geez
[Top] [All Lists]

Re: Range Check error?

To: The MacLeods <lmacleod@earthlink.net>
Subject: Re: Range Check error?
From: Byron Short <bshort@AFSinc.com>
Date: Tue, 08 Jun 2004 08:30:21 -0700
Hey Larry,

Strangeness!  I will easily admit that since I'm not competing right 
now, and not doing any testing work either (having to work on my real 
job these days...my current project is up to 116,000 lines of code...and 
is probably 2/3 through),  I don't know that I know as much as I used to 
know when I was actively using GEEZ as much as many of the users are.

Only one thing jumped out at me.  You said that you just put in 
batteries after a recent loss of everything on the Palm and re-sync.  So 
that means (I gather) that the calibration string is a restored string 
that you got from the sync.  I don't know why or if that should make a 
difference, but I would bring the cube and Palm inside to your 
desk/dining room table, and do a full calibration again.  It's possible 
that your calibration values drifted, although I've used the same 
Palm/cube combo a year later and found the calibration data to be within 
a few (8-10) points in each direction, which is near nothing.  But just 
the same, I would re-do it, and see what happens.  Avoid doing the full 
calibration in the field; too many other variables.  Then, since your 
old plan seemed to work great, just follow your old plan, as far as when 
to hit AutoStart.

The symptom that you got 5 starts out of 3 runs (assuming that wasn't 
normal with your previous setup) is also an indicator that something is 
different.  I would try a recalibration, and see what happens.  Also, 
when I recalibrate, I like to do about 2 or 3 of them, and write down 
the string each time.  What I'm looking for is to see if the string 
moves around much, or if it is stable.  It makes sure that I'm doing 
things the same way AND that the cube is doing things the same way.  
Note that a difference of 8-10 points on each direction is less than 
.01g, and pretty insignificant.  Do 2 or 3 calibrations until you are 
satisfied that your numbers are not jumping around any.

Let us know!

--Byron

The MacLeods wrote:

> Byron Short wrote:
>
>> 1.  Go to Setup | Data and make sure that you have a check in the box 
>> for "Display Raw Unadjusted Data".
>
>
> Yep, got that.
>
>> 2.  Check to verify that your Palm had a valid calibration string on 
>> it when you did the recording.
>
>
> I hadn't done a recent calibration (definitely not the day of the 
> recording).  But the displayed calibration string is:  1A18A13B36B64
>
> Do I need to do it again just because it had been so long?  Maybe it's 
> something with that?
>
>> 3.  If it's neither of the above, then I would look for either a bad 
>> piece of data in the download from Palm to PC (but this isn't too 
>> likely to occur on more than one run.
>
>
> I tried to re-download the runs, but that didn't help.
>
> Here's some more info.  I took 4 runs on Sunday, but GEEZ recorded 
> six.  The first one is probably from my grid spot to the start line 
> (37 seconds), but never deleted itself.  What's wierd is that one 
> doesn't give me any error!?! So something with a low-load run let it 
> work okay.  The other 5 runs all give the error.  I'm not sure which 
> one of the last 5 wasn't from a racing run.
>
> Also, these were fresh batteries in the palm just after a complete 
> palm re-boot and hot-sync (had dead batteries).  The battery in the 
> GEEZ was also a fresh 9v.
>
> I'm running the Pro in Toledo this weekend and plan to try again.  
> This time I'll do a full calibration right before I hit autostart to 
> see if that helps.
>
>
> -Larry

///  Archives at http://www.team.net/archive
///  Check http://www.team.net/mailman/listinfo for other lists of interest.


<Prev in Thread] Current Thread [Next in Thread>