PDA

View Full Version : know this diode? IBM 5153 color monitor repair



tezza
August 21st, 2008, 02:25 AM
Hi,

I've finally got around to taking apart my IMB 5153 color monitor (the one I use with my CGA 5150) to see if I could find out why I have to give it a thump on the side occasionally to restore power. In fact it had got so bad, it had simply stopped working althogether.

Within the vertical PSU at the side of the cabinet (the side I used to thump occassionally) I found the following corroded diode.

http://www.classic-computers.org.nz/blog/images/corroded-diode-in-PS-of-5153.jpg

I first checked the diode with a multimeter. no connection at all, unlike the other diodes in the circuit board. When I touched the corroded pin gently it simply came away from the main body of the diode. Bingo! (I hope!). I suspect this is the problem.

But..what are the specs of this diode? It has the code TJ4D on it. A search of the web reveals nothing?...nor do any catalogues I've got show anything that even looks like this. It definitely is a large diode though, as it has the diode symbol underneath it and is called D815 on the circuitboard.

Failing that, does anyone know where I can get a circuit diagram of this PSU card . I have a scanned copy of the technical reference manual but that just has a logic diagram of the main board of the color monitor.

I need to know what the modern equivalent of this (large) diode is so I can replace it? I've never seen anything like it before.

Anyone know what a replacement would be?

There seem to be a bit of capacitor leakage around that area too (which I assume caused the corrosion in the first place). The caps don't appear blown though in that the heads and flat and look fine. Worth replacing or better to just leave them?

Tez

nige the hippy
August 21st, 2008, 12:27 PM
looks like a fairly standard fast recovery diode to me, I found this circuit diagram on the web, but it doesn't seem to correspond with your component numbers.... That said, it's a fairly generic monitor supply, so if you could identify the diode in about the same position on the circuit diagram, I could probably give you a number.
And yes, replace the cap(s) if it's juice is running out it's getting electrically leaky too (and it's capacitance is draining away as well). Things could get fried fishy very soon.
I'm away this weekend, but will try to help!

Denis
August 21st, 2008, 12:28 PM
The brown gunge is actually glue that's decayed and become conductive. Cleaning (chiseling!) all traces of it from the board and components would be a good start, it is quite corrosive as you can see from the diode leg. Agree any 3A fast recovery diode should be ok.

tezza
August 21st, 2008, 02:06 PM
The brown gunge is actually glue that's decayed and become conductive. Cleaning (chiseling!) all traces of it from the board and components would be a good start, it is quite corrosive as you can see from the diode leg. Agree any 3A fast recovery diode should be ok.

Denis,

Am I right in assuming this glue is leaking from the large black upright capacitor (and hence it should be replaced?)

Tez

tezza
August 21st, 2008, 02:12 PM
looks like a fairly standard fast recovery diode to me, I found this circuit diagram on the web, but it doesn't seem to correspond with your component numbers.... That said, it's a fairly generic monitor supply, so if you could identify the diode in about the same position on the circuit diagram, I could probably give you a number.
And yes, replace the cap(s) if it's juice is running out it's getting electrically leaky too (and it's capacitance is draining away as well). Things could get fried fishy very soon.
I'm away this weekend, but will try to help!

Thanks Nige,

I did have that PDF already and couldn't match it to what I was seeing. I'll have another go at interpreting it though.

The diode is a little larger than others. There are other diodes on the circuitboard with the same shape and colour but they are smaller. You can see another slightly smaller (with a red band) on in the lower right of the picture.

These posts are helpful. Anyone else got any more comments?

nige the hippy
August 21st, 2008, 02:30 PM
I'm inclined to disagree with Dennis on the glue. Sometimes caps are glued down but I've never known Evostik cause corrosion, however I have known leaking electrolyte to do so. It's probably worth just lifting the cap just to see what state it's in. Give it a smell too, electrolyte tends to smell fishy. If there's no problem with the cap, just solder it back in.
If it's on the secondary side of the transformer you'll probably see the effects of a duff cap in either twinkling/variations in intensity of the display, or wavy edges.
On the diode, yes, pop something like a UF5408 (3A 1000V 75nS <26 pence!) in and you won't go far wrong. Without the volts/amps trade offs of 20 years ago we can afford to be lazy nowadays!

Denis
August 21st, 2008, 02:38 PM
Honest chaps it's glue- or it used to be glue before it decayed. tezza, put your meter on ohms and measure between two random points on the old glue, you will get a reading I promise. It's not leaking from the cap, it was put there during manufacture to hold it in place.

nige's advice is good though, best to remove the cap and smell it to be on the safe side.

Ec1564
August 21st, 2008, 06:36 PM
That brown stuff was at one time a yellow colored glue to keep the large cap from vibrating off the PC board. I see this in many TV's and other electronics. Most times just picking it off and cleaning with a good ckt cleaner will cure most of that conductive brown glue. It is best to pull that cap off the board and clean under it too. But with age, I'll suggest replacing that cap. They do leak with age.

tezza
August 22nd, 2008, 07:29 PM
Just to wrap up this thread I can tell you the monitor is now fixed and everything works as it should.

http://www.classic-computers.org.nz/blog/images/ibm-monitor-fixed.jpg

I replaced the diode with the 3A 1000V equivalent suggested, and the unit is now resotred to complete working order.

I'm of the opinion that the brown stuff was indeed glue, that had oxidised. One thing that drew me to this conclusion was that it was in a circular pattern near the base of the capacitor, yet the circuit board was actually mounted vertical in the monitor (with the cap horizontal). If the cap was leaking from that position, the residue pattern would have been different.

The capacitor seemed fine (no smell, no sign of leakage). I tested the residue to see if it was conductive but it didn't seem to be. As it would have been a major job to desolder all the surrounding components to clean the board I decided to just leave it. The diode leg hole is the only place that glue touches other components. It's obviously corrosive, but I figure if I have to replace the diode again in the distant future, then that's the lesser of the two evils.

Thanks for the help guys!

Tez

nige the hippy
August 26th, 2008, 01:40 PM
Great stuff, as it happens I've just picked up the 5150 (it wasn't an XT after all!) & 5153 from my sister's house so I could have had a look tomorrow!