Dead HP 35

[krazyk]

So I decided to scrape clean reflux and reflow the solder underneath both of the display ICs.

   

[Hiwi]

Hi Kris,
have you already replaced the cables? Although voltages are in range now, but you'll never know.
And in the picture in your post #38 i still see some green oxid spots.
Please clean up everything thoroughly, please.

Ralf

[Hiwi]

Ah Ok,
everything clean now, but the soldering does not look perfect (for me), please reflow again.
What kind of solder do you use? With or without PB? This can be an issue.
Ralf

[krazyk]

All of this to no avail.

Absolutely NO CHANGE to the signals and it appears that one or both chips are non functional, hence the 45mA power consumption refuses to change. I have checked the easily replaceable components and they all seem to be okay. Continuity seems good everywhere I can check.

Such a shame really ut who knows what abuse this unit has been exposed to.

This is the slightly broken prong on the keyboard connector I had mentioned.

   

I checked continuity on all prongs last week and they all checked out OK including that prong.

Any other ideas guys?

[krazyk]

I would have preferred it was bright and shiny but I AM ALMOST THINKING I need to remove both chips completely.

The solder on the other side is dark brown/grey. I still think the anode IC is not outputting anything intelligible though. It just resolves as a 5volt signal with a little noise on those PHI rails.

I have spent 10 to 12 hours on this so far and its beginning to get really ANNOYING.......

Not good for a hobby really. :-(

Ahhh well thats electronics sometimes thats 50 years old.

[krazyk]

I think its very fine leaded solder but its probably 20 years old

[teenix]

(10-16-2024 10:02 AM)krazyk Wrote:  Any other ideas guys?

If the mess around the Anode/Cathode driver pins prior to re-soldering was solder splatter then it could have caused shorts and damaged the drivers. Not sure how something like that would have got there though.

If you run out of ideas, I'm east on the other side of the Dandenongs, you could pack and post the calculator and I could take a look if you wanted to. I might have a spare set of drivers if I can find the box that they would be in, or I could swap some parts around and try to narrow the problem down.

[krazyk]

This hp35 had definitely been opened b4 I recieved it as the rear label on the back had 2 definite creases.

As you saw the corrosion in the battery box was shocking and there was also fairly bad corrosion/oxidation on the button metal overlay so who knows where this calculator had been stored for 20 to 30 years of its 50 year life.

The external condition was outstanding though, as if it stopped working decades ago and was put in storage.

I would love to confirm though that the LOGIC board worked as it looks as pristine as the outside of the calculator. Its just its keyboard PCB that suffered very badly. I have no idea about the LED components.

This is the keyboard PCB traces now as best as I can remove the green oxidation.

   


Thank you for the offer Teenix as you dont sound far away up on the Yarra Valley side of the Dandenongs.

Does anybody know if there are any Oscilloscope examples of Phi1 and Phi2 either directly from the Anode driver IC or from the cathode output, keyboard INI and INI2 or the PHI1 or PHI2 test points. Being able to compare or know what I should be seeing would help enormously.

Kris

[teenix]

(10-16-2024 12:34 PM)krazyk Wrote:  Does anybody know if there are any Oscilloscope examples of Phi1 and Phi2 either directly from the Anode driver IC or from the cathode output, keyboard INI and INI2 or the PHI1 or PHI2 test points. Being able to compare or know what I should be seeing would help enormously.

http://teenix.org/ClassicNotes.pdf

[krazyk]

After that reflow session of both the Anode and Cathode ICs I think I may have got a measurement of Phi1 from the testpads of the logic board.
   

This definitely looks like a low level clock signal and is reproducable now every time the calculator starts. The vertical scale is 200mV so the clock signal is only around 40 to 50 mV. That sounds quite low.

I think reading it correctly that each square represents 5 micro seconds so 4 pulses occur every 3 squares or 15 microseconds which equates to around 266 khz. I am still learning so please correct me if I am wrong.

Phi2 is simply outputting voltage noise at this stage so something is still very wrong.

What Custom Logic IC spares do any of you have for a HP35 as everything implies some sort of IC failure.

Kris

[AndiGer]

Just to be sure: We suppose you attached the scope‘s ground lead to ground in the calc?
Otherwise you get unpredictable measurements …

[krazyk]

Yep re earth lead location. Used the dual prong location on the logic board as it was convenient.

I could try using the earth test point location but I really need somewhere I can safely use a CLIP so i have a hand free to use the DSO.

Kris

[Geoff Quickfall]

Prior to removing the ICs with suspect contamination, I will share with you an, albeit, anecdotal story.

Picked up an early 12C for $3. Don’t usually mess with 12C as they are still in production, but this was early American build.

The batteries were fused into a single cylinder from leakage. The leakage seeped in the top left quarter of the logic / keyboard PCA. After cleaning as much as possible with vinegar, mechanical picking (dental probe), then testing continuity on suspect traces I managed to resurrect a random display of nonsense on the LCD.

Aha, progress.

After much testing the display driver, affected by corrosion was the suspect. So, because Voyagers keyboards are not made to be disassembled without risking reassembly, I did a Katie Wasserman suggestion.

She mentioned that DeOxit red has an ability as an active deoxidizer, to “repair” ICs by mechanically weeping under the IC and the PCA board. This results in the corrosion being removed, thus isolating any traces that were shorted by the corrosion.

So I added quite a few drops of DeOxit red (squeeze bottle with needle) around the suspect IC and let it weep throughout the underside as an experiment.

Left it over night and the next day had 90% of the display left. By 48 hours the entire calc functioned as new! That was about 5 months ago. It is still fully functional today!

So, before major surgery try the DeOxit trick on the ICs at the display end and on the logic PCA IC’s, particularity around the suspect solder joints and ICs.

Geoff

Here is the link to the post on facebook.

DeOxit red and the HP 12C ca 1987

[krazyk]

Hi Geoff

I purchased and used DeOxit Red in the nail varnish style bottle a while ago and used it to great effect to restore my partially working HP 25 keyboard. I had read that it causes a chemical reaction to neutralise damaging corrosion in hard to get at places like sealed keyboard contacts.
I too left it overnight and after much excercise the key board came back to life.

Re the HP 35, I have reassembled it and put it aside as too hard without any spare parts to further test and eliminate custom IC failure. All continuity around the anode and cathode ICs checked out so the resoldering certainly helped.

I did take some pictures of the general area around the display generation ICs on the HP35 and it looks great really.





   


I started inspection of a HP21 I got mainly for its leather bag with broken zip and thinking I could use it for Pannamatiks Mod but I did NOT EXPECT THIS TOTAL DISASTER.

   

[krazyk]

I wont show you any other pics of the HP21 as it would make you cry.

It had most probably been stored in its bag face down perhaps in a garden shed or garage so that 30 plus years of condensation had settled into the keyboard and display VIAs PCBs and turned 4 or 5 of the gold metal tracks into green flowers. I knew I was in trouble when 5 or 6 of the keys FELT DEAD SOLID when pressed even b4 I started dissassembly.

Interestingly though after painting the seperate raised electronics board in deoxit RED overnight the electronics PCB with the ACT and display ICs looked almost reasonable.

   

So I have had a 1970s HP BAPTISM OF FIRE really and its truly shown me the difficulty in purchasing these items on Ebay. The HP21 cost a LOT more than $3 thats for sure so my luck truly ran out.

Ahh well thanks for the advice and Teenix if you read this, if you could look thru your collection of parts, if you see any HP35 components or would like to have a look at my HP35 please let me know. I am happy to purchase them to try and ressurect one more piece of HP history.

Kris

[teenix]

(10-23-2024 10:40 PM)krazyk Wrote:  Ahh well thanks for the advice and Teenix if you read this, if you could look thru your collection of parts, if you see any HP35 components or would like to have a look at my HP35 please let me know. I am happy to purchase them to try and ressurect one more piece of HP history.

You are welcome to send both of them and I will take a look, only costs a bit of postage.

I have replacement Woodstock main boards, assembled example image attached, and if you try Harald Pott, (his email should be in the Forum contacts list) he may still have replacement CPU boards for the 35.

cheers

Tony

[Hiwi]

but he needs a new display for his 21...

[teenix]

(10-24-2024 06:41 AM)Hiwi Wrote:  but he needs a new display for his 21...

I wasn't sure if he tested the LEDs after cleaning it. If cleaned properly and the tiny internal wiring is intact, I would think it should work.

One problem with these displays, is if corrosion gets under the LED chips that are glued to the circuit board before wire bonding. The glue is conductive and bonds to a ground plane. Sometimes the chips break loose from the board and the digit stops working. It is wise not to drop these aged displays lest the digits break loose. I have used conductive glue to get one working again. You have to be really careful not to move the LED chip around too much or the tiny wires can fracture at the weld joints.

cheers

Tony