanks Craig. Your explanation was very helpful. I think I have all of the needed parts for the 2400L plugged into my next Mouser order.

This morning my MK XXX preamp decided to drop a channel. . I'm opening it as I am typing this. Not sure if it is a repair or a restoration yet. I've also got a 9B that hums that is on the restoration list. I'm older than this SAE equipment and so far I'm in better shape ;).

Gary
 
Since my last post in this thread two months ago, I've been busy -- Had my kitchen remodeled, rebuilt my MK XXX preamp, and graduated one of my daughters from college. Now am back to my bench and the 2400L project.

Work done so far:

Display board -- I replaced the three 4.7uf 50v caps and all 30 LEDs. All of the resistors and diodes checked out, so I left them. About 4 or 5 of the LEDs were out when powered directly, so I replaced them all to make sure the color and brightness was consistent.

Gain switch board -- cleaned (it was really gunky) replaced all resistors, diodes, and cap, and cleaned the switches.

Now working on the relay board. I noticed that the four small electrolytic caps don't match the 2400L schematic, and/or the board. Caps installed on the board are 100uf 10v, which match the board markings, but the schematic shows these (C37/C38) as 100uf 50v. The other two, installed are 1uf 50v, which matches the schematic (C35/C36), but the board is marked 4.7uf 50v. Also the (5W) resistor on the back of the board is 680 ohms 5W, and the board is marked 680, but the schematic shows this (R71) as 1k 5W. The larger cap (C41) is 100uF 50v, and will be replaced with a 100uF 63v.

This seems like a bit of mess, but it all looks factory installed-- see photos in post #1. Looking for advise on replacements for the following:
C37/38 -- my original plan was to replace the 100uF 10v caps on the board with 100uF 16v. Do these need to have a higher voltage rating?
C35/C36 -- I'm leaning toward using 1uF 50v here, since that is what is installed on the board, and what the schematic says. So I will ignore the 4.7uF 50v marking on the board.
R71 -- I have a 680ohm 5W ohmite 45F to install here.
 
The protection board I have in front of me has 100uF and 1uF, factory. These four capacitors along with the 10K resistor form a low pass filter, shunts all but the lowest frequencies and DC to ground. Any really low frequency and/or DC will trigger the protection relay.
 
You must log in or register to reply here.
Back
Top Bottom