ezChronos headcount and first 2011 project start ? - Page 35
Like Tree3Likes

Thread: ezChronos headcount and first 2011 project start ?

-->
Page 35 of 37 FirstFirst ... 25 26 27 28 29 30 31 32 33 34 35 36 37 LastLast
Results 341 to 350 of 361
  1. #341
    Member
    Join Date
    Jan 2009
    Posts
    3,039

    v3rc1

    Finally I managed to get things with the new v3 to a point where we can have a new Release Candidate

    The "external changes" are minimal compared to the final v2 - however there have been very serious internal changes - starting with the fact that the CPU is now running at a more conservative 8 MHz (instead of the 12 MHz before) to the point where the rounding approximations and actual internal implementation of the algorithm was seriously improved. The major possible drawback is that the old timing fine-tuning (which you probably had in your EZC2PARAM.INI) is still working but (due to the changes in rounding and the elimination of certain timing errors in v2) is going to provide good TC but with a pretty visible offset So again a certain degree of fine-tuning will be needed

    Also aging was added to the firmware - since we had huge problems with the size of the code only the simplest form of aging algorithm was added (a linear shift) but even that one should keep thing just fine for a pretty long time!

    That also needed a change in the INI file (now called EZC3PARAM.INI) - where there are two new parameters for aging - a signed one that is expressing the amount of aging per day (in s/year) * 100000, and a second unsigned one which is indicating the "0-point of aging" (and is expressed in days since 2000-01-01 - for instance you can use about 4383 to set that point to 2012-01-01). In order to use the new information new PC-companion programs are provided.

    A bug (inherited from the old Texas instruments firmware) was also fixed where 1 second was being added to the stopwatch at the 20 minutes point.

    So here are the files - for each zone-type (433 MHz, 868 Mhz and 915MHz) we now have 4 flavors (datalogger is currently missing):

    - Bluerobin sportswatch - like the original TI firmware but with the HEQ part and a few other things = just like in v2; however since the new aging code took some space I had to remove from this one the repeater

    - WorldTime plus Sidereal plus MoonAge and Worldtime plus CountDown and Interval Timer - just like in v2 but with aging added; here we had to give-up the "altitude alarm" (but everything else is still there);

    - Astronomy Watch - aging was also added here; however keep in mind that this flavor (just as in v2) can not write EZxPARAM info into flash - so any parameter change must be made by first writing another firmware, use EZ3PARAM and then write back the ASTRO firmware!

    433MHz_v3rc1.zip

    868MHz_v3rc1.zip

    915MHz_v3rc1.zip

    ezc_pc_v3rc1.zip
    Last edited by Catalin; January 14th, 2012 at 10:35.

  2. #342
    Member
    Join Date
    Jan 2009
    Posts
    3,039

    Small bug discovered (and fixed for the next RC)

    A small bug was discovered - sidereal time (in both WorldTime+Sidereal and Astro) was wrong in leap years with one (solar) day - meaning it was off by about 3 minutes 56 seconds (that got past the testing from 2011 since that one was not a leap year).

    Anyway that will be fixed in the next RC - probably in 1-2 weeks or so!

  3. #343
    Member
    Join Date
    Mar 2011
    Location
    Bergen, Norway
    Posts
    25

    Some communicating error with ezc3param

    Have upgraded the watch with new software (Bluerobin). When I run ezc3param_rc1 most of the time I get"Error communicating with the watch!" It worked one time, reply "n" and then I had to try about 8 times, before it worked again (The ini is now updated). No com. error with ezc3set or get. (Would just let you know.) I assume I should wait for the new logging-prog before starting the measurement.

  4. #344
    Member
    Join Date
    Jan 2009
    Posts
    3,039

    Re: Some communicating error with ezc3param

    Quote Originally Posted by kjellTiWatch View Post
    Have upgraded the watch with new software (Bluerobin). When I run ezc3param_rc1 most of the time I get"Error communicating with the watch!" It worked one time, reply "n" and then I had to try about 8 times, before it worked again (The ini is now updated). No com. error with ezc3set or get. (Would just let you know.) I assume I should wait for the new logging-prog before starting the measurement.
    Is that the EU version (868 MHz)? What is the battery level just after a sync? I have sometimes seen random communication problems (with the USB adapter - which is NOT using Bluerobin) - but those have NOT been more frequent with the new version compared to older versions. Also those seem to go away if I unplug the USB adapter from the computer and then plug it back.

    I don't remember very well - do you have any Bluerobin device? (heartband?) If yes - how is that one working?

    Regarding the logging - you can use the old datalogger firmware - BUT BE 100% CERTAIN THAT YOU RESET THE PARAMETERS TO the content of EZC2PARAM_0.INI !!!

    If you already have a temperature curve from the last 1-3 months you can probably directly use that one (3 pairs - warm/cold/room - send me a private message if you already have those) - all the effort in getting a new graph makes more sense after 6-12 months in order to determine aging.

  5. #345
    HAQ moderator
    Join Date
    May 2005
    Posts
    3,164

    Re: Some communicating error with ezc3param

    I have the same problems after flashing new firmware. I remove and reinsert the USB dongle and it is back to normal.

  6. #346
    Member
    Join Date
    Jan 2009
    Posts
    3,039

    v3rc2

    A new Release Candidate is available - the main differences are that the Sidereal/Astro 3m56s bug (in leap years) was fixed and the datalogger was also added to the (now full) mix of (now at 8 MHz) firmware 'flavors' - something not that simple as it seems and which also involved changes to the PC-companion programs (all dates for v3rc2 - firmware and PC programs - should be January 24).

    As a side-note - so far a huge effort was made with the PC-companion programs so that the latest versions still work with any previous firmware of any kind (from our HAQ branch, not from the Texas Instruments branch).

    Also the (deep) internal changes done for the HAQ part seem to be paying-off - I will have a more detailed post on just that during the weekend but the variability in accuracy (at the same temperatures or at different temperatures) is now VASTLY reduced and as a result the final accuracy results can be a huge lot better!

    The latest firmware can be loaded from:

    433MHz_v3rc2.zip

    868MHz_v3rc2.zip

    915MHz_v3rc2.zip

    and the latest set of PC-companion programs from:

    ezc_pc_v3rc2.zip

    Please post around here your your own results/feedback - currently the plan is to have this in tests for 2-4 weeks and after that if zero problems are found just have the final release! As a side-note the part where still a HUGE lot of testing should be done is the Sidereal/Astro part for future years at leap+0/1/2/3 (2012/2013/2014/2015), also the day/date correspondence for those years before and after March 1st and the two automatic DST changes in those years (assuming no changes in the national DST rules used - I might need to check again how Russia ended in that regard).

  7. #347
    Member
    Join Date
    Jan 2009
    Posts
    3,039

    Rolling eyes v3rc3 - but we should really test this one

    Just one day after posting the previous rc2 I learned a number of surprising things - and it seems that we might not have anybody testing the WorldTime firmware from Australia/Brazil - or generally from outside EU/US

    Anyway - the DST change was wrong for Australia and did not exist for the bigger part of Brazil - something that is now corrected (but keep in mind that in a timezone where changes are supposed to take place at 23:59:59.999 things will instead take place at 23:00 in the watch - since in the watch at 00:00 there are other conflicting things done). Also a quick research showed two other WorldTime/DST 'problems' - there are simply too many small countries with very unusual/inconsistent rules, and at least one bigger country - Brazil - which decided inconsistent rules are a must - so the WorldTime firmware now has 'User DST' - unfortunately that can not be set manually (which however would also be incredibly tedious/error-prone), but instead using the PC-companion programs - EZC3SET_rc3.EXE now will detect if the DST rules in your (Windows-based) computer are different from any of the DST rules in ezChronos WorldTime firmware and if that takes place just create 'custom DST rules' - you will know that in the watch if the USER timezone will now appear instead as UDST timezone (there will also be an unfriendly debug message from the PC program, I will probably make that friendlier in the final version). If the rule from the watch is OK of course that it will be used - and in that case the TimeZone will be one of ANCH,L A ,DENV,CHIC, NYC,RIOJ,LOND,PARI,ATHN,ADEL,SYDN - other timezone do not have automatic DST in this firmware!

    Also ASTRO and Bruerobin Sportswatch have seen a very minor bugfix added, but the Datalogger did not have that so it remains at the previous rc2 build.

    So here we have it - v3rc3 firmware:

    433MHz_v3rc3.zip

    868MHz_v3rc3.zip

    915MHz_v3rc3.zip

    and the corresponding PC programs:

    ezc_pc_v3rc3.zip

    Please post somewhere around here if you had any problems when initially setting the time with EZC3SET_rc3.EXE - and generally please test if the DST rules seem to be working as expected for you! (and if you want to test the DST change in the future please keep in mind that DST is listed as DSA0/1 for Auto Off/On, but once the DST is applied for the rest of that day UNTIL THE NEXT DAY 00:00:00 it will be shown as DSC0/1 - when you see that you need to go to 23:59:00 and wait for the day to change to the next - which will reset DSCx back to the 'normal' DSAx, and only after that manually go to another DST date to test things!).

  8. #348
    Member
    Join Date
    Mar 2011
    Location
    Bergen, Norway
    Posts
    25

    Funny battery voltage drop when running log-prog

    The watch battery-voltage is dropping fast now, when doing logging. I use ver 2 for logging. Was wondering if installing _rc2 could have anything with lager current consumtion, or could it be low temp that is causing it? Installed _rc2 31 jan to try. It was after that, current comsumtion/voltage drop started. This show startdate, start and end voltage, deg C, days
    18 feb, 3.01, 2.81, 4.0C 5days, new battery,
    03 feb, 3.05, 2.90, 2.9C, 4days, new battety,
    28 jan, 2.94, 2.92, 7.5C, 3days
    25 jan, 2.92, 2.94, 32C, 2days

  9. #349
    Member
    Join Date
    Jan 2009
    Posts
    3,039

    Re: Funny battery voltage drop when running log-prog

    Quote Originally Posted by kjellTiWatch View Post
    The watch battery-voltage is dropping fast now, when doing logging. I use ver 2 for logging. Was wondering if installing _rc2 could have anything with lager current consumtion, or could it be low temp that is causing it? Installed _rc2 31 jan to try. It was after that, current comsumtion/voltage drop started. This show startdate, start and end voltage, deg C, days
    18 feb, 3.01, 2.81, 4.0C 5days, new battery,
    03 feb, 3.05, 2.90, 2.9C, 4days, new battety,
    28 jan, 2.94, 2.92, 7.5C, 3days
    25 jan, 2.92, 2.94, 32C, 2days
    The voltage is clearly linked to the low temperature - but the latest v3 might help a little with not hitting a dangerously-low level - I believe in the old v2 the undervoltage reset was set around 2.6V, now in v3 is around 2.4V. Once you are done with collecting the fridge data you should be fine anyway.

    Please send me a private message if you need any help with determining the quadratic coefficients! (I also hope that you have correctly reset the watch parameters to 0 before starting the logging tests - that was extremely important for accurate results in v2!)

  10. #350
    Member
    Join Date
    Jan 2009
    Posts
    3,039

    Version 3 'official release'

    There have been no problems reported with the latest Release Candidate (which was here together with some information on how that is handling unusual DST zones), so today (since is a "once in four years occasion" suggested by our moderator Ronald ) we have the 'official release' of version 3

    Technically the version number is 3.0 for 12 MHz firmware and 3.1 for 8 MHz firmware, but since we now only use the 8 MHz version (for reasons of power consumption and avoidance of low-power reset) the current version has an internal number of 3.1 (the minor version number is used by the PC programs to "know" what frequency the firmware has).

    The 5 'firmware flavors' (astro, datalogger, bluerobin, worldtime+countdown and worldtime+sidereal) are grouped as usual for the 3 possible watch version:

    433MHz_v3_1.zip

    868MHz_v3_1.zip

    915MHz_v3_1.zip

    and the PC companion programs can be found here:

    ezc_pc_v3_1.zip

    If you have not used any of the v3 release candidates the overall differences from the previous v2 are split in two parts - external differences which are very-very small, and internal differences which are quite big but 'invisible'.

    Other than a number of bugfixes the easy to see exernal differences are the presence of "aging compensation" in the thermocompensation part and in the INI file plus eventually a number of small differences in the timezones used by the WorldTime firmware - including now the possibility for 'user-DST' (set from Windows automatically) which changes the old USER timezone to be listed as UDST!

    I will now leave my own ezChronos in timing tests for a few months to clarify the effectiveness of "age-compensation" - currently I am starting from under +1 s/y and we will see in a few months the actual evolution of that.

    Have fun with the new version and let me know if there is any problem!
    Last edited by Catalin; February 29th, 2012 at 12:33.

Page 35 of 37 FirstFirst ... 25 26 27 28 29 30 31 32 33 34 35 36 37 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

Search Engine Friendly URLs by vBSEO 3.6.0 PL2