SAE 2600 restoration/recap.....need assistance.

Okay, so I'm finally getting around to repairing and restoring my SAE 2600. It's been a while and may take a bit to complete depending on when I have to ship out again. This amp has had a very hard life on the road, has undergone previous repairs and looks as though it was stored in a garage for many years. Unfortunately I don't have the money to have it professionally restored, but in spite of my lack of experience I think I will be able to do it justice. I do have basic soldering and schematic reading skills but this will still be a challenging endeavor for me.

This will be my first full on repair and restoration of any audio gear so please be patient with me. My background is as a Merchant Marine Engineer and while I do a lot of electrical work aboard ships, my electronics knowledge however is limited but I think I will be able to manage with a bit of help. Please correct me when I use incorrect terminology etc.

I have just ordered the schematics from Stereo Manuals.com so I can't provide those yet. I won't be using ultra high end caps but will be using high quality caps from Nichicon, Elna, or Panasonic. I will try to stay within the same series on each set of boards.

I have done some preliminary testing of the input(?) and driver boards. On both left channel boards all of the to-3 transistors have been replaced with FET transistors. All of the transistors on all four boards tested good. I have not yet tested the output transistors. I've found two resistors that need replacing and another that although tests fine, appears badly burned. There are a couple damaged traces that look like I will need to solder in a jumper as a replacement.

After repairing and recapping the input(?) and driver boards I'll move onto the relay/protection, main board, meter, and lighting boards as well as checking the output transistors.

I do have several nooby questions I'll post after the pictures. I hope you don't mind. :worried:

Left driver board:


The 62Ω 1w resistor is shot and burned so badly that the boards fiberglass weave is exposed on both sides of the board. The trace underneath is damaged to the point I'll need to solder a jumper in it's place.

This leads me to my first set of questions.

1. Are the orange colored resistors like the one pointed out in the picture flame resistant or otherwise a special kind of resistor?

2. On the right driver board, one of the orange resistors has been replaced with a carbon resistor but I can see that it got so hot that it too has exposed the boards fiberglass weave. Should the carbon resistor be replace with the orange original type?

Left input(?)board:



I believe the grey resistors are called precision resistors? Anyway, of the two circled in the photo, they are both supposed to be 750Ω 5w. The one on the left measures 802Ω and the one on the right measures 734Ω.

Round two of the questions.

3. Is this difference significant enough that I should replace the resistors to achieve closer to the specified 750Ω?

4. I'm I incorrect for testing resistors in circuit?

5. Both of the left boards have had their to-3 transistors replaced with FET transistors, is it acceptable to keep the to-3's in place on the right boards or should they be changed for the appropriate FET replacements to match the FET's on the left boards?

6. There are two pots on each driver board. Are they something I should look into replacing or are they not likely to be bad.

7. All I can read from some very poor schematics is that the rectifiers are rated at 25A. I have some new 50A 1000V rectifiers I could use. Is it critical the ratings for these be exact or will my over rated bridge rectifiers work fine?

I know other issues exist beyond these boards and will deal with them once these boards are up to snuff. If anyone thinks improvements can be made with regard to parts materials/types please feel free to let me know. I am certainly open to making any improvements possible. This amp is really "the one" for me. It is the amp I have wanted since I was in my teens and anything I can do to make it shine I will. Any help or input will be greatly appreciated.

edit:
I just placed an order from Mouser and got all Nichicon FG caps for the boards. I also ordered several resistors but I'm not sure if the types of resistors I ordered are correct. The resistors are pretty cheap so if I need to I'll order the specialized resistors.

Not sure if this will help much:
http://amp8.com/tr-amp/foreign/sae/sae2600.htm
Build a dim bulb tester too.
http://antiqueradio.org/dimbulb.htm

Thanks for the links ryuuoh. Yes, the first link has been helping me and I've downloaded all of the pics. I will use a dim bulb tester for sure when I get to the testing phase.

Regarding upsizing the rectifier, going to a bigger one should not be a problem. Resistors, you should measure out of circuit, unless they are super low values, like emitter resistors.
Good luck with that beast, I've heard they are fine amps.

Good, I'll use the new rectifiers. I guess I have my work cut out for me with the resistors. I'll replace the ones that are clearly burned and go from there with the dim bulb tester once I get there. Thanks again.

Second thought:
If the board is burned to the point where the fibers are showing, lay some epoxy on the area. This makes the area stable, and resists moisture.

Wii do. One board has two small spots like that and another board has one spot.

Make sure you clean off any carbon soot from the burn spots on the boards. It will likely be electrically conductive.

Bongiorno recommends replacing the MPSU06/U56 parts with 2sa1837/2sc4793 in a thread on SAE Talk.

I would recommend replacing any resistor that looks compromised regardless of how it tests. No point in leaving a partially burnt part that costs .05 on the board if you're trying to get it back to spec.

To answer a couple of your questions...

2. On the right driver board, one of the orange resistors has been replaced with a carbon resistor but I can see that it got so hot that it too has exposed the boards fiberglass weave. Should the carbon resistor be replace with the orange original type?

Click to expand...

That may be why it failed. The orange may be a 1W or 2W resistor whereas the carbon was a 1 or 1/2W. You'll need to sort out the proper watt value for the resistor there before choosing a suitable replacement. You should find that info on the schematic.

3. Is this difference significant enough that I should replace the resistors to achieve closer to the specified 750Ω?

Click to expand...

If they're all 3% tolerance as the one shows, then 802Ω is a bit out of spec. Wether it makes a difference I have no idea.

4. I'm I incorrect for testing resistors in circuit?

Click to expand...

This depends on the circuit and what you want to know about the resistor. If there's a parallel resistance you'll get an Ωs reading that's inconsistent with the value (always lower) of the particular resistor being measured. If you have a schematic it's usually not too hard to figure out if you're going to get a parallel resistance being measured. If you simply want to know if there is a short or continuity, then reading in circuit should work fine so long as you aren't measuring a 2Ω resistor or something, which most continuity checkers might regard as essentially zero resistance.

FootFungus said:

Make sure you clean off any carbon soot from the burn spots on the boards. It will likely be electrically conductive.

Bongiorno recommends replacing the MPSU06/U56 parts with 2sa1837/2sc4793 in a thread on SAE Talk.

Click to expand...

I will be sure to clean off all of the carbon and I'll look further into replacing the MPSU06/U56 parts as well.

eurekaiv said:

I would recommend replacing any resistor that looks compromised regardless of how it tests. No point in leaving a partially burnt part that costs .05 on the board if you're trying to get it back to spec.

To answer a couple of your questions...



That may be why it failed. The orange may be a 1W or 2W resistor whereas the carbon was a 1 or 1/2W. You'll need to sort out the proper watt value for the resistor there before choosing a suitable replacement. You should find that info on the schematic.



If they're all 3% tolerance as the one shows, then 802Ω is a bit out of spec. Wether it makes a difference I have no idea.



This depends on the circuit and what you want to know about the resistor. If there's a parallel resistance you'll get an Ωs reading that's inconsistent with the value (always lower) of the particular resistor being measured. If you have a schematic it's usually not too hard to figure out if you're going to get a parallel resistance being measured. If you simply want to know if there is a short or continuity, then reading in circuit should work fine so long as you aren't measuring a 2Ω resistor or something, which most continuity checkers might regard as essentially zero resistance.

Click to expand...

Thanks for the information eurekaiv, that makes sense and I will try to make sure the correct values are used. It will be much easier once the schematics arrive.


Today I used wire numbering tags on the labyrinth of wires coming from the transformer and terminal strips etc. as well as drawing diagrams to ensure I can return everything to it's original position. I will double check all of this against the schematics and do my best to understand each circuit.

So I now have almost everything disassembled allowing access to the main board and output transistors. There was a wire on the bottom of the main board that had come disconnected due to what looked like a bad solder joint. I have repaired that connection.

This all leads me to a few more questions with regards to the output transistors.

Here is the picture:


I looked online to find out about testing transistors in circuit and read to use the diode function on the DMM so that's why my readings are in volts. :scratch2: If this is way off please let me know. If it's not way off then it looks like I have 6 to 8 bad transistors. If my data has any value please let me know.

As far as I know there are no FETS in that amp. All transistors are bipolar and perhaps a few Darlingtons. You need to remove the output transistors to test them properly. The b554 d424 pairs are from a previous repair. They look like counterfeits to me.

grateful said:

As far as I know there are no FETS in that amp. All transistors are bipolar and perhaps a few Darlingtons. You need to remove the output transistors to test them properly. The b554 d424 pairs are from a previous repair. They look like counterfeits to me.

Click to expand...

Sorry, I thought the ones in the picture below were FET.



I will remove the output transistors. Would I be better off replacing the T0-3 transistors with their modern equivalents?

The Onsemi MJ21194G and MJ21193G would be great and are better than the originals. They are about $4.00 each. I have heard they can be purchased directly from Onsemi for less. I think you will find a special bias IC on the output transistor heat sink. It may be hard to source. What part did you use for the predrivers?

grateful said:

The Onsemi MJ21194G and MJ21193G would be great and are better than the originals. They are about $4.00 each. I have heard they can be purchased directly from Onsemi for less.

Click to expand...

These are replacements for the output transistors? I am leaning in the direction of replacing all of them with new like the Onsemi mainly if there are bad transistors and partially because some may be counterfeit. Also, I would like to have a better measure of consistency. Plus, as you said, they are likely better than the originals as well. Thank you for the info. :thmbsp:

Yes they are T03 outputs.

You need to replace all output transistors that show 0V DC in the diode test. They are shorted. Try to get replacements from reputable sources such as Mouser or Digikey.

Resistors on the left input board (and probably on the right one too) should be changed all; however, you should look at the schematics to see if they are of "Fireproof"/"flameproof" type and change to same. If they are not fireproof, I would use Vishay 1% metal film with the wattage a bit higher than the originals , such as 2.5 or 3W instead of 2 W etc.

Also change pots to multi-turn Bournes.

Thank you . I will do this.

Those 2SB554/2SD424 look like counterfeits from here. You've got a mixed set of outputs on the right, although it probably doesn't matter if the 2 different pairs are correctly in series. It would be best to replace them all. AVNet/Arrow and maybe Future carry the MJ21193/4/5/6 parts significantly cheaper than Mouser or Digikey. They don't always have each part in stock though.

kotofei said:

You need to replace all output transistors that show 0V DC in the diode test. They are shorted. Try to get replacements from reputable sources such as Mouser or Digikey.

Resistors on the left input board (and probably on the right one too) should be changed all; however, you should look at the schematics to see if they are of "Fireproof"/"flameproof" type and change to same. If they are not fireproof, I would use Vishay 1% metal film with the wattage a bit higher than the originals , such as 2.5 or 3W instead of 2 W etc.

Also change pots to multi-turn Bournes.

Click to expand...

Thank you kotofei, I will follow your suggestions. Resistors are cheap and I really just want to do this right so it lasts for many years to come.

Should all of the resistors be changed on the driver boards as well? While we're at it, I've heard there is usually no reason to change the little ceramic disc caps. Does anyone change them anyway while they are doing the work or would that be totally unnecessary?

Replace them all they seem like they have been thermally stressed. Electrolytic caps should be replaced. Disks should be OK. Of course all the driver transistors have to be all tested and replaced as necessary. You also need to check the pre drivers and every transistor and diode on every board.