Threaded Mode | Linear Mode

Doctor_Jacek · 09-03-2019, 09:39 AM

Goodmorning everyone,

After connecting my T-58C to an external NiMh battery pack and arranging the foam on the keyboard, the calculator came back to life.

If I run 2nd-Pgm 01 Sbr = after 30 seconds the display shows a 1.

If I run Pi (3.14xx) STO 05, I turn off the machine then immediately turn it back on and run RCL 05 I get Pi.

Same behavior if I wait 15 minutes before turning it back on.

All this could suggest that the calculator is working properly.

But if I wait about 30/45 minutes (or more) and I repeat the command RCL 05 I get a meaningless and non Pi number.

What could be the cause of this operation?

Could it be the electrolytic capacitor?

Regards

jlind · 09-03-2019, 03:29 PM

(09-03-2019 09:39 AM)Doctor_Jacek Wrote: Goodmorning everyone,

After connecting my T-58C to an external NiMh battery pack and arranging the foam on the keyboard, the calculator came back to life.

If I run 2nd-Pgm 01 Sbr = after 30 seconds the display shows a 1.

If I run Pi (3.14xx) STO 05, I turn off the machine then immediately turn it back on and run RCL 05 I get Pi.

Same behavior if I wait 15 minutes before turning it back on.

All this could suggest that the calculator is working properly.

But if I wait about 30/45 minutes (or more) and I repeat the command RCL 05 I get a meaningless and non Pi number.

What could be the cause of this operation?

Could it be the electrolytic capacitor?

Regards

If an electrolytic is the mechanism that provides a voltage to perform the RAM keep-alive when the power switch is turned off, that would be my first suspect. I presume you are leaving the external NiMH battery power supply you have connected and are just using the power switch. I don't claim to know what the TI-58C uses for the keep-alive. If it is a mondo electrolytic, at 40 years of age it wouldn't surprise me if it's a bit leaky with the electron pixies slowly dancing across the electrolyte internally. Sometimes electrolytic caps will reform the electrolyte internally to the point of working well if they're left with power applied. Getting back to an "as new" state is another matter. Looked in the hardware docs I have for the 58/59 and there was nothing in them about how the 58C performs the memory keep-alive. I do know a standard 58 will do OK for a few seconds with the battery pack removed to place it onto the printer, but don't dawdle.

I've got a TI-58 (not a "C") and if I need to have it keep its its current state alive, I leave it plugged in to the charger. The original pair of BP-1A battery packs I have for it were modified for removable AA NiCd cells. If you've got original battery packs, you can have them rebuilt for this. I hope you didn't pitch any out. They're impossible to find. eBay sellers typically dump them if they've leaked and then sell the calculator without them - "no batteries" is not unusual. A man in Georgia makes replacement battery packs with a printer but it's better if you have one that can be rebuilt, whether or not the batteries have leaked. Note that the 58/58C/59 should not be operated with the charger plugged in without a battery pack or batteries in place. The batteries are the power supply filter on the full-wave bridge rectifier circuit.

Out of curiosity, what did you use for keyboard foam . . . did you replace it? If so, with what and most important, how thick is it. Mine is in need of it.

John

Doctor_Jacek · 09-03-2019, 04:56 PM

I bought the T-58c on ebay, it had no external power supply.

I was lucky because the battery pack, even if exhausted, was free of leaks.

The batteries were empty and when I turned it on the first time I could only see one '0' for a few tenths of a second, then it went off.

On the advice of Nigel (UK), I opted for an external battery pack loaded with its own charger.

When I turned it on the first time the keyboard tests are unsuccessful results, many bounces and keys that didn't work.

Furthermore, the first test on the module resulted in 0.

I took the whole calculator apart and noticed the lack of foam :-(

I also found a second processor (!!!) soldered to the math processor, maybe it's a repair job.

In desperation, I tried to take a plastic case of a CD, cut it to the right size and inserted between the keys and the keyboard.

As if by magic the keys are working again, even the test on the SS module gave a positive result.

Only the same behavior remains on the memorized variables that appear when switched on a few hours after switching off.

Bye

[attachment=7650]

Nigel (UK) · 09-03-2019, 07:53 PM

Here's a good thread about the TI58c.

Section 2.4.1 in the first post of this thread says that when the calculator is turned off the static RAM chip is powered by a 470uF electrolytic capacitor, and that this capacitor is kept charged via a germanium diode connected to the battery. Looking at the chip specs online gives a typical standby current of 0.2 uA. Without the diode to keep it charged, this current would drain the capacitor in a couple of hours, which is roughly what you are observing. I suggest that the electrolytic capacitor in your calculator is probably fine, but that the germanium diode keeping it charged might need replacement.

Nigel (UK)