Threaded Mode | Linear Mode

Xorand · 03-31-2019, 09:54 PM

(03-30-2019 02:44 PM)Maximilian Hohmann Wrote: Very hard to find and expensive is the Ti-58C. Still missing in my collection. I have a Ti-58 and several 59s, but have so far been unable to get a 58C for a price which I am willing to pay.

I've had little problem finding 58C's in the US and at a good price. The 58C in my photo was purchased along with excellent case, adapter, original box, original books (minus the big book for the Master Library Module), and an unused pad of programming sheets for $35 US (plus shipping).

I had recently purchased another in a lot with a couple less desirable calculators along with several accessories. I cleaned it up and rebuilt the battery pack. I gave it to a good friend for his birthday (he's young enough that the calculator was older than he is).

**rprosperi** · 04-01-2019, 12:19 AM

(03-29-2019 09:58 PM)Gene Wrote: Re: the TI-66, I disagree about it's speed vis a vis the TI-58C / TI-59.

From the loops of addition speed benchmark:

TI-66
Count: 210
Code: + 1 = RST

TI 58C
Count: 387
Code: + 1 = RST

TI-59
Count: 635
Code: 1 + RST

I believe the TI-66 despite having alpha key legends in programs and LCD and two-three years younger was a BIG step back in speed compared to the 58/59.

I did not realize the 59 was so much faster than the 58; about 1.6 times faster here, but was it that much faster in general? And are these counts for 1 minute?

badaze · 04-01-2019, 01:08 AM

(03-29-2019 09:58 PM)Gene Wrote: ....
I believe the TI-66 despite having alpha key legends in programs and LCD and two-three years younger was a BIG step back in speed compared to the 58/59.

Indeed. However, consider TI 66 appeared in 1983 after TI 88’s cancellation. In my opinion TI 66 was meant to compete with HP 11C not to replace TI 58/59. Then in 1986 the TI 95 (note the figures Wink

) came. It had little success although it was a good machine.

**Gene** · 04-01-2019, 01:34 AM

(04-01-2019 12:19 AM)rprosperi Wrote: I did not realize the 59 was so much faster than the 58; about 1.6 times faster here, but was it that much faster in general? And are these counts for 1 minute?

Gene: Yes, for one minute. The TI-59 was that much faster in general.

Garth Wilson · (This post was last modified: 04-01-2019 02:20 AM by Garth Wilson.)

(03-31-2019 02:13 PM)toml_12953 Wrote:
(03-29-2019 01:46 AM)Garth Wilson Wrote: I never, ever, had any problem with the card reader. 0.00% failure. I also never ran the head-cleaning card. I never touched the magnetic surface with my fingers though.

How did you get the cards out of their very tight vinyl sleeves in the storage cases? I always had a bit of finger oil on the edges of my cards from extracting them. I never had a read or write error, though. In fact I still have a 59 and it reads and writes perfectly every time.

Oh, yes, I forgot. I'd grab it by the smallest bit I could. I don't know how close any data come to the end, but I'm sure I was not getting any skin oils on the recorded part. When the card was out of the "wallet," I held it only by the edges.

Garth Wilson · (This post was last modified: 04-01-2019 02:26 AM by Garth Wilson.)

(03-29-2019 01:54 PM)Maximilian Hohmann Wrote: Hello!

(03-29-2019 01:46 AM)Garth Wilson Wrote: But of course it's not RPN, and the keyboard is kind of awful.

But it had AOS which I personally prefer over RPN (if I am allowed to write that here...). And the keyboard was indeed a bit bouncy, but no calculator keyboard ever made before or after that time matched that of the pre 1980 HP calculators.

That was Dave Britten who wrote that, not me. As far as AOS versus RPN goes, I could have gone either way for a short time after I had both; but RPN quickly won out when it came to controlling instruments on the workbench and taking data from them, which is what I really got the HP-41cx for, with HP-IL and the various interface converters. For many years now I have wanted to never see an algebraic language again. I sure liked the looks of the 58c and 59 though!

Thomas Okken · 04-01-2019, 02:36 AM

(04-01-2019 01:34 AM)Gene Wrote:
(04-01-2019 12:19 AM)rprosperi Wrote: I did not realize the 59 was so much faster than the 58; about 1.6 times faster here, but was it that much faster in general? And are these counts for 1 minute?

Gene: Yes, for one minute. The TI-59 was that much faster in general.

Did they use the best chips in the 59 and the marginal ones in the 58, allowing them to run a faster clock in the pricier model? I think I read somewhere that HP did something like that in the 48 series.

However, the benchmarks you used above for the 58 and 59 are not the same. How much is the difference when you run exactly the same code on both machines?

xerxes · 04-01-2019, 01:12 PM

(04-01-2019 02:36 AM)Thomas Okken Wrote: However, the benchmarks you used above for the 58 and 59 are not the same. How much is the difference when you run exactly the same code on both machines?

Comparison of the TI-58, TI-59 and TI-66 with exactly the same test code (n-queens) to the other indirect addressing keystroke programmables from TI with almost the same test code:

Code:

 - 1:50:00   TI-66         

 -   58:20   TI-58C      

 -   56:15   SR-52          

 -   43:40   TI-59          

 -   18:25   TI-88     

 -    3:53   TI-95

Xorand · 04-07-2019, 01:20 AM

(04-01-2019 01:12 PM)xerxes Wrote: Comparison of the TI-58, TI-59 and TI-66 with exactly the same test code (n-queens) to the other indirect addressing keystroke programmables from TI with almost the same test code:

Code:

- 1:50:00 TI-66 - 58:20 TI-58C - 56:15 SR-52 - 43:40 TI-59 - 18:25 TI-88 - 3:53 TI-95

I decided to do a side-by-side comparison of my own. I used the program for determining the factorial of a number (i.e. x!) as listed on page IV-74 of the Personal Programming book. I typed the program into each of my two calculators (TI-58C and TI-59), verified they were identical and set them both up to compute 69!

The times were 30.88 seconds for the TI-59 and 38.88 seconds for the TI-58C. This gives a factor of about 1.27 to 1. The stats posted in the quote were around the same ratio (about 1.33 to 1) between the two calculators.

Xorand · (This post was last modified: 04-16-2019 01:25 AM by Xorand.)

And then there were three. I was looking for another 58C on the auction site and had one pop up for a cheap buy it now price. Looked in pretty good condition, so I nabbed it. Though the auction listing said 58C, it actually was a 58. And it works really well. So now I have a 58, 58C and 59.

One thing I had noticed when I got the 58C is that its display was a little dimmer than the 59. I subsequently found that the display on the 58 is dimmer still. I think there is a component (or components) that the internal power supply has that might need refreshing. (All three are running newly rebuilt battery packs, so it isn't battery voltage). I'll dig up my service manual and give them a check over the next couple days.

[Image: i-TnspJL8-L.jpg]

Jlouis · 04-17-2019, 06:10 PM

(04-16-2019 01:23 AM)Xorand Wrote: And then there were three. I was looking for another 58C on the auction site and had one pop up for a cheap buy it now price. Looked in pretty good condition, so I nabbed it. Though the auction listing said 58C, it actually was a 58. And it works really well. So now I have a 58, 58C and 59.

One thing I had noticed when I got the 58C is that its display was a little dimmer than the 59. I subsequently found that the display on the 58 is dimmer still. I think there is a component (or components) that the internal power supply has that might need refreshing. (All three are running newly rebuilt battery packs, so it isn't battery voltage). I'll dig up my service manual and give them a check over the next couple days.

What a nice, beautiful and superb machines!

I have recently got a TI virus and have collected some untis, like several 59, a 58, a 30 (original) and a 55. Love the LED displays!

Cheers

ijabbott · (This post was last modified: 04-17-2019 10:57 PM by ijabbott.)

(04-16-2019 01:23 AM)Xorand Wrote: One thing I had noticed when I got the 58C is that its display was a little dimmer than the 59. I subsequently found that the display on the 58 is dimmer still. I think there is a component (or components) that the internal power supply has that might need refreshing. (All three are running newly rebuilt battery packs, so it isn't battery voltage). I'll dig up my service manual and give them a check over the next couple days.

The usual goto fix for old electronics is to replace the electrolytic capacitors which tend to either dry out or leak over time. I don't know it applies in this case.

Xorand · 04-18-2019, 03:54 AM

(04-17-2019 10:56 PM)ijabbott Wrote: The usual goto fix for old electronics is to replace the electrolytic capacitors which tend to either dry out or leak over time. I don't know it applies in this case.

Typically, yes. I know I ran across an article or forum thread somewhat recently of someone else mentioning that dim displays are sometimes caused by a bad part or parts in the power supply circuit (between the battery and the rest of the circuits). I really should bookmark this stuff when I see it.

ijabbott · 04-18-2019, 08:59 PM

(03-30-2019 02:13 PM)Gene Wrote: The TI-58C was always my favorite of the bunch. Constant Memory was always better IMO than an unreliable card reader...of course, it had half the memory, but I took that trade off.

The TI-58C is among the three (or four) TI models I recommended all HP collectors have from that side of the competition.

The others were an SR-50 original, the SR-56 and perhaps a TI-30 because of its ground-breaking price / performance point.

The TI-30 is an engineering marvel of product design. (Apart from the crappy keyboard.) The way it just clips together and all implemented on a single chip (plus LED display) with no discrete components is wonderful!

Xorand · (This post was last modified: 04-19-2019 12:48 AM by Xorand.)

So I found some specs for power supplies in the TI-58/59 series.

The battery voltage is (nominally) 3.6V. There is a voltage board that also creates -10V and -16V. I found where to measure those on the TI-58 and they were -6.5V and -12.5V, respectively.

I was using my variable DC power supply to provide the 3.6V input to the calculator, so I carefully ramped up to 3.9V input. Interestingly, the two voltages remained somewhat constant if not dropping just a little. Hmm.

For reference, I pulled apart a working TI-59 that I have in the current parts pile (missing a card reader) and performed the same testing on it with about the same results. However, the display on the TI-59 was fairly brighter than the TI-58.

The TI-58 is a very early model (date stamp 4677, which is around November of 1977 - they were introduced in May of 1977). Wonder if they were just using dimmer LEDs in earlier models. My good TI-59 has the brightest display of the bunch of calculators but is a 1982 build.

Maximilian Hohmann · 04-19-2019, 08:39 AM

Hello!

(04-18-2019 08:59 PM)ijabbott Wrote: The TI-30 is an engineering marvel of product design. (Apart from the crappy keyboard.) The way it just clips together ...

Yes indeed! However when I first took one apart (if I remember correctly to see if anything could be done about the keyboard bounce) I was a little bit disappointed to see that there is almost nothing inside!

(04-18-2019 08:59 PM)ijabbott Wrote: ... all implemented on a single chip (plus LED display) with no discrete components is wonderful!

They cheated a little bit by putting all the discrete stuff inside the battery pack where it is used to convert the 2.4V of two NiCd cells into stabilised 9V. Later models from Ti did that with a single cell inside the pack. I wonder why no one else came up with that concept.

Regards
Max