MAMEWorld >> The Loony Bin
View all threads Index   Flat Mode Flat  

italieAdministrator
MAME owes italie many thank yous, hah
Reged: 09/20/03
Posts: 15246
Loc: BoomTown
Send PM
Re: Xtal compensation theory 101: Part 2
07/03/12 05:29 PM


> > > "Reverse error generation"? "Cancellation signal"?
> > > You mean, as opposed to a lookup table containing values such as "500000" and
> > > "500012"?
> >
> > Yes. One solid difference is that most of these error corrections are generated
> real
> > time (analog). A digital look-up table would require a huge resolution to be
> > effective at anything other than one very precise fixed temperature (see oven
> > oscilator, the only way this principle works to any degree of accuracy).
>
> What do you mean by a "huge resolution"? Do you mean lost of entries? Or do you mean
> that the entries themselves would have a lot of digits?
>
> What you could do is: instead of thinking in terms of cycles per second, think in
> terms of (pico- or femto-)seconds per cycle. The table would have entries like:
>
> (figures pulled out of ass, for illustration only)
> 20.0 deg C: 0.000 000 999 999 95 sec
> 20.1 deg C: 0.000 001 000 000 00 sec
> 20.2 deg C: 0.000 001 000 000 07 sec
> 20.3 deg C: 0.000 001 000 000 18 sec
>
> and you're smart enough to figure out what happens with each oscillation of the
> crystal.



So you have built your uber precise table based upon a very controlled data collection in a stable environment...

Then your dog sneezed on the clock...

Your table has just been invalidated, enjoy your recalibration.


Xtals are extremely sensitive to many factors, many of which I haven't even touched on. If you aren't compensating for them in some way, your tweaks are random and inconsequential minutes after they are implemented.







Entire thread
Subject Posted by Posted on
* What time is it? (rant) amused 07/02/12 12:46 PM
. * Did an elephant sit on your fence? GatKongModerator  07/02/12 09:51 PM
. * For reasons I can't explain, I love this thread. *nt* StilettoAdministrator  07/02/12 05:51 PM
. * Re: What time is it? (rant) Gor  07/02/12 04:21 PM
. * Re: What time is it? (rant) amused  07/02/12 05:05 PM
. * Re: What time is it? (rant) Gor  07/02/12 06:45 PM
. * Re: What time is it? (rant) JWJr  07/02/12 05:25 PM
. * Re: What time is it? (rant) amused  07/02/12 07:18 PM
. * Re: What time is it? (rant) italieAdministrator  07/02/12 07:30 PM
. * Re: What time is it? (rant) amused  07/02/12 07:36 PM
. * Re: What time is it? (rant) italieAdministrator  07/02/12 07:48 PM
. * Re: What time is it? (rant) amused  07/03/12 03:45 AM
. * Xtal compensation theory 101: Part 1 italieAdministrator  07/03/12 05:45 AM
. * Xtal compensation theory 101: Part 2 italieAdministrator  07/03/12 06:23 AM
. * Re: Xtal compensation theory 101: Part 2 amused  07/03/12 11:04 AM
. * Re: Xtal compensation theory 101: Part 2 italieAdministrator  07/03/12 01:42 PM
. * Re: Xtal compensation theory 101: Part 2 amused  07/03/12 04:24 PM
. * Re: Xtal compensation theory 101: Part 2 italieAdministrator  07/03/12 05:29 PM
. * Re: Xtal compensation theory 101: Part 2 Bekki Doll  07/07/12 12:05 AM
. * Re: Xtal compensation theory 101: Part 2 Vas Crabb  07/03/12 04:27 PM
. * Re: Xtal compensation theory 101: Part 2 Gor  07/03/12 05:20 PM
. * Xtal compensation theory 101: Part 3 italieAdministrator  07/03/12 07:25 AM
. * Re: What time is it? (rant) lharms  07/02/12 10:01 PM
. * Re: What time is it? (rant) amused  07/03/12 03:50 AM
. * Just one question (yes, only one) italieAdministrator  07/02/12 02:02 PM
. * Re: Just one question (yes, only one) amused  07/02/12 02:40 PM
. * Re: Just one question (yes, only one) Tomu Breidah  07/02/12 03:52 PM
. * Re: Just one question (yes, only one) italieAdministrator  07/02/12 03:47 PM
. * Re: Just one question (yes, only one) amused  07/02/12 04:02 PM
. * Re: Just one question (yes, only one) italieAdministrator  07/02/12 04:22 PM
. * Re: Just one question (yes, only one) amused  07/02/12 05:07 PM
. * Re: Just one question (yes, only one) italieAdministrator  07/02/12 07:51 PM
. * Re: Just one question (yes, only one) TriggerFin  07/02/12 08:36 PM
. * Re: Just one question (yes, only one) italieAdministrator  07/02/12 09:07 PM

Extra information Permissions
Moderator:  GatKong 
1 registered and 191 anonymous users are browsing this forum.
You cannot start new topics
You cannot reply to topics
HTML is enabled
UBBCode is enabled
Thread views: 2810