You do this hobby to a much finer level than I and while I hope to one day be to that size I am currently assembling my first 200Ah 7s pack that is properly designed vs cobbled together like what my house is currently running off, I live off grid so I needed something quick.
Alright the bus bars were installed and the tedious process of pack testing is complete.
I must say the results are more than spectacular. Total difference between the low and high pack is 1.40Ah. I am very pleased with this result. Just goes to show you the power of nemo's excel based repackr and choosing the best cells for the battery.
I started taking pictures of the iCharger results from pack 3 on just so that you can see that I am not fudging the numbers.
Not that any of you would doubt me but I always like showing proof.
Next are the lexan sides, longmon installation, and into the battery box they go.
So you all know that I am storring and monitoring my Batrium Battery data with influx and have grafana visualize it.
Well an anomaly happened today and it has me puzzled. Actually once I did a 5 day trace there were 2 issues.
The first one was no data collected on 09/04 from 22:30 to 09/05 00:30. 2hrs of no data. I kinda dismissed it believing it may have been a wireless/Batrium issue. Now today this other anomaly has me wondering. Do I have a Batrium, a longmon or a battery issue.
As you can see in the trace below pack #4 had some weird readings for about the same amount of time as the first no data issue.
The odd readings happened from 09/08 12:00 to 09/08 14:00. After that the pack snapped back into the fold and has been following the normal trace since. Obviously I will be watching this #4 Pack very closely. I have to wait for Batrium to roll over tonight for me to download today's session history. I will do that tomorrow and plot that in excel and see if there are any clues there.
All pertinent traces and pictures below.
Anyone have a clue what may have happened here? Solar flare? Neutrinos? Electrons just got tired?
Trace with just pack 4 and cell voltage difference shows data drop and pack anomaly.
5 day trace all 42 packs
2 day trace just for clarity
And a closeup
And another view
I mean the packs are well balanced so no apparent problem
Have you checked the wires that the longmon uses to monitor the pack? Check continuity from the "pin" on the longmon (if you can) all the way to the pack and move the wire around. Maybe the solder is cracking on the pin on the board.
Which version of the longmon you have, V2? I know you have had it for awhile now.
Its hard to imagine a single pack's voltage at you're size would dip so much. If it were power surges, the other packs in series should show it as well. Easier to imagine that longmon connections are flaky as @Korishan suggested OR the longmon itself is a bit flaky. If the longmon connection wire doesn't seem to be the problem - maybe replace the longmon next.
Getting 'way out there' / on the theme of longmon being flaky - was there any particular event such as bypass (heat) going on at that time period that might affect the longmon itself?
@Korishan and @OffGridInTheCity
I kinda was going that route as far as a possible wire/longmon issues. For my last longmon order I got 4 spares just for this type of incident.
You know I have no idea what version longmons I have. They all look the same to me. Kinda got a dumb look on my face when Kori mentioned that and not knowing, as I like to keep up on all of that. I just ordered them when I needed more and looking through the orders now I don't see a version number or anything that points to a difference.
The way out there possibility.... bypass heat, no not that I am aware of as I havent had the need to run a "balance" for ever, the "bypass" voltage is also set to 4.08V and my charge controllers throttle down at 4.078V (57.1V). Also the differential gap is set to 0.03V and at the "top" voltage all my packs run in the 0.01V to 0.02V difference range so there really is no need to balance. But you bring up a good point as far as temperature is concerned. I will add the individual longmon pack temp and possibly the bypass temp to my node red flow and send that info to influx and then I can monitor that with grafana.
Good thinking Offgrid.
All I see pre the anomaly is I am charging the battery, the battery box humidity and temp rise slightly.
Here is a trace of the main parameters:
Shunt Amps ±
Battery box temp
A longmon version is visible on the Cell Telemtry page.... I've made 5 or 6 different longmon purchases over the last 3 years - the last one was early this year. I notice that all my versions are the same - "HW6.2 FW2.2 BL2.0".
I did damage a couple of longmons (waco readings) by touching a screwdriver shank across the edge of a longmon (on it's body) and the buss of a pack while tightening a screw. The black coating doesn't necessarily protect against shorts? or whatever happened.
Could it be possible that that longmon has moisture on it and is shorting out contacts when the temps right enough??
Perhaps you could simulate it by warming up the air around that longmon and see if it repeats the issue.
I wouldn't blow warm air directly on it as that could possibly dry out the board, or possibly keep the moisture from effecting it. Not sure how you could heat up just that longmon and nothing else, tho
Thanks for your input and possible causes. I have checked all of the potential issues making sure of no pinched wires, shorts, loose connections, and short of giving it a humidity shower with a hose have not really found anything. I will let it float so to speak and keep a close eye on it as it hasn't reared its ugly head again in the last 6 days. I will attribute it to a neutrino strike and call it good after all the longmons surely don't have any ECC memory or components.
On another note I almost made it into the 100A charge club just 2.9A shy of it. Maybe next year
OK this is weird. The anomaly happened again. This time on pack #38. I actually caught it in the act. This time humidity and all box parameters were nominal. Nothing drastic other than Batrium was spazzing out on cell/pack 38. Voltages where jumping from ≈3.9 to ≈3.6 in a very random manner. I did a reboot of the Batrium system also did a longmon sync where you can see the line @ ≈16:10 didn't change anything. About 2 hours later it was as nothing ever happened. Of course I have been paying close attention to see if cell/pack 4 would have an issue again so I spotted this right away. But if I had not been logging the data with influx I would have never been the wiser if I just depended on the Watchmon toolkit and only looked at it every so often . Very odd.
The spazzing in closeup.
that is really weird. almost looks like a bad connection. but I'm sure thats not the case. The only other thing I can think of is data corruption in the Com Bus but you'd think it happen to more cells then.
You know.... when I originally setup my longmons I used the longmon connectors but I had a 6-8ft distance from the edges of the battery to the Watchmon - so I used pairs of flexible wire (20awg?) and twisted the length with a twist every 1/2inch - 1inch. The communication would work OK for months but then just go 'crazy' and do weird things. It seemed like heavy humidity (like a rain storm forming) would be associated with things failing - e.g. this is under the house with vents to outside air - so heavy humidity 'could' have been a factor somehow.
Eventually I replaced this with regular cat-6 ethernet wires - just a stripped out twisted pair with no insulation. Wa la - not a single problem since. I don't understand why ethernet twisted wires did so much better than me just twisting some wires myself... but it did. The ethernet has more twists than I used - but the Batrium folks OK'ed my 1 twist per inch when I first set things up - so I don't know.
The moral of the story - are you using ethernet twisted wire pairs from the edge longmons to the central core?
There is also shielded Cat6 that may also help but Something that a batrium Log may have to be looked at as it could be as simple as a bad sector on a hard drive (or ssd) that may cause Issues with data.
Interesting So I dug into the Batrium log only to find that batrium does not keep a "minute" by "minute" log or so it seems. It does however keep Min and Max cell Voltages and of course cumulative parameters such as Peak charge/discharge current. So for the type of recording that I do with node red and influx, what I have in grafana is the only visual proof. The Daily session info did show the drop in min cell voltage down to 3.51V which should have been technically impossible with a min SOC of 73%. You can compare the day before and the day after for comparison. So I guess the PC SSD drive is OK as the data concurs with Batrium.
Here is the Daily History data pulled with the Extractor.