What's new
What's new

Resurrecting 1960 10EE with 460V Sabina drive

May have spoken too quickly.
Diode's are not all marked the same - though they may be equivalent?
Diodes on heat sinks (1 is farthest left when facing the units end as initially pictured)
1 = SW15PCN075 0025
2 = semikron 9025U SKR1M40/14. D0-5
3 = semikron 8929U SKR 1M40/14 D0-5

There is also an additional board. Diode:
SW15PCN075 9728
Suspect this is the same as the one on the first heat sink - and that the last 4 digits may be a lot number of some kind.
full
 
Be careful when scraping those resistors. There is only paint covering the resistive material. You could easily scrape too deep and alter the resistance.
 
The Semikron diodes are 50 Amp at 1,400 volts. The 9025U might be the data code, 25th week of 1990. The 3 diodes on the 3 heat sinks should be the same. If they are not, then likely some were replaced in the past.
Do the diodes test OK?
If you need to replace the diodes they should be easily available. So long as they meet or exceed the current and volt rating. Be aware that diodes were made in both polarities. Check the diode symbol for the ones that you have. The "R" in the Semikron diodes part number stands for reverse polarity.
 
Be careful when scraping those resistors. There is only paint covering the resistive material. You could easily scrape too deep and alter the resistance.
Thanks.
I used an extremely light touch- the camera allows me to see. Re-tested - resistance remains 2.5KΩ.
Good heads up tho - had no sense of it.

The Semikron diodes are 50 Amp at 1,400 volts. The 9025U might be the data code, 25th week of 1990. The 3 diodes on the 3 heat sinks should be the same. If they are not, then likely some were replaced in the past.
Do the diodes test OK?

Repairs - perhaps - but - weird that 2 would be reverse and two would be normal?
Perhaps rather than repaired - original builders simply used different brands/ sizes because they wanted reverse ones for those two positions?
Does it make any sense that heat sink 2&3 are reversed but heat sink 1 is normal?
Or if it was repaired - someone used the wrong direction diodes - but it was unaffected?
(it ran well on an RPC for more than a decade until this issue).

That said - some surprise from the test.
#1 heat sink - no burn marks/ not involved in short circuit - has a bad diode. Caught me off guard. Tested 3x: .001V in both directions. No voltage drop/O.L. reading like the others.
Diode #2 tests .379 in one direction/O.L. in other
Diode #3 tests same (.389)
And the diode on the non heat sink tests .470/O.L - but I have it still soldered to components.

So I guess the question is - what am I shopping for here?
Considering trying following for low cost test:
- Use Mr_CNC's SCR's
- Buy single diode to replace dead one in exact specs
- Plug in 3 new LM741CN's
- See if it fires up.

Then determine if I want to properly rebuild the SCR box with new components.
Keep in mind - I am a weekend warrior - using this a handful of hours/month.
Of course - don't want to trash it - precious to me...

Thoughts?
 
Last edited:
The 3 diodes on the 3 heat sinks are almost certainly the same polarity. Different manufactures construct their part numbers in different ways. Look on the diodes for the diode symbol. That should confirm their polarity.
Does this system run from 3 phase power? If so, that is why you have 3 heat sinks. The 3 phases will have 3 identical circuits that drive them. That can help in testing when you apply power. The voltages on the 3 circuits should be the same. Any significant difference can point you to the problem.
If a diode was replaced in the past with a diode of the wrong polarity, you would know it. It would act like a dead short.
I would replace the bad diode that you found. The other ones are likely OK.
Put the SCR's that I sent you in and give it a try. Ensure that there are fuses between the Sabina drive and incoming power. Give it a try and you might be lucky. These SCR's and diodes are pretty robust (except for when you short them out).
If it does not work, try swapping the 741 chips.
There are a number of small diodes on the circuit board, have you checked them? Most of the time you can check them in-circuit and get usable readings. One that is shorted will stand out. One that is blown open will probably test bad, but the results can be confusing.
 
the white square components in the drive with 1k on them are power resistors, they're used (somehow) to set the output current of the drive. Something to do with a feedback circuit. Either way, if they look melted, they probably saw alot more current than they should have. I'd suggest swapping them out. You can get similar ones for KB DC drives.
 
A short like you had is like dropping a hammer into a machine. You do not know where it will land, but it will land hard someplace!
Ha! Indeed it did. Good analogy.
The 3 diodes on the 3 heat sinks are almost certainly the same polarity. Different manufactures construct their part numbers in different ways. Look on the diodes for the diode symbol. That should confirm their polarity.
I'll take a closer look.
TBH - I wondered at first if the R in SKR stood for SemiKRon rather than Reverse.
Quick good search however turned up the same exact diode model in SKN.
I'll look for the markings to see if they show.
The two SKR models is a mystery until then.
Does this system run from 3 phase power? If so, that is why you have 3 heat sinks. The 3 phases will have 3 identical circuits that drive them.
Yes. 3P. Ran on an RPC for at least 15 if not 20 years - the last 2 of which were in my possession.
Moved into a proper shop with 3P, bought a used transformer to up the power, and it ran for about 5 min beautifully - before it didn't...
That can help in testing when you apply power. The voltages on the 3 circuits should be the same. Any significant difference can point you to the problem.
Great tip. Tested the 3 legs coming in before start up. ~426V ea. Had no idea how or where I would have tested inside the machine before this whole walk.
There is a shut off box on back of machine with 3 @ 20amp fuses - so yes fused.
If a diode was replaced in the past with a diode of the wrong polarity, you would know it. It would act like a dead short.
I shot some images earlier.
I'm pinching the lug end.
I simply snipped wire rather than unsolder - at the time of removal I had assumed I would find all bad.
You can see the polarity marked here on this one which was the SKR.
full

There are a number of small diodes on the circuit board, have you checked them?
As soon as I can figure out which one the diodes are... I'll be all over it.
🙂
 
the white square components in the drive with 1k on them are power resistors, they're used (somehow) to set the output current of the drive. Something to do with a feedback circuit. Either way, if they look melted, they probably saw alot more current than they should have. I'd suggest swapping them out. You can get similar ones for KB DC drives.
Thanks Matt. Appreciate you weighing in.
I'll take a closer look at those components tomorrow - and perhaps take a slightly damp rag to them to see if it is burn or dust.
 
I have marked out the diodes and transistors on your board. I think that I got them all.
RED - ordinary diodes, the printing on the board is D7 for example.
BLUE - zener diodes, the printing on the board is DZ7 for example.
YELLOW - transistors, the printing on the board is Q7 for example.
The diodes and the zener diodes will both test as diodes. Often the diodes can be tested while they are still soldered into the board. The transistors generally cannot be tested in the board. You can get cheap multimeters that have a transistor test socket. You would have to remove them to do the test.
Removing components from circuit boards requires some care to avoid damaging the copper traces. A solder sucker helps, but require some skill to get good results. There are videos that show how to use them. I use one of these;
Here is a Chinese knock-off;


Sabina_drive.JPG
 
the white square components in the drive with 1k on them are power resistors, they're used (somehow) to set the output current of the drive. Something to do with a feedback circuit. Either way, if they look melted, they probably saw alot more current than they should have. I'd suggest swapping them out. You can get similar ones for KB DC drives.
They do look somewhat like ceramic power resistors, but they are not. Ceramic power resistors are square in one dimension, these are rectangular. Ceramic resistors do not melt, they blow up!
 
The black device at the top left of the board is probably a full wave rectifier. It will have 4 connections if so. They, too, can be tested as diodes. I missed that.
All in all, there is nothing on this board that is mysterious and cannot be fixed. Between a voltmeter and swapping parts, you should be able to get it running.
 
Great.
I will board test all that can be today.
Desoldering iron is in my future - classic "I need a tool for this" moment for me.
Am sure we've all been there...

Looks to me like the Semikron diode's are not reverse based on printed direction on side?
Thumb is pinching threaded lug end in this photo - which is apparently also example A of why I need a desolder gun...
full




EDIT* - Mouser sells a slightly higher rated diode for the heat sink. 1600 Volt 85 Amp. Hard time finding anyone who sells the 1500V 75Amp version. Assume it will work for my purposes?
 
The Mouser part will work OK. Note that the diode symbol is pointing down, reverse polarity. You must get a diode with the symbol in the same direction. Getting one with the same stud size makes things much easier. If you use some thermal grease, that will help control the temperature.
I am assuming that you have a circuit something like this;
If so, then the 3 SCR's and the 3 diodes should be the same. If they are not the same, then you have something else. By "the same" I mean that the SCR's point in the same direction and that the diodes point in the same direction. They don't have to be the exact same part.
 
The Mouser part will work OK. Note that the diode symbol is pointing down, reverse polarity. You must get a diode with the symbol in the same direction. Getting one with the same stud size makes things much easier. If you use some thermal grease, that will help control the temperature.
I am assuming that you have a circuit something like this;
If so, then the 3 SCR's and the 3 diodes should be the same. If they are not the same, then you have something else. By "the same" I mean that the SCR's point in the same direction and that the diodes point in the same direction. They don't have to be the exact same part.
Clear.
When I got into the shop today - did two things.
1. Compared diodes (both brands) from the 4 positions on the board. Though not denoted in the SW diode's name(Westcode brand), the symbol is in the same direction for both it and the Semikron.
Think that supports your first theory - a repair has been done.
full

My guess - the Westcode (SW) were there original. They look to be better quality. Only a guess.
2. I tested the SCR's as well. All three tested as good.
full


However - the SCR's are normal - meaning reverse polarity to the diodes.
full


My plan of action - up for debate:
- Get the bad diode replaced with like style reverse in same or better spec.
- Unplug & replace 741's * (noted you had suggested that as second step - any reason I wouldn't just plug in to start?)
- Utilize existing SCR's. Now that I've tested - unless there is reason to believe the test isn't valid?

Board testing is next. Thanks for posting the pic. & definitions.
 
It makes sense that the SCR's and diodes would have the opposite polarity from each other. That way, there need be only one heat sink per phase, as is shown in the schematic at the link I provided.
I like to replace one thing at a time, that way I know that "I got it" and that one part was the bad one. Another reason is that good parts and not identical even if they have the same part number. Their specifications will vary, even though they are within tolerance. This matters, especially when circuits have adjustments as yours does. The board was calibrated in the factory, which will have accounted for the variations in the component values. It is best not to mess with that unless you have to.
 
Board test results.
Not certain I am testing correctly - may take another round.
First number is always negative lead to end of diode with bar.
Some Std diodes tested at 2.6 when leads were swapped (positive lead to diode bar end).
Some Std diodes tested at O.L.
Some tested differently all together (orange boxes).
See below:
full


Zener diodes would not test with diode tester setting on multimeter.
Moved to std OHM setting to get their numbers.
 
The diodes that have a reading of about .6 in one direction are probably good. I am surprised that the zener diodes did not test properly using the diode setting of your meter. The diodes that have a very low reading are suspect. But the only way to know for sure is to unsolder one end and test them that way.
I would replace the power diode that you know is bad and give the whole system a try.
 








 
Back
Top