search

Safecast / bGeigieNanoKit

bGeigieNano is a kit version of the bGeigie mobile survey geiger counter designed to fit into a Pelican Micro Case 1010.
https://safecast.org/devices/bgeigie-nano/
114 stars 43 forks source link

possible data corruption in log file #31

Open thinrope opened 10 years ago

thinrope commented 10 years ago

While investigating data from Nano with ID 2028 (https://api.safecast.org/en-US/bgeigie_imports/14580) a possible corruption was identified. After a strangely formatted longitude, the counts jump, sometimes wildly. The longitude should have been formatted "%05ld.%04ld" per https://github.com/Safecast/bGeigieNanoKit/blob/master/bGeigieNano.ino#L754 but it has more digits, e.g. (see 4th line, 11608.429496726):

$BNRDD,2128,2014-02-21T04:38:09Z,80,10,11344,A,3213.3686,S,11608.1638,E,282.40,A,5308426,0*53
$BNRDD,2128,2014-02-21T04:38:14Z,77,6,11350,A,3213.3046,S,11608.1079,E,276.20,A,5308426,0*61
$BNRDD,2128,2014-02-21T04:38:19Z,81,10,11360,A,3213.2410,S,11608.0521,E,272.00,A,6029321,0*5B
$BNRDD,2128,2014-02-21T04:38:24Z,80,4,11364,A,3213.1803,S,11608.429496726,E,278.80,A,6029321,0*5B
$BNRDD,2128,2014-02-21T04:38:29Z,145,71,11435,A,3213.1247,S,11607.9427,E,281.30,A,6029321,0*68
$BNRDD,2128,2014-02-21T04:38:34Z,146,7,11442,A,3213.0695,S,11607.8827,E,282.00,A,6029321,0*51
$BNRDD,2128,2014-02-21T04:38:39Z,149,10,11452,A,3213.0105,S,11607.8262,E,288.30,A,5505034,0*67

It looks like lines 749 and 750 need explicit cast to long? Also unsigned long number constants should be suffixed with UL per http://arduino.cc/en/Reference/IntegerConstants

May be that is the culprit that has been hunting us? It seems that "429496726" comes from the maximum unsigned int (2^32=4294967296") divided by 10 per https://github.com/Safecast/bGeigieNanoKit/blob/master/bGeigieNano.ino#L750 and a bit off. It seems that the unit in question has used older firmware, 1.3.1 changed the code a bit, but still needs the above fix.

robouden commented 10 years ago

Kalin,

I think you see a bug that Joe found before. What version software was #2038 running at that time?

regards rob

Regards, Rob Oudendijk Yuka Hayashi http://yr-design.biz http://oudendijk.biz http://about.me/robouden tel +81 80-22605966 Skype: robouden Facebook:roboudenhttp://on.fb.me/QeKw2P twitter:robouden http://bit.ly/RAiSTC

On Mon, Feb 24, 2014 at 4:38 PM, Kalin KOZHUHAROV notifications@github.comwrote:

While investigating data from Nano with ID 2028 ( https://api.safecast.org/en-US/bgeigie_imports/14580) a possible corruption was identified. After a strangely formatted longitude, the counts jump, sometimes wildly. The longitude should have been formatted "%05ld.%04ld" per https://github.com/Safecast/bGeigieNanoKit/blob/master/bGeigieNano.ino#L754but it has more digits, e.g. (see 4th line, 11608.429496726):

$BNRDD,2128,2014-02-21T04:38:09Z,80,10,11344,A,3213.3686,S,11608.1638,E,282.40,A,5308426,0_53 $BNRDD,2128,2014-02-21T04:38:14Z,77,6,11350,A,3213.3046,S,11608.1079,E,276.20,A,5308426,0_61 $BNRDD,2128,2014-02-21T04:38:19Z,81,10,11360,A,3213.2410,S,11608.0521,E,272.00,A,6029321,0_5B $BNRDD,2128,2014-02-21T04:38:24Z,80,4,11364,A,3213.1803,S,11608.429496726,E,278.80,A,6029321,0_5B $BNRDD,2128,2014-02-21T04:38:29Z,145,71,11435,A,3213.1247,S,11607.9427,E,281.30,A,6029321,0_68 $BNRDD,2128,2014-02-21T04:38:34Z,146,7,11442,A,3213.0695,S,11607.8827,E,282.00,A,6029321,0_51 $BNRDD,2128,2014-02-21T04:38:39Z,149,10,11452,A,3213.0105,S,11607.8262,E,288.30,A,5505034,0*67

It looks like lines 749 and 750 need explicit cast to long? Also unsigned long number constants should be suffixed with UL per http://arduino.cc/en/Reference/IntegerConstants

May be that is the culprit that has been hunting us?

Reply to this email directly or view it on GitHubhttps://github.com/Safecast/bGeigieNanoKit/issues/31 .