Is my TI-58C hosed?

[Dave Britten]

I've got a 58C that I snagged from ebay a few years back, and finally got around to restoring it to operation. Mainly that just required building a battery pack: a 3x AAA receptacle soldered to the two contact pads on the board, with a trio of NiMH cells inserted (didn't want to spend a lot on a more sophisticated battery until I could be sure it works). I don't have the A/C adapter, so I'm not worried about the chemistry being different.

It worked great pretty much immediately when I did this last week, and I messed around writing a couple small programs, and trying stuff in the Master Library module. All the included diagnostics looked good, and I could get a couple hours useful life from the AAA NiMHs.

Now, a couple days after using it last, I suddenly can't store any program steps or register values. Everything reads back as 0. Also, the Master Library diagnostic routine (2nd Pgm 01 SBR =) returns a flashing 1. I checked the batteries, and tried a spare set, so it isn't low on power, I don't think. Is the memory chip or the constant memory circuit fried?

[Paul Berger (Canada)]

Have you tried removing the batteries and letting it go completely dead? Also make sure you battery holder is not shorting anything everything is exposed inside the battery compartment.

The 58C contains two RAM chips and with the default partition it should be a 50/50 split between registers and program memory, so from what you describe it would suggest that there is a problem with both of them or the problem is elsewhere. There is a scan of a service guide for the 59 available, it is the same hardware except it has 4 RAM chips and the extra circuitry for the card reader. A search "ti59 service guide" should bring up a few hits. There is also some useful information on Joerg's datamath.org site in particular a block diagram at http://datamath.org/Sci/WEDGE/JPEG_TI-59...is%20block If you need parts let me know I have a few boards from 58 and 59 that are in various states of disrepair.

[Dave Britten]

Yeah, I let it sit unpowered overnight and got the same results this morning.

A bit more weirdness: If I enter LRN mode and start mashing the 1 key to fill program memory, at step 145, it starts saving properly, except for a couple of holes that stay zero. This pattern of ones and zeroes seems to repeat in a cycle of 8 steps. I haven't tested the higher storage register numbers yet. I tried reflowing some of the solder on the Toshiba CMOS RAM chip in the bottom right corner of the board, but that didn't change anything. If there are two RAM chips, I'll have to hunt for the other one and try it again.

[Paul Berger (Canada)]

Hmmm well now that I actually look at the board in a 58C I see that they are in fact quite different in layout, it would seem the service guide is referring to the earlier TI-58 that did not have continuous memory, so the one you see is likely the only RAM chip. I have 2 58Cs here and the board layout in the two is quite different, one I think is likely the same as yours as it has the same TC5047 RAM chip in it, the other would seem to be a slightly older one going by the date codes on the the chips and it has a house number 18 pin chip for RAM and the CD2400 and CD2401 are stacked . Since the TC5047 is a regular 1Kx4 SRAM chip with separate 10 bit parallel address bus and 4 bit data bus additional logic would be required to interface it to the processor and it seems that the TMC0591 that is immediately below it provides that function. The regular pattern like you are seeing might be an address issue such a a hot bit on the address bus or it could also indicate an internal problem with the RAM chip such as a defective row or column in the array of memory cells.

[Dave Britten]

I'm leaning more toward a failure in the RAM chip itself, at this point. It seems like if it were an addressing problem, it would still read/write in a deceptively normal fashion, but would actually be stuffing data in the wrong part of the chip. (i.e. Storing something in reg 00 would appear to work, but the stored value would also appear at a different address.)

There's still a tiny bit of battery residue near a couple solder points not far from the RAM chip, so I'm wondering if stray current taxed the chip and killed part of it. Every now and then when I would turn it on, it would behave extremely erratically, showing C, or flashing other nonsense patterns. I'd have to power cycle it a few times to restore its wits.

Seems like the easiest fix at this point would be to swap the main board with one from a 58C with a damaged keyboard or screen. That would require a nice evening with the ol' desoldering braid, though.

[Paul Berger (Canada)]

I would not use solder braid you would be better off with a decent pump style solder sucker, I use an OK DP-100 it is fairly big but has lots of suction some of the smaller ones are not so good. Recently I picked up a used Unger 4000 Hot Vac and that works even better.

[Paul Berger (Canada)]

I would not use solder braid you would be better off with a decent pump style solder sucker, I use an OK DP-100 it is fairly big but has lots of suction some of the smaller ones are not so good. Recently I picked up a used Unger 4000 Hot Vac and that works even better. You need to be a bit careful with the amount of heat you apply too I found the pads lift fairly easy on these boards.

[Paul Berger (Canada)]

I was just playing with my two 58Cs a bit and I noticed on the one that seems to be the same version as yours I was seeing the same sort of odd stuff in program memory when it was on the AC adapter alone I then put it on the printer cradle and initially I was seeing the same thing there until I clear program memory (CP) and then it started to behave. I also noted that the diags would fail as well but worked fine with it on the cradle after I had done a CP so maybe there was something stuck there that CP cleared out.

[hp41cx]

If you want this gift.
I'm in Brazil.
You only pay shipping

[jebem]

(04-01-2014 12:57 AM)Dave Britten Wrote:  I've got a 58C that I snagged from ebay a few years back, and finally got around to restoring it to operation. Mainly that just required building a battery pack: a 3x AAA receptacle soldered to the two contact pads on the board, with a trio of NiMH cells inserted (didn't want to spend a lot on a more sophisticated battery until I could be sure it works). I don't have the A/C adapter, so I'm not worried about the chemistry being different.

It worked great pretty much immediately when I did this last week, and I messed around writing a couple small programs, and trying stuff in the Master Library module. All the included diagnostics looked good, and I could get a couple hours useful life from the AAA NiMHs.

Now, a couple days after using it last, I suddenly can't store any program steps or register values. Everything reads back as 0. Also, the Master Library diagnostic routine (2nd Pgm 01 SBR =) returns a flashing 1. I checked the batteries, and tried a spare set, so it isn't low on power, I don't think. Is the memory chip or the constant memory circuit fried?

Any further development on this issue, Dave?

Thanks,
Jose

[Dave Britten]

Actually, yes. I'm pretty sure it was a failed memory chip, or something that's part of memory I/O. I ended up grabbing another one from ebay (the 58C is dirt cheap), and everything seems normal. Normal, but slow.

[jebem]

(10-24-2014 10:06 PM)Dave Britten Wrote:  Actually, yes. I'm pretty sure it was a failed memory chip, or something that's part of memory I/O. I ended up grabbing another one from ebay (the 58C is dirt cheap), and everything seems normal. Normal, but slow.

Thanks, Dave.

I have the same issue with mine.
The previous owner left the batteries inside and it leaked badly.
I have dismantled the machine piece by piece, put everything in the kitchen sink in warm water and dish soap for 15 minutes, then used a soft brush to remove the battery remains from the PCB and plastics.

Two days later I powered it on and it worked normally, except that it doesn't store in the registers or in the program steps. Just get zeros when reading.

However the current consumption is about 220mA, which seems too high for me, but I do not have another machine to compare.
Do you have a multimeter capable of reading in the 200mA region to check yours?
I would be very grateful.

A few things I have checked:

The voltages from the DC-DC power supply for the Texas IC's are fine:
Vdd = -10VDC (Pin 2 at CROM socket);
Vgg = -16VDC (Pin 1 at CROM socket);
Common point = Vss at +Vbatt. (Pin 8 at CROM socket).

Also the power supply for the TC5047AP-1 is fine:
Vdd = +4.8VDC (Pin 20);
Common point (GND at TC IC pin 8);

The large 470uF capacitor near the DC-DC power supply is a energy storage for this 5VDC SRAM IC supply, allowing the battery to be removed without losing the SRAM contents.

So my next step is to order one SRAM chip.

[DANTONVB]

(10-24-2014 10:31 PM)jebem Wrote:  Thanks, Dave.

I have the same issue with mine.
The previous owner left the batteries inside and it leaked badly.
I have dismantled the machine piece by piece, put everything in the kitchen sink in warm water and dish soap for 15 minutes, then used a soft brush to remove the battery remains from the PCB and plastics.

Two days later I powered it on and it worked normally, except that it doesn't store in the registers or in the program steps. Just get zeros when reading.

However the current consumption is about 220mA, which seems too high for me, but I do not have another machine to compare.
Do you have a multimeter capable of reading in the 200mA region to check yours?
I would be very grateful.

A few things I have checked:

The voltages from the DC-DC power supply for the Texas IC's are fine:
Vdd = -10VDC (Pin 2 at CROM socket);
Vgg = -16VDC (Pin 1 at CROM socket);
Common point = Vss at +Vbatt. (Pin 8 at CROM socket).

Also the power supply for the TC5047AP-1 is fine:
Vdd = +4.8VDC (Pin 20);
Common point (GND at TC IC pin 8);

The large 470uF capacitor near the DC-DC power supply is a energy storage for this 5VDC SRAM IC supply, allowing the battery to be removed without losing the SRAM contents.

So my next step is to order one SRAM chip.

First, my excuses for revive an old thread, but I have the same problem as you had. Have you find an solution?

My problem:
I have a Texas Instruments TI-58C (made in Brazil). I bought this calculator new in Brazil on April 22, 1981.
It has been stored without batteries for 20 years or more. Recently I took it out to use it.

My calculator is displaying strange memory and program storage behavior. Sometimes it works as expected, but most of the time it does not store the correct number. For example, if I store the number 99.99 in a memory, it returns something like 9999000090.

Do you think you may have a damaged memory IC? My calculator has a NEC D445LC memory IC, as opposed to the common ToshibaTC5047AP-1. I think in replacing the D445LC for a ToshibaTC5047AP-1 IC. What do you think about? Or may another problem?

Thanks,

Danton
Brazil.

[jebem]

Olá Danton, como vai?

Yes, I fixed my TI-58C because luckily it was the SRAM failed chip that one can still easily find as new old stock on eBay.

Please see this:
Texas TI-58C: (in)famous constant memory issues

Good luck with your rebuild.
We do this for fun, right?

(10-01-2018 03:05 AM)DANTONVB Wrote:  

(10-24-2014 10:31 PM)jebem Wrote:  Thanks, Dave.

I have the same issue with mine.
The previous owner left the batteries inside and it leaked badly.
I have dismantled the machine piece by piece, put everything in the kitchen sink in warm water and dish soap for 15 minutes, then used a soft brush to remove the battery remains from the PCB and plastics.

Two days later I powered it on and it worked normally, except that it doesn't store in the registers or in the program steps. Just get zeros when reading.

However the current consumption is about 220mA, which seems too high for me, but I do not have another machine to compare.
Do you have a multimeter capable of reading in the 200mA region to check yours?
I would be very grateful.

A few things I have checked:

The voltages from the DC-DC power supply for the Texas IC's are fine:
Vdd = -10VDC (Pin 2 at CROM socket);
Vgg = -16VDC (Pin 1 at CROM socket);
Common point = Vss at +Vbatt. (Pin 8 at CROM socket).

Also the power supply for the TC5047AP-1 is fine:
Vdd = +4.8VDC (Pin 20);
Common point (GND at TC IC pin 8);

The large 470uF capacitor near the DC-DC power supply is a energy storage for this 5VDC SRAM IC supply, allowing the battery to be removed without losing the SRAM contents.

So my next step is to order one SRAM chip.

First, my excuses for revive an old thread, but I have the same problem as you had. Have you find an solution?

My problem:
I have a Texas Instruments TI-58C (made in Brazil). I bought this calculator new in Brazil on April 22, 1981.
It has been stored without batteries for 20 years or more. Recently I took it out to use it.

My calculator is displaying strange memory and program storage behavior. Sometimes it works as expected, but most of the time it does not store the correct number. For example, if I store the number 99.99 in a memory, it returns something like 9999000090.

Do you think you may have a damaged memory IC? My calculator has a NEC D445LC memory IC, as opposed to the common ToshibaTC5047AP-1. I think in replacing the D445LC for a ToshibaTC5047AP-1 IC. What do you think about? Or may another problem?

Thanks,

Danton
Brazil.

[DANTONVB]

(10-01-2018 08:33 PM)jebem Wrote:  Olá Danton, como vai?

Yes, I fixed my TI-58C because luckily it was the SRAM failed chip that one can still easily find as new old stock on eBay.

Please see this:
Texas TI-58C: (in)famous constant memory issues

Good luck with your rebuild.
We do this for fun, right?

Olá José Mesquita,
Yes, it's just for fun. I used the TI-58C during the Engineering College, from 1981 to 1984. It helped me a lot.
After that, I bought an 11C.
I will replace the RAM chip. I just found it (Tc5047ap-1, Dip 20 Pin) at MercadoLivre.com.br for R$ 26.40 (US$ 6.57) for 5 pc shipping to Brazil included.

Thank you, abraços, Danton