Second day..

This time it was my fridge. Everything was just like with freezer yesterday. 10 second intervals (for laughs, even 10 mins would be enough for something that slow). In real life i was logging 22 hours, but only got ~two. And time started from zero, actually first line on log is “4.361”. Only last two lines had real timestamps (accurate for when i stopped logging).

Even weirder, as i did show my friend how little changes would show on a graph after about 10 hours of logging, with app at graph mode and it seems it has not affected anything, if values on a card are from first two hours that i attempted.. Or there was no logging at all for first ~10 hours and it started when i was demoing. And still there are those 2 lines with real timestamps at this morning.

Temperature data from last two lines do skip a little for internal temp, which may or may not tell that there was a (long?) pause between datapoints.. Thermocouple voltage seems to match perfectly.

So, what would it need to get mooshimeters clock to zero itself? And how much connection time with app it needs to sync again? After yesterdays freezer logging, i got memorycard out, copy+delete only logfile, card back in and set probe to fridge. Just remembered, that i did a graph demo to my dad also, as he happened to visit at just as i was setting things up. Does graph-mode do some harm to logging or timesync?

As for the math for thermocouples, i think i quit and go for pt100/1000 probes.. Just need to find one, i’m sure that i have some in very deep storage. I hope those are easier :D

ville,
“first line on log is 4.361” — I have seen this in the distant past, maybe release 23, Firmware 1466039460. It was unrelated to thermocouple. James suspected the firmware watchdog timer issue. It was fixed in later releases and now it must have returned.

Also, I thought James said logging and graphing were never intended to be performed together.

There is lots of good historical info in this forum, but I haven’t found a good way to search it. And relying on my memory is…well…I forgot what I was going to say. —- Duane

And story goes on.

Did not find any pt100 or 1000 sensors, but in that 2 minutes i was searching, i found few ntc probes for heated floor thermostat. The one that i tried, seems to be 33k@25C.

For testing that probe, i set logging for 1 sec intervals and tried boiling water and melting ice. Cant remember very accurate values, but there is a bigger problem, as there was no log file after that session.. I had live values or graph on a phone for all that time and logging should have been on. At least i did not see anything wrong.

How is led supposed to blink in those conditions? 1 blink/1 sec for connection AND 3 blinks/1 sec for writing samples? If every sample is written in realtime, there would have to be way too much blinking :D

I had to reset mooshimeter to get it logging again.. I should have checked how led is blinking in different situations, but did not. As it is hard to remember and I think there should be more feedback visible on a app side of what the meter is doing or failing to do. After all, on most cases, there is couple million pixels to tell us a story (while setting things up or checking what is happening in a middle of multiple days of logging), rather than to do it only with one amber led..

Seems odd to not be able to graph on a phone while logging, if we stay out of buffer mode (which would be a totally different case).. We are able to stay on main screen with same live values and as far as i know, mooshimeter does not need to know what we do at the receiving end of bluetooth connection..

And yes, (working) forum search would solve so many problems, for us who try to break everything (like proud beta testers should) ;) and for new users also.. Maybe James has to start answering our questions directly to wiki :D

Hi!

I will check that pdf and try to learn how those k-probes work.. Or just quit and continue to search for pt1000 probe from my storage :D

And freezer vs fridge.. Mooshimeter was outside of both, only that fluke k-probe with tiny and flat cord went in, hardly even denting door gaskets. So room temperature for sd-card. And my “experiment” was to see how much there is variation in temperatures, as that fridge+freezer is one unit with only one compressor, so other has to suffer when switching it on/off.. I try to be creative with my beta-testing, as nice graphs are just nice graphs, but when i measure something that somehow interests me, i probably find different kinds of bugs or at least be more angry for crashes and report them, as i have lost something.. Its so easy to just reset your meter and try again what the meters internal temperature inside storage case on your workbench is :D

Then the blinking/chirping. I really dont have a clue how it was blinking :( .. Should have checked, but after all, failed logs before this had been total crashes and this time i was able to see live values and graph few times during that logging, so i did not suspect that there is anything wrong.. Meter uptime had to be from that last firmware update, it succesfully logged my freezer for 7300 lines (10s interval) and i do stop logging before getting card out, every time. Maybe i should reset mooshimeter while copying files to computer from now on.. At least when doing semi-interesting measurements.. Internal temp at workbench would have to do for uptime records ;)

I did try google site search yesterday, after Duane and i had mentioned searching.. I got bit odd results and did not bother to try again, but today it seems to find real threads. Yesterday it gave me page full of results to some “admin/reply” view with only posts from James and i had to click each again to see what the question was :D .. I am so busy on vacation that google has to give me right result on a first try ;D

Aaargh.. I dont even like writing or know english, where did all this came from? :D

Hello,
Where can I buy a MS3206 connector for the K thermocouple in France?
Thanks & regards,
J.C. De BIEVRE