diy solar

diy solar

How do i capacity test with no equipment?

Lt.Dan

Solar Wizard
Joined
Dec 25, 2020
Messages
3,528
Location
Tulare, Ca
Hey guys, now that I've got my 16s 272ah CATL pack together and have put it through a few cycles, id like to do a capacity test, but I have no equipment to measure total current used.

I have the Smart Daly 200a BMS, which in the app shows current be drawn at that time, but its highly innacurate, as I will pull a constant 50a with the inverter, but it will bounce around, from 24a to 67 to 38 to 52, etc. The app does show a battery percentage that seems to be very accurate though, from 100% all the way to 0% (which i believe are based on my LVD and HVD in the settings)

My LV6548 does the same, it shows me instantaneous current, but not over a period of time.

What can I do? I'd rather not buy the little 150w capacity testers, as 14kwh of battery would take way to long and im far too impatient.

Suggestions?
 
Use a constant power load like a heat gun and multiply time by instantaneous load current.
 
Last edited:
Thank you! I've been looking for where Will got that tester from. I just ordered one ?
 
Thank you! I've been looking for where Will got that tester from. I just ordered one ?
Please let me know when you receive it. I ordered two ($11 ea with free shpd) last year from Ali and now that version of bare board is no longer available from Ali or Ebay, they now only sell the one in the case. So if this company still have them I will want to order a couple more.Thanks.
 
Last edited:
Hey guys, now that I've got my 16s 272ah CATL pack together and have put it through a few cycles, id like to do a capacity test, but I have no equipment to measure total current used.

I have the Smart Daly 200a BMS, which in the app shows current be drawn at that time, but its highly innacurate, as I will pull a constant 50a with the inverter, but it will bounce around, from 24a to 67 to 38 to 52, etc. The app does show a battery percentage that seems to be very accurate though, from 100% all the way to 0% (which i believe are based on my LVD and HVD in the settings)

My LV6548 does the same, it shows me instantaneous current, but not over a period of time.

What can I do? I'd rather not buy the little 150w capacity testers, as 14kwh of battery would take way to long and im far too impatient.

Suggestions?
Use a shunt, you should have one anyway. I use this one. https://www.amazon.com/gp/product/B0824X5MKM/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

Put a load on it and let it run until the BMS hits low cell voltage cutoff.
 
Please let me know when you receive it. I ordered two ($11 ea with free shpd) last year from Ali and now that version of bare board is no longer available from Ali or Ebay, they now only sell the one in the case. So if this company still have them I will want to order a couple more.Thanks.
Will do ? i chose the 12-14 day shipping for $0.11 lol
 
Wouldn't a shunt tell me my SOC just like the BMS already does? Or does it do anything else?

Id rather keep my system as simple as possible, without more connections and points of failure.
No, you're thinking of a voltmeter, not the coulomb meter I linked. It will show Ah, just like any of the little testers many use here with the fan.

I can tell you my BMS SOC is not as accurate as the coulomb meter I linked or the AiLi others here use (they are the same item). Close but if you really want an accurate test of capacity, you need to use a coulomb meter.
 
No, you're thinking of a voltmeter, not the coulomb meter I linked. It will show Ah, just like any of the little testers many use here with the fan.

I can tell you my BMS SOC is not as accurate as the coulomb meter I linked or the AiLi others here use (they are the same item). Close but if you really want an accurate test of capacity, you need to use a coulomb meter.
The Daly BMS SOC is not based on Voltage from what I can tell. With a constant load the percentage rate drops consistently to match. If it were based on voltage, it would be all over the place.

The Daly BMS does seem to keep track of AH used, but it doesn't show/record it. I wonder if hooking it up to a laptop and using the software will yield more results than the Android App?
 
This post and the ones that follow are about testing the capacity of my battery and three that were purchased from the same lot by @MrNatural22. The posts talk about two devices used to do it. Mr. Natural has, and used, the second of the devices. If I recall, Will Prowse has also used it.

Here's the link: https://diysolarforum.com/threads/s...rks-valence-for-dc-ac-power.20458/post-244206

My personal view of this is that Watt hours is the meaningful measure of capacity, and that the way to measure Watt hour capacity is to connect a device to the battery with a known draw in Watts and find out how long it runs.

The first thing I do with a video camera that I want to run on a lithium-ion battery is to set up the camera for a known draw and run it until it stops. Notwithstanding what's frequently said about voltage readings and lithium-based batteries, it's also useful to take regular voltage readings while doing this. In reality, voltage readings are quite useful in practice. You just have to know the battery's behaviour.
 
The Daly BMS SOC is not based on Voltage from what I can tell. With a constant load the percentage rate drops consistently to match. If it were based on voltage, it would be all over the place.

The Daly BMS does seem to keep track of AH used, but it doesn't show/record it. I wonder if hooking it up to a laptop and using the software will yield more results than the Android App?
I never said the coulomb internal to the Daly is voltage based. But I can tell you it isn't as accurate. I can tell you that the resistance of coulomb meters varies and all are subject to drift under the right conditions. But I can also tell you that members of this forum are nowhere close to meeting those conditions. In an EV, yes, you can get some drift under extremely high load conditions.

I thought you wanted simple, now you want to track Ah using a laptop?
 
x2 but if you need some cheap stuff anyways:

Someone asked what Will was using in this video to measure stats >

I found this for him, you can temporarily bolt inline. > https://www.fasttech.com/product/9615245-juwei-300v-100a-multifunction-voltmeter-ammeter

The second clamp meter, that is AC current only, just FYI.

The first is great and what you are looking for, but really only good up to about 80 amps.
 
I never said the coulomb internal to the Daly is voltage based. But I can tell you it isn't as accurate. I can tell you that the resistance of coulomb meters varies and all are subject to drift under the right conditions. But I can also tell you that members of this forum are nowhere close to meeting those conditions. In an EV, yes, you can get some drift under extremely high load conditions.

I thought you wanted simple, now you want to track Ah using a laptop?
I apologize, just the way I read it I assumed you meant it was voltage based. You are correct though, I don't know how accurate it is, but there is a "Zero Drift Calibration" in the app, is that what that is for?

I do have a Klein clamp meter that I bought just for my solar adventures. Could that help in this?
 
Use a shunt, you should have one anyway. I use this one. https://www.amazon.com/gp/product/B0824X5MKM/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1

Put a load on it and let it run until the BMS hits low cell voltage cutoff.
I second the shunt idea, they are pretty cheap.
This is what I use:

 
I think there's some misunderstanding going on here. All of these devices: the BMS current measurement, the $18 shunt from Alibaba, a Victron shunt, and a AiLi shunt on Amazon are "voltage based". All of them measure a small voltage difference across a low-value resistor in series with the flow of current and use that to compute current using Ohm's Law. The only one which is not using this methodology is a clamp ammeter, which works based on magnetic induction.

If you want any shunt meter to be more accurate:
- Calibrate the shunt resistance
and
- Calibrate the offset of the analog circuitry sensing that tiny voltage, or use a component like a TI INA chip which has extremely low offset by design
or
- Calibrate the combination of them (shunt + electronics) using a known reference current. You can do this on a good BMS, like the JBD ones from Overkill. You can also do this on anything, by just correcting the reading using a table that you have produced by passing known currents through it and recording its output reading.

These monitors are just doing an integral / summation over time for you, sampling many times per second. They have pretty displays. But, if you use a constant-power load and a calibrated monitor (or just a "post correction" table as I said above), then you don't need very many data points because: the battery voltage barely changes, the inverter efficiency barely changes, and the load power consumption barely changes. Even the most basic linear interpolation gives a very accurate depiction of the current and voltage profiles. I think if you already have a clamp ammeter with reasonable accuracy, a simple way to do this kind of test is:

- use a constant-power load, like a heat gun
- record an average current reading over 10-15 seconds every 15 minutes
- record how long it takes to empty the battery
- integrate your data over time (there are online calculators to do this, or do it manually using the trapezoid method, or use your TI-83 from high school math, or it's about 5 lines of python in a computer terminal)

EDITs:

Of course, using a clamp ammeter requires you to sit there and record data. Probably a good idea anyway if you're running a heat gun! If not, you could also set up your phone to videotape the ammeter output and review the video later, assuming your ammeter doesn't shut off after 10 minutes.

I'm not saying buying more electronics isn't useful; it could make your life easier. But now you have potentially 2 uncalibrated Chinese circuitboards instead of just 1. If your BMS already supports current calibration, you should just calibrate it and use that, or correct its reading using a table if it does not support calibration directly.
 
Last edited by a moderator:
Most shunts are pretty accurate, most clamp meters are ballpark. That's my experience with the cheaper ones, anyway.

Shunts can run hot when run near their capacity, since they are resistance based. Get a bigger shunt if that is a problem, generally speaking you want to run about midrange of the shunt rating. Shunts are quick and easy to calibrate in the factory, so accuracy is rarely a problem in bundled units.
 
Back
Top