Any Osborne I gurus?

[cosam]

The "bad" chip is marked NEC D416C
The replacement (same position on the other board) is marked AM9016FPC 8228GMP

Is that a match?

It would certainly appear so; both are 16K x 1 DRAMs and being used in the same places on similar (identical?) boards is a good sign.

I would however second Dave's tip of reflowing the solder first, especially as the thing was working not long ago and only started playing up after that long drive.

It's of course up to you, but if it does need replacing, I'd suggest tracking down a loose replacement RAM chip, unless the donor machine is a complete write-off. Removing an IC from a board intact can be tricky. If you do decide to do it, it's worth practicing on a junk board first. There are some tips on the wiki.

PS - by "piggybacking" do you mean simply holding one chip over another so all the upper chip's legs touch the lower chip?

That's the way, yep. Incidentally also on the wiki ;-)

 

[NutmegCT]

Cosam - thanks very much for all the pointers.

You mention "reflow" - I can't find that in the wiki. Does that mean letting a bit of solder flow down on each pin? Or does that mean just touching a hot solder iron to each pin to "re-set" what's there?

I can't see any non-shiny solder points under that chip - which I think is what Dave was referring to in his reflow suggestion.

Thanks.
Tom

 

[cosam]

You mention "reflow" - I can't find that in the wiki. Does that mean letting a bit of solder flow down on each pin? Or does that mean just touching a hot solder iron to each pin to "re-set" what's there?

You'd probably get away with just melting what's there and let it solidify, although feeding in a touch of new stuff will often help the old solder flow more nicely, giving you a better joint.

I can't see any non-shiny solder points under that chip - which I think is what Dave was referring to in his reflow suggestion.

A non-shiny ("dry") joint would definitely suggest a problem, but it's not always all that obvious. As reflowing a few pins is so quick and easy, it's worth a try. Worst case scenario: you need to replace the chip after all. Even then reflowing wouldn't be a complete waste of time as the "fresh" solder will likely be easier to remove than old stale stuff.

 

[NutmegCT]

Touched a hot electronics iron to each solder point (under the board) on that chip. Saw the solder "melt" and quickly moved away.

Ran the entire memtest series (refresh, galpat, v-bar, diagonal) again. Results:

Refresh - 4799 errors
Galpat - 25188 errors
V-bar - 14706 errors
Moving Diagonals - 4027

And all the errors are at board location UB:27.

Also did a "continuity" test between similar pins on the UB:27 chip and other chips. Continuity all good - which to me means there are no breaks in the circuit at the tested points.

So my beginner's brain says there's a problem in the chip at the UB:27 location. And my brain also asks: Is there a way to simply "lock out" that failed chip from use?.

Again, it's obvious I'm new to all this. Is there any other test I should do before trying to remove the failed chip and solder in a replacement?

Thanks.
Tom

 

[cosam]

So my beginner's brain says there's a problem in the chip at the UB:27 location. And my brain also asks: Is there a way to simply "lock out" that failed chip from use?.

Again, it's obvious I'm new to all this. Is there any other test I should do before trying to remove the failed chip and solder in a replacement?

Sounds pretty definitive to me. I'd go ahead and replace the chip at UB:27. Even if you did somehow lock the chip out, you'd be missing one bit from every byte in the 16Kbyte it's part of, which I don't think will get you very far ;-)

 

[dave_m]

Tom,
I'd hate to see you have to rob parts from a vintage Osborne board that might be fixable later.

I recently repaired a Commodore PET which had a bad 4116 RAM chip. I have some left over chips and 16 pin sockets. Send me a private message to me (dave_m) on this forum with contact info so I can send you the needed parts.

 

[NutmegCT]

OK - here goes. I'm heating up my iron to remove a good chip from my spare m/board. This is the board that won't boot up at all - just a garbage screen. I may have to remove several of those chips for a piggyback test, 'til I know I've got a good one.

Question: How do I remove a good chip? I understand the removal of a bad one: cut off (I assume with a Dremel and a cutting wheel) each leg of the bad chip, then de-solder each leg separately from the m/board so I have spot with empty holes.

But I'm pretty foggy on removing a good chip, i.e. whole. Do I somehow work on the underside of the board, heating each pin bottom 'til the solder "falls" out? Then when all have "dropped their solder", turn the board over and tap the chip out?

It seems that de-soldering from above would just melt the solder back into place, as I can't see how to "pull" the chip up with all the rest of the pins still soldered in place.

I did check the Wiki and other links, but they seem to concentrate on removing a *bad* chip - not on removing a good one.

Thanks.
Tom

 

[NutmegCT]

Dave - I just sent you a PM.

 

[barythrin]

Yup, that's the right concept on piggy backing a chip. er.. ok I replied to the last post on page two. Nice. Anyway, I'd suggest getting a desoldering iron http://www.radioshack.com/product/index.jsp?productId=2062731, I know there are solder suckers but they're a little more of a pain to use. This one is nice although you do have to be careful not to burn the traces or PCB.

Alternatively if you can push the solder around and away from the pins/hole you can sometimes use some non-rough screw driver action to get the chip off the board. I suppose some folks would cut the legs but I'm not big on that since fresh solder and connection joints would be the ideal situation. It'd be ashame to solder a new dip onto old legs and bad solder joints.

 

[dave_m]

I suppose some folks would cut the legs but I'm not big on that since fresh solder and connection joints would be the ideal situation. It'd be ashame to solder a new dip onto old legs and bad solder joints.

That's not quite what we meant. One cuts the legs away from the body with a small diagonal cutter tool. This allows one to unsolder the legs one at a time without overheating the board. You then clean the holes with solder wick before one installs the new chip.

 

[tezza]

Question: How do I remove a good chip?

Tom,

In your case there are two things you want to do....

1. Removing the faulty ram. Clip the legs off the IC, using fine snippers then desolder each leg from the board one at a time. After they are removed, clean up around the holes with solder wick.

2. Extracting a good IC from the second board.

This is tricker than 1.

I've used solder wick for this. Do a leg at a time. Start from the back side of the board where the pin protrudes out. Press the solder wick against the solder on the pin. Put the hot iron on the wick until the solder flows into the wick (you'll notice when this happens). On the first attempt, you probably wont get all the solder but don't hold the iron there too long. Just a second or two. Cut off the used wick and try to collect more solder with another attempt. Work through all pins. Then flip the board over and do the same thing with the legs as they go into the board as there is probably a little solder around them too.

Then you'll probably need to revisit the bottom of the board again and maybe the top too. Once you think you've got all the solder, get the edge of a screwdriver under the IC and just gently try and lift it. Don't force it. If it doesn't move, there is still more solder to remove.

It's laborious, exactling and takes a long time.

I would strongly, strongly advise you to try this first with both a board and an IC you don't want just for practice.

I would also advise that you don't use your other Osborne board as the source unless you are SURE you want to turn it into junk board. A good Osborne board is probably salable. If it shows garbage characters then it may be simply a single RAM IC which needs replacing on that board. This was the case with mine. You may find you soon have the skills to repair it.

All the IC's I've removed for REUSE (hence I extracted rather than just clipping them off) have been from junk boards. Hence my only concern was the IC, not the board or the tracks. Unless you are very skilled and have all the right tools, its almost impossible not to wreck the board when you are trying to extract a chip. You will probably destroy a good Osborne board which only needs a single IC replacement.

It's up to you, but if it was me, I would take up Dave_m''s offer even if it meant waiting a few days for the chip to arrive.

Good luck anyway

Terry

 

[NutmegCT]

Terry - this is excellent. Thanks for the detailed instructions.

I've contacted Dave and hope to receive the items I need in a few days. I had figured that the first board was pretty much useless; as I didn't have a way to test/diagnose/repair it, I thought it was usable for parts only. Thus I used the 1.4 ROM from it to replace the 1.3 on my working board, and was able to make the working board functional for double density. Now I'll have that non-working board relatively untouched - and may be able help someone here one day.

Have a good weekend everybody. I'm an 1830s New England farmer at Old Sturbridge Village on Saturdays and Sundays; great retirement job if ever there was one.

Tom
www.osv.org

 

[NutmegCT]

Wednesday update: Yesterday I got two new 4116s and sockets. I tried the piggyback test. With the original chip, I was getting about 4000 refresh errors in a cycle of refresh tests (FOG diagnostic disk). With the new chip pressed down over the old chip, I got about 700 refresh errors on that same testing cycle. The diagnostic still indicates the error in exactly the same location as before I piggybacked the chip - UB:27.

Removed the replacement, and tried the second replacement chip. Pretty much the same result: much fewer errors than before piggybacking, but they're still in the same location.

Obviously a new chip makes a difference. But I had (naively?) thought the new chip (piggybacked) would eliminate all errors.

Does it still seem that replacing that original chip will end the problem? Should I go ahead and remove it, put in a socket, and add the new 4116? Or is the continuing (but reduced) error problem evidence that the new replacement chips are faulty also? or even that there's still another problem?

Paranoid in Connecticut.
Tom

 

[cosam]

I think we can safely establish that the original chip has had it and needs to be replaced, so I'd go ahead and do that. Piggybacking won't uncover every kind of fault, but the fact that it made any difference at all is encouraging. If you're lucky, the RAM replacement will be all that's necessary (and even if it isn't, it most likely needed doing anyway).

 

[dave_m]

Tom,
I agree with cosam. With piggybacking, one is only looking for a change for the better. It is difficult to maintain 16 solid connections with this technique for the entire test.

 

[NutmegCT]

11:42am EST update:

Carefully clipped off the pins of the faulty chip. Using copper wicking, got about half of the board holes cleaned of solder. But there are a few holes where the solder is just inside the hole - in other words, not flush with the surface top or bottom.

Could someone give me a pointer on how to get that tiny bit of solder out? I've been using my 25w iron and the copper wicking, but the solder in the hole isn't touching the wicking - so there's no heat transfer. I'm sorely tempted to try to drill it out with a Dremel, as poking it with a needle or toothpick isn't successful.

Thanks. Seems I'm 90% there.
Tom

 

[cosam]

Just drop a bit of new solder in/over the hole and you should be able to pull the old stuff out with it.

 

[dave_m]

Tom, please follow cosam's advice. Do not use a drill as you might damage the plated-through hole and printed circuitry. Solder wick works by capillary action and so the wick must be in contact with all the solder.

I do not know if you have removed the system board from the chassis. If so you can always try to remove the solder by working on the flip side of the board.

 

[NutmegCT]

I used Cosam's advice and dropped a bit of solder on each of the problem holes. Finally got the problem holes cleared; sure seemed there was some mighty persistent solder down in a few of those holes, so I had to do several cycles of the "solder drop and wick" bit.

Checked for continuity to the adjacent IC to be sure there was no damage - all OK. Carefully laid in the socket and soldered each pin. Checked continuity again. Carefully pressed in the new IC and checked continuity to adjacent IC.

Reinstalled all the lines, plugged in the power cord, girded my loins and flipped the red power switch.

Silence.

Not even a beep.

Totally flustered on this. I've checked all the lines - can't see anything I'm forgetting. Again checked continuity of new IC and adjacent IC. Even removed the new IC and left the socket empty.

Still total silence on power. Fuse ok. Power up but no drive light, no CRT filament.

If I've now got a working system board, but the PSU has just decided to give up the ghost, I'm not a happy camper.

Geez Louise.
Tom

 

[NutmegCT]

The fun continues. I disconnected every cable from the logic board, removed it. Decided to test the power outputs (5v, 12v) from the PSU.

Plugged into wall, flipped red power switch - flash, zap, smoke and sparks from R5 on the PSU. R5 is a 1K half watt resistor. Powered off immediately. Said a few choice words.

I cannot believe this is all happening - especially when this morning I could boot up the system fine (altho' with faulty ram). Now I've got the ram replaced, and the fog has descended once again.

T.