Threaded Mode | Linear Mode

vikram86 · 03-10-2015, 01:09 AM

Hi everyone!

I'm trying to repair a broken HP 32E I just received on eBay with a friend.

We opened it up and rewired corroded connections.

When I connect batteries (2xNiMH) the display lit up all zeros and did not respond to key presses. But after a couple attempts, we could not replicate the straight zero display. Instead, what we are getting now is just an "apostrophe" in the upperleft corner of the screen when we move the switch from on to off.

Any idea what is wrong? I know about the thread with the 25C, the ACT chip being fried...but this isn't a Woodstock - so I'm not sure what could be the problem.

Thanks!
--Vikram

vikram86 · 03-10-2015, 02:50 AM

So, we figured out the problems.

1. There was corrosion on the battery terminals - this was fixed by bypassing them completely with 3 new wires.

2. The small IC (third from the top) had one leg broken off completely. My friend soldered a wire to it, and soldered that to the board.

3. Because of this extra wire in (2), the other ICs were not contacting the board properly. A bit of Scotch tape on the plastic frame to get them more snugly on there, and voila. The calculator powered on like nothing had ever happened.

The jury is still out on how that single pin was broken clean off the IC.

--Vikram

brouhaha · 03-10-2015, 09:22 AM

(03-10-2015 02:50 AM)vikram86 Wrote: The jury is still out on how that single pin was broken clean off the IC.

I've seen that same failure mode on several of the press-fit 30-series calculators.

PANAMATIK · 03-10-2015, 03:28 PM

(03-10-2015 09:22 AM)brouhaha Wrote:
(03-10-2015 02:50 AM)vikram86 Wrote: The jury is still out on how that single pin was broken clean off the IC.

I've seen that same failure mode on several of the press-fit 30-series calculators.

I have a non functional press-fit HP-33E where the 1820-2105 processor Pin 7 is intentionally cut, because there is no signal trace leading away from this pin. But there is another 8-pin chip labeled 00031 A0008SGP where Pin 8 is completely off, no chance to solder a wire. So I have to give up repairing this calculator the easy way?

At that unfortunately I was clumsy when opening the case and the three 8-pin chips were falling out suddenly, so I don't remember in which order they should be replaced. They are labeled MA10003, MA10004 and the above.

Can anyone help?

Bernhard

ElectroDuende · 03-10-2015, 03:37 PM

(03-10-2015 03:28 PM)PANAMATIK Wrote: I have a non functional press-fit HP-33E where the 1820-2105 processor Pin 7 is intentionally cut, because there is no signal trace leading away from this pin. But there is another 8-pin chip labeled 00031 A0008SGP where Pin 8 is completely off, no chance to solder a wire. So I have to give up repairing this calculator the easy way?

When that happens to me, and the chip is not easy to find (that I suppose is the case) what I do is using a "dremel" tool to remove a bit of the plastic packaging of the chip exposing about a milimeter of the pin base where a wire can be soldered.

A high power, temperature controlled, thin tip soldering iron is very recommended, but it doesn't guarantee a 100% success... anyway, it seems that you have nothing to loose.

Dave Frederickson · 03-10-2015, 03:50 PM

(03-10-2015 02:50 AM)vikram86 Wrote: A bit of Scotch tape on the plastic frame to get them more snugly on there, and voila. The calculator powered on like nothing had ever happened.

Scotch tape can generate a high static charge and cause Electro Static Discharge (ESD) damage. Low-Static Kapton is preferable.

PANAMATIK · 03-10-2015, 03:53 PM

(03-10-2015 03:37 PM)ElectroDuende Wrote: When that happens to me, and the chip is not easy to find (that I suppose is the case) what I do is using a "dremel" tool to remove a bit of the plastic packaging of the chip exposing about a milimeter of the pin base where a wire can be soldered.

A high power, temperature controlled, thin tip soldering iron is very recommended, but it doesn't guarantee a 100% success... anyway, it seems that you have nothing to loose.

Good idea. Thank you. And the number of permutations for the order of the three chips is only three factorial 3!=6. I will try that.

Bernhard

Marcus von Cube · 03-10-2015, 10:42 PM

(03-10-2015 03:53 PM)PANAMATIK Wrote: And the number of permutations for the order of the three chips is only three factorial 3!=6. I will try that.

They might work in any order if they are all wired in parallel and decode their addresses themselves.

PANAMATIK · (This post was last modified: 03-11-2015 11:42 AM by PANAMATIK.)

(03-10-2015 10:42 PM)Marcus von Cube Wrote:
(03-10-2015 03:53 PM)PANAMATIK Wrote: And the number of permutations for the order of the three chips is only three factorial 3!=6. I will try that.

They might work in any order if they are all wired in parallel and decode their addresses themselves.

You are right! I should have known it by myself. I made the experiment with a working 31E, which has two chips and I interchanged them. It worked in both positions.

Then I tried to repair the HP-33E. I "dremel"d the defective chip and soldered a wire to the board but without success.

Then I exchanged components from the 31E to 33E and back to find the defective part. It was indeed the (I assume RAM) chip 00031 A0008SGP. I replaced it by 00031 B0008SGP from 31E and my HP-33E works completely!

But I need another 0031 A0008SGP for the HP-31E now.

Thanks for your suggestions.

Bernhard

PANAMATIK · 03-11-2015, 12:14 PM

Success!

I tried again the original chip of the 33E and discovered, that there was just a contact problem. When I pressed the chip against the board it worked and now I made it that it has contact without force.

Cheers to all from a happy 31E and 33E owner
Bernhard

ElectroDuende · (This post was last modified: 03-11-2015 12:37 PM by ElectroDuende.)

(03-11-2015 12:14 PM)PANAMATIK Wrote: Success!

I tried again the original chip of the 33E and discovered, that there was just a contact problem. When I pressed the chip against the board it worked and now I made it that it has contact without force.

Cheers to all from a happy 31E and 33E owner
Bernhard

Congratulations!

You did the dremel trick quite well... I wasn't so successful on my first attempt!

Regarding the contact problem... I suppose it is something one can expect; the calculator may be working properly, but once open, it is easy to touch the IC's pins or PCB pads and cause local corrosion (well, just a thin "patina") that may affect conductivity. And, as the chips are just pressed and don't slide (like HP48GX cards), the patina is not "self cleaned".

vikram86 · 03-11-2015, 01:38 PM

What my friend did was rub an eraser on the traces to make sure they were clean.

vikram86 · 03-11-2015, 01:40 PM

(03-10-2015 09:22 AM)brouhaha Wrote: I've seen that same failure mode on several of the press-fit 30-series calculators.

How do you think it happened? We didn't see the leg floating around inside the body of the calculator at all.

PANAMATIK · 03-11-2015, 01:40 PM

(03-11-2015 12:35 PM)ElectroDuende Wrote: Regarding the contact problem... I suppose it is something one can expect; the calculator may be working properly, but once open, it is easy to touch the IC's pins or PCB pads and cause local corrosion (well, just a thin "patina") that may affect conductivity. And, as the chips are just pressed and don't slide (like HP48GX cards), the patina is not "self cleaned".

Because it was not "self cleaned" and the pads had small black patina spots, I used this brass brush "dremel" tool to remove the patina from all pads and pins. Also the display contacts had to be brushed. This was very helpful. May it last another 30 years now.

Bernhard