Threaded Mode | Linear Mode

GanjaTron · 02-05-2023, 06:59 PM

Dear all,

I got an HP35 in bad shape very cheap and decided to take on the challenge to revive it. (Read: I gotta get a life!)

I dunno what the previous owner did to this calc, but it suffered severe corrosion, presumably from a leaked battery pack. As a result, two tracks (RCD and Vss) on the display/keyboard PCB were corroded through. I soaked the whole thing in citric acid (coffee machine descaler) and distilled water to neutralise and clean up the mess, and repaired the broken traces with a combination of copper tape and solder (to keep the profile flat). AND IT WORKED... for 15 minutes or so. Then it just suddenly locked up with a stuck display. When I power it up now the display faintly flashes for a few msec and stays dark.

I've consulted Jacques Laporte's excellent website on the HP-35, and verified that all derived voltages (Vcc, Vgg and Vss) are in spec, and that both clocks (phi1 and phi2) are present, ditto SYNC. The keyboard matrix is also being scanned via the R1-R8 C1-C5 lines. However, there is no activity on the D0-D4 lines going to the anode driver, so nothing is displayed. The IA and IS lines to/from the ROMs are stuck in a fixed pattern, even immediately after powerup, so there's no activity there.

Is there anything further I could check or should I just give up and assume one of the ROMs or the C&T chip has failed? (To be fair, the logic PCB looks remarkably clean and free from corrosion).

Any hint appreciated. Many thanks and have a good week,

--Roland

teenix · 02-05-2023, 10:02 PM

Are you able to determine what ROM address the IS bus is stuck on. This 10 bit data is inside the SYNC pulse, LSB first. An example can be found on page 16 of my Classic Notes

It might give a clue if a ROM has failed.

If any ICs get warm after a few minutes of operation that might mean it is faulty.

You might consider changing the capacitors in the power supply.

If the board has failed and you cannot replace the faulty part, Forum member Harold Pott may be able to help with a replacement CPU board.

cheers

Tony

GanjaTron · (This post was last modified: 02-11-2023 09:09 PM by GanjaTron.)

Following up on this, here's a snap of what's going on in the IA line (top) and the IS line (bottom), with the scope externally triggered by Φ1:

Note that this pattern never changes, even immediately after powerup. Compared to the traces shown in Tony's Classic Notes and Jacques Laporte's website, that doesn't look right at all! The ROM address on the IA line (apparently all 1's here??) should alternate with the fetched instruction on the IS line below. My understanding is that IA and IS should never be active at the same time. But instead, IS is constantly oscillating, just like Φ. Infact, it actually appears to be shorted to the latter, presumably inside one of the ROMs or the A&R chip which buffers and interprets IS. That's my theory anyway.

So from this, and the fact that I've run out of ideas, I conclude this calc's logic board indeed buggered itself beyond repair after just 15 minutes of normal operation. So I may get in touch with Harold -- thanks for the hint!

--Roland

teenix · (This post was last modified: 02-12-2023 12:03 AM by teenix.)

Even if the IA bus was outputting all 1's, it should look similar to the PHI buses.

A 2.2uF capacitor controls the circuitry reset at switch on, pretty sure top RHC of board, it may need replacing, say with a modern tantalum 16 - 25V.

cheers

Tony

GanjaTron · 02-12-2023, 12:42 AM

(02-11-2023 11:24 PM)teenix Wrote: Even if the IA bus was outputting all 1's, it should look similar to the PHI buses.

A 2.2uF capacitor controls the circuitry reset at switch on, pretty sure top RHC of board, it may need replacing, say with a modern tantalum 16 - 25V.

Hi Tony,

yeah, it's that 2.2µF at upper right. Checks out ok, but I can try with an electrolytic as a quick test.

--Roland

teenix · 02-12-2023, 02:04 AM

I don't know if you detected a warm IC after being on for a minute or so. If one is warm it is probably defective.

GanjaTron · (This post was last modified: 03-15-2023 01:09 PM by GanjaTron.)

Hi,

just to recap on this (sorry for the pun): changing the reset cap did nothing, and no chips got warm. The calc is now reassembled as a display piece for the time being.

Will contact Harold. Thanks again,

--Roland

Harald · (This post was last modified: 03-15-2023 07:18 PM by Harald.)

Hi Roland,

I am still waiting for new PCBs.
Since you got the calculator cheap you might not want to invest a lot of money to get one of my boards. If it is of interest to you I could have look in the parts box, maybe I have a working 35 board that I removed from a calculator when installed one of my boards.

Cheers,
Harald

GanjaTron · 03-15-2023, 09:14 PM

(03-15-2023 07:17 PM)Harald Wrote: Hi Roland,

I am still waiting for new PCBs.
Since you got the calculator cheap you might not want to invest a lot of money to get one of my boards. If it is of interest to you I could have look in the parts box, maybe I have a working 35 board that I removed from a calculator when installed one of my boards.

Thanks, Harald. Let me know when you get round to it. Failing that, I'm also ok with donating this calc to anyone who needs the still useful parts (display board, keycaps) for the cost of shipping.

--Roland

John Garza (3665) · 03-15-2023, 10:05 PM

Any chance the components were not sealed tightly?
Could acid leaking into a chip cause some damage?

GanjaTron · 03-15-2023, 10:23 PM

(03-15-2023 10:05 PM)John Garza (3665) Wrote: Any chance the components were not sealed tightly?
Could acid leaking into a chip cause some damage?

Possibly, tho the CPU board is actually relatively free of corrosion -- and the calc _did_ have its 15 minutes of fame. ;^)

--Roland