Load/Capacity Test 1s90p cell

Hey guys;

So I have 14 - 1s90p cells.

Is it possible/practical for me to load/Capacity Test each 1s90p seperately?

I am asking this question because I do not have a BMS as of yet. I have been running experiments with:

Solar Panels -- 4 240 Watt 37Vo 8.5 Ao Wired in Parallel 37Vo 34 Amps(open)
MPPSolar Charger m(PCM60)
Ames Inverter 24V 6000W
2 7S90P in parallel -- 28V (24V) Battery ~360 Amp Hours (1 cell x 90 cells x 2 A/cell == 360Amps)

The above setup has been under load and PV recharge for about 3 days.

The individual 1s90p cells are starting to diverge more than is healthy.

Taking Voltage Measurements, I can see what are my low performing 1s90p cells.

I guess a more general question is:

How to trouble shoot the above problem of diverging voltages (unhealthy outcome).

I have shut the system down. I plan on pulling the lower performing 1s90p cells to individually check 18650s and fuses.

In putting the 1s90p cells together, I did not run the calculator for organizing 18650 cells. I sort of went on the philosophy of my processes being random enough.

Thoughts/advice/admonishments ???

First: you can say you have a 14s90p setup. No need to split it as 14x 1s90p.
Second: each 90p grouping is a pack, not a cell. Each pack is made up of 90 cells.
Third: Yes, you can load test each pack separately

So you are getting packs that are starting to get out of balance with the other packs. This would indicate that you have a weak cell(s) in those packs. You know which packs are getting low. The next step would be to try to figure out which cells are the culprit. This is very hard to do with a soldered setup. I don't know how your packs are built as I can't find any posts on them. Seems like your "Hillbilly" thread you never updated.

If you are using a classic U shaped bus bar on your pack like most people are doing, then possibly disconnecting half of the pack from the other half. So, electrically you have 2 45p in one pack. Charge them up and see which one holds its voltage longer or goes wonky compared to the other.
However, it's also possible that your voltage differences could be caused by blown/broken fuses. If you used fuses on your packs, check them to make sure they are all in tact. A few blown/broken fuses can cause the pack to not stay in balance with the others. AveRage Joe had that issue with a few of his packs.

What are the exact numbers? Is it just one pack that gets out auf balance or several? How big are the differences? You haven't sorted the cells, but have you tested them properly and know for sure that you only used decent cells within a certain capacity threshold?

Of yourse you can do a capacity test now. That would certainly tell you something useful. What you can do about that then is a completely different story. If you find out that the packs differ because their capacities aren't similar then you have to take everything apart and rebuild it. Maybe from scratch if it is really bad.

Another solution would be to bottom balance your battery. You take all packs down to 3.1V individually, reconnect them and charge them up until the first pack reaches 4.0V. You can also do 3.0V/4.2V or whatever you prefer, the exact number isn't important for the process as long as it is the same on all packs. You stop charging at this point and take the battery voltage, which isn't max voltage by then of course, and use that as the end of charge voltage for the battery. You still have to monitor this closely and make sure it stays in unbalance, so to say, but that way you can get away with uneven cells and without a BMS.

Even if you add a BMS now you have to make sure that the differences between the packs aren't too big for the BMS to handle. So capacity testing the packs now is probably a good idea anyway. You can make this a very efficient process with something like a iCharger 4010 or any other charger/discharger that allows bigger currents. Otherwise you can still do it, but it might take quite a while with a small device.

As said did you sort so the packs were at same total capacity from start?

Use a proper charger/tester and you can test them. Both IR and capacity if you want. I do that with my iCharger DUO. With that one I can push 20-40A without issues when discharging.

I logged each 18650 (capacity,begin/end voltage, resistance). I built a charging station using the standard chips and I have a RC Charger that ran 4 at a time with the charge/discharge and resistance.

I will send some pictures in a bit. I am at work currently. I have not pulled a low performing pack to examine yest (fuses, individual 18650 voltages).

What would a load/capacity DIY tester for a 3.7V 1s90p pack be composed of?

What would a load/capacity DIY tester for a 3.7V 1s90p pack be composed of?

Click to expand...

Really big load resistor. If you plan on doing a capacity check of the whole pack, then you are probably gonna wanna do it manually. If you plug it into a regular charger, it could take a day to process 1 pack. Maybe longer.
If you take a 100W incandescent light bulb (filament style) (or use several automotive bulbs hooked in parallel) and hook it up, monitor the amps going to it and monitor the voltage of the pack as they drop. Check the pack every 5 mins or so, just keep it consistent.

And again, you don't need the "1s" portion on your designations. It's already assumed

***********
If you take a 100W incandescent light bulb (filament style) (or use several automotive bulbs hooked in parallel) and hook it up, monitor the amps going to it and monitor the voltage of the pack as they drop. Check the pack every 5 mins or so, just keep it consistent.
**********

I can do the above with a 4.0V 180AH pack?

I guess, if I use minimum 6awg wire, heat should not be an issue (100W/4V = 25 Amps)?

Is a resistor still needed? I guess I am out of my knowledge range.

Just because the lamp is 100W it doesnt state its 100W at that voltage. The lamp will only be 100W at the specific voltage. You need to use Ohms law to calculate the wattage at the new voltage.

BobinTN said:

***********
If you take a 100W incandescent light bulb (filament style) (or use several automotive bulbs hooked in parallel) and hook it up, monitor the amps going to it and monitor the voltage of the pack as they drop. Check the pack every 5 mins or so, just keep it consistent.
**********

I can do the above with a 4.0V 180AH pack?

I guess, if I use minimum 6awg wire, heat should not be an issue (100W/4V = 25 Amps)?

Is a resistor still needed? I guess I am out of my knowledge range.

Click to expand...

25A / 90 cells = 0.277A/cell So it'd be safe

daromer said:

The lamp will only be 100W at the specific voltage. You need to use Ohms law to calculate the wattage at the new voltage.

Click to expand...

Exactly. So, aside from the above math, we'd have to do a different math to get a more accurate reading. If the bulb was pulling 25A truly, you'd still be safe. Since you'd be pulling far less, you're still safe.
That is why I said to make sure you monitor the amps during the test. Otherwise, you will get way wrong numbers if you just go off of voltage and "assumed" wattage.
You won't be dumping a full 25A into the 100W bulb. I'm going to guess more like 15A, maybe less.

If the lamp is 100W at 12V you got a resistance of:

100/12 = 8.3A
U/I=R => 12/8,3= 1.44Ohm.

At 4V => 4/1.44 = 2.77A => 2.77 *4 = 11W

So basically a 100W 12V bulb will at 4V produce/use 11W.

Hey guys;

correction: these are 80p packs, not 90p as stated in previous posts.
I have removed two of the low performing packs (see pictures).

After checking voltages of cells, one pack had 5 broken fuses. I don't think the fuses blew, I think they were abraded due to me struggling with a battery holder design. Those 5 cells had 4.12V charge. Which leads me to think that they never participated as part of the battery thus leading to this pack having a lower voltage overall after a load was applied (5 cells short of 80p pack). I am happy with this result.
I have hooked up this pack to my thunderhead Ac6 balance/discharger/charger. It will charge at 5Amps.

The other low performing pack has me puzzled and worried. There are no broken fuses on this pack. But there are 33 cells that are at 4.12V or close 3.9V. The rest of the pack is at 3.64V. Like I said, no broken fuses or connections found.
But, this is one of the 1st packs I made. I can tell, because I had used brush-on flux (plumbers flux). This has produced corrosion on my copper bus wires. I have/had done some re-soldering on this pack. I had sanded the bus bars before re-soldering (but I can't say I did a really thorough job).
I am now in the process of reviewing all the packs to find non-contributing cells.

Apparently, at least for me, I need to put my packs under load, then pull them off load and look for 'non-particpating' cells.

Great find on the first pack. Looks like we diagnosed that one properly!

Now, the second one, that's a prime example of how much resistance corrosion can have on a connection. It didn't just slow down the voltage, it stopped it all together. This is definitely the main reason for using the right flux/solder on these things

Btw, there's waaaay to much free space in that trailer. You need to add more packs

l have the fever. More packs on trailer.

I re-built one of 2 - 7s80p batteries (No BMS).
1. My Inverter is a 24V inverter.
2. I charged the battery up with my solar Charger (PCM60) (44240W solar panels on roof hooked up in parrallel (some shade issues)).
3. Ran a space heater(1500W) with a Watts-Up connected.
4. I got about 2 KWh out of the battery.
5. so that means about 83Amps/hour out of my 7s80p battery.

I guess some of my batteries energy is lost powering the Inverter and the pcm60.

Measuring the final Volts on each pack ranged from a low of 3.29V to 3.76V

Does that sound like what I should be getting if the battery is a healthy batter?

:-/

No, not really. Well, sort of, it depends on the definition of healthy. For a battery that has been randomly cobbled together it could be a valid result. Looks like you might have used pretty low capacity cells (between 1000 and 1500mAh) and/or your 80p packs don't match very well in terms of their total capacity.
What are the exact results for the seven packs? By comparing them you can get an idea which ones are the weak link. I guess the trigger for cutoff was that pack that is now at 3.29V because it hit 3.0V, or whatever you have used, under load first?

Earlier you have asked about testing individual packs, have you done that or was this the first test of these packs?

In addition to DarkRavens comment, testing each packs voltage will yield a better result of what's going on. It's possible 1 or more of the packs are actually lower voltage than the others in the string.
This is part of the reason for a bms. Perhaps you should use some digital voltmeters on each pack for awhile to keep an eye on them; like Average Joe did at first.

This was the first test.
**********
Trigger for inverter cutt-off
*********
I guess this is the big concern. The Trigger is the Inverter pulling under 24V from the whole battery (7s80p).
So, potentially, one pack could be pulled down to an unhealthy level before the battery as a whole gets down to 24V.
And,
The inverse. The PCM60 charging up to 28.8 volts the whole battery, over-charging the high performers and while trying to bring up the under-performers.
Thus the need for balancing the packs (on the front end using a capacity calculator) and on the back-end (a BMS).

Please read the above with sentences ending in question marks. Please correct if my assumptions are wrong.

That's absolutely correct and that's why you match the packs and/or do balancing and/or use a BMS. The maximum for 7s is 29.4V and ideally this is 7x4.2V but it could also be 5x4V + 2x4.7V which would be bad. Same on the discharge, you shouldn't go below 21V, 7x3.0V, but even at your inverters 24V cutoff it could be like 6*3.2V + 1.8V which would be bad as well.

I have some digital voltmeters, as Korishan recommends, and will wire them up and run through another Charge/discharge/Charge sequence of the 7s80p (paying better attention and recording changes perhaps every 5 minutes)

Thanks for the support guys.

---

Battery now has voltmeters.

BobinTN said:

I have some digital voltmeters, as Korishan recommends, and will wire them up and run through another Charge/discharge/Charge sequence of the 7s80p (paying better attention and recording changes perhaps every 5 minutes)

Thanks for the support guys.

---

Battery now has voltmeters.

Click to expand...

---

Hey Bob,

Gotta say I admire your determination and desire to get this figured out. I've learned a lot from reading this thread too. I'm really close to building out my first 14s string, so this has been really useful for me.

Cheers.