Peculiar speed differences between two 11Cs

[Dave Britten]

I've got two 11Cs, both made in USA, one in 1982 and the other in 1985. The 1982 model has a somewhat more sluggish response in its LCD, most noticeable in the leftmost digit, where the segments tend to be a bit delayed in updating. Also, when "running" is flashing, the speed of flashing is slower than the 1985 unit.

But the weird thing is that, despite the faster display updating, the 1985 seems to run about 10% slower (very rough estimate based on just eyeballing the completion times of the 8-queens benchmark - it was a good few minutes' difference).

Is this a hardware design/revision difference, a manufacturing/tolerance difference, aging components, or something else?

Slightly unrelated, the 1982 has heavier but darker orange paint for the keyboard labels, whereas the 1985 is thinner but much brighter.

[Ren]

(04-12-2022 05:05 PM)Dave Britten Wrote:  Slightly unrelated, the 1982 has heavier but darker orange paint for the keyboard labels, whereas the 1985 is thinner but much brighter.

No, I don't think that is unrelated at all, it takes longer for the calculations to go through the heavier calculator.
More atoms for the electrons to negotiate around.
But, it could be the darker orange paint is absorbing more photons or flogistan.

B^)

[KeithB]

Fresh batteries in each?

[Dave Britten]

(04-12-2022 09:09 PM)KeithB Wrote:  Fresh batteries in each?

The 1982 with the sluggish display has fresh batteries, but the ones in the 1985 have probably been in there a year or two.

[Marcovecchio]

This page lists various chip versions for the Voyagers during production years. I think you can open your calculators and easily check the chip serials: http://www.brouhaha.com/~eric/hpcalc/voy...iants.html

Perhaps we can find a relationship between revisions and performance. I have about 6 HP11c calculators, from various revisions. If you share your tests, times and chip revisions, I can test here and share my results to help find a pattern.

[damaltor]

The devices get their clock from a simple parallel LR circuit, and not from a precise clock source like a crystal. The inductivity is a part with fairly high tolerances, so 10% is probably within the tolerance.
It is easy to overclock them by the way. More info here:
http://www.brouhaha.com/~eric/hpcalc/voy...eedup.html

[Dave Britten]

(04-13-2022 06:17 AM)damaltor Wrote:  The devices get their clock from a simple parallel LR circuit, and not from a precise clock source like a crystal. The inductivity is a part with fairly high tolerances, so 10% is probably within the tolerance.
It is easy to overclock them by the way. More info here:
http://www.brouhaha.com/~eric/hpcalc/voy...eedup.html

Not too surprising. But what surprised me is that one had a faster LCD refresh and "running" flash rate while also having slower execution speed. Very weird!

[OlidaBel]

Do they have the same memory content when running your tests ?
As you know, a running program possibly parses the whole programming steps when looking for a LBL.

[Dave Britten]

(04-13-2022 01:16 PM)OlidaBel Wrote:  Do they have the same memory content when running your tests ?
As you know, a running program possibly parses the whole programming steps when looking for a LBL.

As far as program content, yes - storage registers may have differed. I just typed in the n-queens benchmark from a CLEAR PRGM state. I'll have to give it another run and time the results.

[damaltor]

(04-13-2022 12:20 PM)Dave Britten Wrote:  Not too surprising. But what surprised me is that one had a faster LCD refresh and "running" flash rate while also having slower execution speed. Very weird!

I faintly remember reading about two different display types used over the years, but i cannot find that info again unfortunately. if i am not only imagining it.

[Dave Britten]

(04-14-2022 09:40 AM)damaltor Wrote:  

(04-13-2022 12:20 PM)Dave Britten Wrote:  Not too surprising. But what surprised me is that one had a faster LCD refresh and "running" flash rate while also having slower execution speed. Very weird!

I faintly remember reading about two different display types used over the years, but i cannot find that info again unfortunately. if i am not only imagining it.

The two displays have a very subtly different appearance, both in contrast and shape of the segments, so there could be something to that.

Also I misspoke earlier - these are 1982 and 1987 units, not 1985. I've got too many Voyagers laying on my desk and I'm getting the dates mixed up.

[David Hayden]

Did the ROM software in the 11C change during its lifetime? Maybe there was a bug fix that slowed it down.