Page 7 of 8

Re: DM42- clock losing time

Posted: Wed May 15, 2019 8:54 pm
by akaTB
H2X wrote:
Wed May 15, 2019 12:57 pm
akaTB wrote:
Wed May 15, 2019 11:49 am
toml_12953 wrote:
Wed May 15, 2019 10:37 am
I want the portrait version! :D
I could concur, in this case.
8-)
You are strangely quiet about the placement of the arithmetic operator keys... :D
Yeah: I am thinking about the vertical red LED bar... :P

Re: DM42- clock losing time

Posted: Sat May 18, 2019 6:31 am
by Jebem
keithdalby wrote:
Tue May 14, 2019 7:09 pm
Ah yes, but what was the average temperature? Humidity? Where in the lunar cycle are we? What was the background radiation during this period? These are all important to know as well.
Yeah it is funny :)

But one working in the electronics field would know that using regular smd capacitors in the oscillator load circuit as seen in the dm42 pictures can result in permanent capacitance drift during the soldering operation.
This is well understood by the industry.
Therefore adequate low drift temperature capacitors are a must in this position.
Nothing new under the sun.

So the specified 20ppm crystal is the base drift value, but the actual value can be lower or higher depending on component tolerances, and yes, temperature variations.

Not so sure about the large effects of humidity and radiation changes but it will affect somehow, even considering that the crystal and capacitors are encapsulated hermetically in kind of sealed cases.

And let us not forget the vibration effects on any crystal resonant period.
A constant strong vibration coming from the key domes being smashed by busy/ speedy users will for sure have some impact on the poor crystal inside, no?
:)

Re: DM42- clock losing time

Posted: Sat May 18, 2019 8:07 am
by Walter
Our clowns are all right regarding short time intervals. Though when you put date and time on default display you must not wonder at people bothering about it.

Re: DM42- clock losing time

Posted: Tue May 21, 2019 6:13 am
by H2X
Walter wrote:
Sat May 18, 2019 8:07 am
Our clowns are all right regarding short time intervals. Though when you put date and time on default display you must not wonder at people bothering about it.
Granted. There is plenty of space (once the date and time are gone), and other potentially interesting stuff (to some, if not all users) to put on the status bar, which might have been manageable in Settings.

... but I still want my DM-01! :D

Re: DM42- clock losing time

Posted: Tue May 21, 2019 10:31 am
by Thomas Okken
Out of curiosity, I measured the accuracy of the clock in my car. It's a very basic clock: 12-hour digital display, hours and minutes only, not even an AM/PM indicator, and a single button for setting the time. But quite accurate: over 7.5 days I couldn't measure any drift, and given that my measurement had an accuracy of about one second, that puts an upper bound on the drift of under 1 minute per year, or under 2 ppm.

Presumably, a car clock will be temperature-compensated, to deal with the wide range of temperatures inside cars.

Calculator clocks don't need temperature compensation because they spend most of their time in temperature-controlled environments, i.e. within the narrow range of temperatures that humans find comfortable. But they could still benefit from a software-based calibration system such as that used in the HP-82182A Time Module or the HP-41CX calculator, and that should be pretty easy to implement. Assuming the crystals used in the DM42 are reasonably constant, their deviations from their rated frequencies should be possible to compensate for in this manner.

The UI for this could be added to the SETUP menu, or handled by implementing the CORRECT, RCLAF, and SETAF functions. Or both.

And for extra points, user-configurable automatic DST adjustment. 8-)

Re: DM42- clock losing time

Posted: Tue May 21, 2019 10:56 am
by H2X
Thomas Okken wrote:
Tue May 21, 2019 10:31 am
Out of curiosity, I measured the accuracy of the clock in my car. It's a very basic clock: 12-hour digital display, hours and minutes only, not even an AM/PM indicator, and a single button for setting the time. But quite accurate: over 7.5 days I couldn't measure any drift, and given that my measurement had an accuracy of about one second, that puts an upper bound on the drift of under 1 minute per year, or under 2 ppm.

Presumably, a car clock will be temperature-compensated, to deal with the wide range of temperatures inside cars.

Calculator clocks don't need temperature compensation because they spend most of their time in temperature-controlled environments, i.e. within the narrow range of temperatures that humans find comfortable. But they could still benefit from a software-based calibration system such as that used in the HP-82182A Time Module or the HP-41CX calculator, and that should be pretty easy to implement. Assuming the crystals used in the DM42 are reasonably constant, their deviations from their rated frequencies should be possible to compensate for in this manner.

The UI for this could be added to the SETUP menu, or handled by implementing the CORRECT, RCLAF, and SETAF functions. Or both.

And for extra points, user-configurable automatic DST adjustment. 8-)
That might make me reenable date & time on my status bar. :-)

I'd still like other options for my status bar. And my DM-01! :D

Re: DM42- clock losing time

Posted: Tue May 21, 2019 2:30 pm
by Walter
H2X wrote:
Tue May 21, 2019 10:56 am

I'd still like other options for my status bar. And my DM-01! :D
How about putting your wish for a DM-01 in your signature? 8-)

Re: DM42- clock losing time

Posted: Tue May 21, 2019 3:42 pm
by H2X
Walter wrote:
Tue May 21, 2019 2:30 pm
H2X wrote:
Tue May 21, 2019 10:56 am

I'd still like other options for my status bar. And my DM-01! :D
How about putting your wish for a DM-01 in your signature? 8-)
I don't really believe it's realistic. I'll stop nagging about it, and instead join the crowd if one emerges... :D

Re: DM42- clock losing time

Posted: Wed May 22, 2019 5:16 am
by grsbanks
Thomas Okken wrote:
Tue May 21, 2019 10:31 am
Presumably, a car clock will be temperature-compensated, to deal with the wide range of temperatures inside cars.
Also, many car clocks are adjusted by RDS if you listen to the radio. This is the case with mine (Vauxhall/Opel Insignia estate) so I don't even need to adjust it when changing to/from Summer time.

Re: DM42- clock losing time

Posted: Wed May 22, 2019 1:02 pm
by Thomas Okken
grsbanks wrote:
Wed May 22, 2019 5:16 am
Thomas Okken wrote:
Tue May 21, 2019 10:31 am
Presumably, a car clock will be temperature-compensated, to deal with the wide range of temperatures inside cars.
Also, many car clocks are adjusted by RDS if you listen to the radio. This is the case with mine (Vauxhall/Opel Insignia estate) so I don't even need to adjust it when changing to/from Summer time.
I think the clock in my car manages its accuracy without any outside help. I can't say with absolute certainty that it doesn't get synchronized by the radio (though the manual makes no mention of such a feature) (and it wouldn't work very well in my case since I don't listen to the radio), but it definitely doesn't handle the DST switch on its own. 12½-year-old Toyota Yaris; your mileage may vary!