Battery Capacity Tester

[robby]

Digging up this old thread for a Florida Man (Poor man) Battery Capacity Tester

Victron Smart Shunt
Standalone Solar Assistant
2 x 2000 Watt 120V Immersion Water Heaters

I have the first two items. The reason I picked the Immersion Heaters is my pool in is upper 50F degrees. We cover it with a solar blanket and try to get it warmed up above 72F. Sometimes this takes until mid April. I don't expect the immersion heaters to heat up my pool but it would help. Pool makes one hell of a dummy load for heat.

120Vac 2000 Heaters should be around 16.7A or 7.2 Ohms

So two is 3.5 Ohm's. 51.2V / 3.5 Ohms = 14.63A or 789W

Why not do 4 for $100

This way I am not wasting the energy, heating the pool, Testing a battery capacity, and cooking pork shoulder on the grill.

Yes, I am Serous.. What's not to Love?

Your setup is what I would use if I had a need to test a lot of batteries.
Never thought of 120V 7.2 Ohm water heater elements, I was think 19 Ohm 240V elements with four in parallel attached to a water drum. The 120V sound much better suited for the task.

 

[BarracudaBob]

Your setup is what I would use if I had a need to test a lot of batteries.
Never thought of 120V 7.2 Ohm water heater elements, I was think 19 Ohm 204V elements worth four in parallel attached to a water drum. The 120V sound much better suited.

I was looking at those and I'd have to build a whole structure a pump whereas these immersion heaters I can take a 2x4 or to drill holes and hang them into the water.

 

[BarracudaBob]

I just updated the link the previously one was wrong.

So not just for load testing during a hurricane if you want to warm bath or something we can throw those in there My water heater pulls way too much power for my current system.

I guess I could always fire up a generator but where's the fun in that.

Alton Brown always talked about single purpose things in the kitchen where I like to get dual purpose out of everything.

2000W Immersion Water Heater

 

[fnnwizard]

So two is 3.5 Ohm's. 51.2V / 3.5 Ohms = 14.63A or 789W

Why not do 4 for $100

This way I am not wasting the energy, heating the pool, Testing a battery capacity, and cooking pork shoulder on the grill.

Yes, I am Serous.. What's not to Love?

You won't need 4 of them

2000W / 51.2V = 39A X 2 (units) = 78A / .90 (efficiency) = 86A likely for boh units running at max.

If you are testing batt units that are already built and have BMS, use the BMS to log via SA, or better yet HA. Just need to calibrate V and A against a decent meter to verify.

 

[BarracudaBob]

You won't need 4 of them

2000W / 51.2V = 39A X 2 (units) = 78A / .90 (efficiency) = 86A likely for boh units running at max.

If you are testing batt units that are already built and have BMS, use the BMS to log via SA, or better yet HA. Just need to calibrate V and A against a decent meter to verify.

at 51.2V they wont be 2000W because the voltage is far lower than 120Vac(rms).

I see what you are thinking.. I wasn't going to use an inverter this was a DC to DC connection so 100% transfer of energy into a resistive heater.

The idea is that I can test batteries directly with a shunt and resistor that is water cooled by my pool. Keep it simple

I used to use a inverter but it would drop out before the battery BMS would. This way the BMS takes full control.

Totally agree about calibrating everything.

 

[Brucey]

I'm going for max efficiency with my immersion heater. Point of use. Red solo cup to try and stop water from draining past leaky plug

 

[BarracudaBob]

I'm going for max efficiency with my immersion heater. Point of use. Red solo cup to try and stop water from draining past leaky plugView attachment 277838View attachment 277839

That is my backup plan!

 

[fnnwizard]

at 51.2V they wont be 2000W because the voltage is far lower than 120Vac(rms).

I see what you are thinking.. I wasn't going to use an inverter this was a DC to DC connection so 100% transfer of energy into a resistive heater.

The idea is that I can test batteries directly with a shunt and resistor that is water cooled by my pool. Keep it simple

I used to use a inverter but it would drop out before the battery BMS would. This way the BMS takes full control.

Totally agree about calibrating everything.

Are you planning on testing built packs from various labels? Individual cells?
DIY Packs?

I've tested batteries in many ways and can provide some info into doing it "good enough" at low costs, always with logging via raw or graphical data and even video too.

 

[BarracudaBob]

Are you planning on testing built packs from various labels? Individual cells?
DIY Packs?

I've tested batteries in many ways and can provide some info into doing it "good enough" at low costs, always with logging via raw or graphical data and even video too.

My main interest is testing 48V battery packs accurately without to much additional expense. I am also interested in 12V AGM, SLA, and LFP but that would need to be a portable system.

The smart shunt and fluke volt/amp meters are my measuring devices. Solar Assistant and Bluetooth app for stat storage.

Instead of heating up a resistor on a heat sink with fans heating the air the immersion heaters in the pool is a double win. I don't need to heat the air up. The biggest challenge we have in South Florida is cooling the house down.

I posted it looking for other interesting ideas and feedback.

So what ideas do you have?

 

[fnnwizard]

My main interest is testing 48V battery packs accurately without to much additional expense. I am also interested in 12V AGM, SLA, and LFP but that would need to be a portable system.

The smart shunt and fluke volt/amp meters are my measuring devices. Solar Assistant and Bluetooth app for stat storage.

Instead of heating up a resistor on a heat sink with fans heating the air the immersion heaters in the pool is a double win. I don't need to heat the air up. The biggest challenge we have in South Florida is cooling the house down.

I posted it looking for other interesting ideas and feedback.

So what ideas do you have?

Testing already built systems are very easy and minimal cost. You essentially use the BMS to log individual cell data into SA or HA and capture on a graph all the needed info.

I've converted some older (mainly JK & JBD) BMS's into logging devices, but most all of them can be converted now days.
Pinned (stalled) starter motors for peak current measurements.

Also have lab equip, but those are a pain to setup to also capture individual cell stats.

I saw your E/W rack bat thread, but afraid to start reading it... too long.. hint: can you update first post to bullet main ideas? What BMS does that use?

Do you have an old computer lying around you don't use?

 

[fuzzmunky415]

I saw your E/W rack bat thread, but afraid to start reading it... too long.. hint: can you update first post to bullet main ideas? What BMS does that use?

It's one of those threads...sometimes buried in a discussion, some real nuggets and Ah-Hah! moments present themselves...it's well worth the time investment. And that's from someone who usu. does 2X on Youtube vids.

 

[BarracudaBob]

Testing already built systems are very easy and minimal cost. You essentially use the BMS to log individual cell data into SA or HA and capture on a graph all the needed info.

I've converted some older (mainly JK & JBD) BMS's into logging devices, but most all of them can be converted now days.
Pinned (stalled) starter motors for peak current measurements.

Also have lab equip, but those are a pain to setup to also capture individual cell stats.

Do you have an old computer lying around you don't use?

I didn't want to go off topic on the E-W thread but we can do that there. I believe it is agreed by most that the Smart Shunt is accurate enough for testing. I like the idea of a non-reactive purely resistive load at an inexpensive price.

The BMS calibration, setting the 'coulomb counter', and SOC meter is one area of interest as well as testing them every year for degradation.

I am not making any money off it so budget is a concern. Makes for a good conversation on the thread as well as Will calling me out on my methodology.

Making it "More Better Every Day"

 

[BarracudaBob]

some real nuggets and Ah-Hah! moments present themselves

I have a couple of ideas that I would love to try

 

[pollenface]

I use one of these and test down to inverter cut-off voltage (10.5v) with a 500w inverter and 40w soldering iron

Pardon our interruption...

 

[BarracudaBob]

I use one of these and test down to inverter cut-off voltage (10.5v) with a 500w inverter and 40w soldering iron

Pardon our interruption...

I have some smaller stuff from the old quadcopter days we used to build them.

 

[BarracudaBob]

I'm interested in your ideas what would I do with the old burner laptop?

 

[Daddy Tanuki]

I still think we should peak power test a bunch of LiFePo4 cells with my solar powered electric chair... the way DOGE is going we are going to need it sooner then later for some of these bureacrats...

 

[fnnwizard]

Makes for a good conversation on the thread as well as Will calling me out on my methodology.

Details, link?

I have a couple of ideas that I would love to try

Share the ideas?

Making it "More Better Every Day"

Laptop is not the best to use. The computer would run Home Assistant (HA) and using a few inexpensive, but very capable devices, you can communicate with just about every BMS out there to grab and log data to 3 decimal places ( of course only as accurate as the BMS) .

You would also, if you want, get a spare bms and set it up to test other configurations with and be able to gather cell level data. You can have HA do the coulomb integration for you and compare it with the BMS's integration, though both would use the BMS data of V and current for the equation so only variable would be sampling rate...

The only issue with the resistive load is as the V drops the current gets a bit lower. You usually get a bit more energy testing this way, but for all practical matters it won't make much difference.

 

[BarracudaBob]

Share the ideas?

The Victron Bluetooth app and the Solar Assistant will do. Don't want to mess with Home Assistant right now. I can export S-A data and see the discharge points 90%, 80%, ect and get a truly tested capacity = voltage to reprogram into the BMS for truly accurate SOC.

As the Voltage Drops from 56.8 to 42.0 the current will change a little 32.46A to 24A over the discharge.
I could easily double that number 64.92A to 50A with more loads in parallel. The water cooled loads sound like a solid idea.

 

[BarracudaBob]

Found some 1500 Watt ones. Ordered.


Update: They came in today and I measured their resistance. 7.9, 7.7, 8.3, 8.2 Ohms so combined it is 2.006 Ohms. This would put a 100ah 51.2V battery at around 0.25C (22A - 28A) discharge. Perfect for a test. Most of my batteries are at that or below in production.

 

[BarracudaBob]

Warning Deep Dive:

So I did a test to verify how the "Coulomb Counter" on the smart BMS stop charging & discharging. I'll do another test in the future on the discharge side. With the weather not getting everything to 100% in a while I turned off my loads and allowed the PV to take everything up to 100%. I have EG4 LifePower4 V1 and the Eco-Worthy V1 & V2 Rack Batteries. The BMS 100% Reset voltage on the LP4V1 is 56.6V and the E-W is 56.8V. I set the AIO to stop at 56.6V for bulk and float so not to Pack OV the EG4. I changed the E-W to match the EG4 to 56.6V. So when all the packs get to 56.6V the BMS automatically registers 100% SOC and stops the charging. The E-W will continue with up to 1.5A until the Coulomb Counter get to the Rated Value programmed in the BMS.

I changed the Cell Capacity from 100AH to 105AH to see how much the BMS would push into the cell. Remember after the 100% SOC Reset Voltage it will only do 1.5A Charging until the Coulomb Counter shuts it down. If the Cells won't take any more at the 56.6V charging would stop. If the cells got close to 105Ah charging would stop.

So here is the result. You can see where the BMS kicked into "Full Charge Protection". I can see where if these number quantities aren't carefully set in the beginning where any intelligent BMS could get lost. Setting the Rated/Full capacity in the beginning when doing charging/discharging balancing would effect it's ongoing operation. Once I reached the top 100% where it stopped I reset everything to 100% and 100AH. My goal wasn't to squeeze and maximize capacity for future use but to accurately set what was on the top end 100%. Now that there is headroom in the cell capacity.

So here is a thought experiment for the future: If the cells discharge down to 100AH and they haven't reached the Cell LV disconnect will the BMS turn off the Discharge MOS leaving capacity on the table. You would think the Pack/Cell voltage would be above the LV threshold. If it keeps going to the LV disconnect threshold(s) and ignores the Coulomb Counter then would it keep going past zero or blow the 20% SOC disconnect threshold. I have seen it get below 20% even though the discharge is set for 80%.

When the E-W charges back up it will continue past the 100% SOC & 100AH but at a reduced current set by the 100% BMS SOC Voltage of 56.6V (programmed). This is as long as it is below the max charge voltage. Factory max charge voltage was 58.4V.

Looking at the image again i see the Charge MOS still activated even though the "Full Charge Protection" state is activated. There is another setting showing any current under 300mA for a certain time will trigger sleep mode. So perhaps it wouldn't take any more current at that voltage and eventually shutdown.

Inquiring Minds would like to know.

 

[fnnwizard]

Warning Deep Dive:

So I did a test to verify how the "Coulomb Counter" on the smart BMS stop charging & discharging. I'll do another test in the future on the discharge side. With the weather not getting everything to 100% in a while I turned off my loads and allowed the PV to take everything up to 100%. I have EG4 LifePower4 V1 and the Eco-Worthy V1 & V2 Rack Batteries. The BMS 100% Reset voltage on the LP4V1 is 56.6V and the E-W is 56.8V. I set the AIO to stop at 56.6V for bulk and float so not to Pack OV the EG4. I changed the E-W to match the EG4 to 56.6V. So when all the packs get to 56.6V the BMS automatically registers 100% SOC and stops the charging. The E-W will continue with up to 1.5A until the Coulomb Counter get to the Rated Value programmed in the BMS.

I changed the Cell Capacity from 100AH to 105AH to see how much the BMS would push into the cell. Remember after the 100% SOC Reset Voltage it will only do 1.5A Charging until the Coulomb Counter shuts it down. If the Cells won't take any more at the 56.6V charging would stop. If the cells got close to 105Ah charging would stop.

So here is the result. You can see where the BMS kicked into "Full Charge Protection". I can see where if these number quantities aren't carefully set in the beginning where any intelligent BMS could get lost. Setting the Rated/Full capacity in the beginning when doing charging/discharging balancing would effect it's ongoing operation. Once I reached the top 100% where it stopped I reset everything to 100% and 100AH. My goal wasn't to squeeze and maximize capacity for future use but to accurately set what was on the top end 100%. Now that there is headroom in the cell capacity.

So here is a thought experiment for the future: If the cells discharge down to 100AH and they haven't reached the Cell LV disconnect will the BMS turn off the Discharge MOS leaving capacity on the table. You would think the Pack/Cell voltage would be above the LV threshold. If it keeps going to the LV disconnect threshold(s) and ignores the Coulomb Counter then would it keep going past zero or blow the 20% SOC disconnect threshold. I have seen it get below 20% even though the discharge is set for 80%.

When the E-W charges back up it will continue past the 100% SOC & 100AH but at a reduced current set by the 100% BMS SOC Voltage of 56.6V (programmed). This is as long as it is below the max charge voltage. Factory max charge voltage was 58.4V.

Looking at the image again i see the Charge MOS still activated even though the "Full Charge Protection" state is activated. There is another setting showing any current under 300mA for a certain time will trigger sleep mode. So perhaps it wouldn't take any more current at that voltage and eventually shutdown.

Inquiring Minds would like to know.

View attachment 279725

Are you able to get cell level data while testing?
Is there a settings page showing all the configurable parameters?

 

[BarracudaBob]

Are you able to get cell level data while testing?
Is there a settings page showing all the configurable parameters?

There is windows software that I can use as well as the smart shunt and solar assistant to gather data

 

[BarracudaBob]

Will's quad box being fed with DC would make a nice interface by the pool for the quad immersion heaters. I could also use it in AC or DC mode. I have some 50A SO Cord.. Money is a little tight right now so I'll probably use some scrap materials I have around here.

 

[BarracudaBob]

Found some 1500 Watt ones. Ordered.
View attachment 279311

Update: They came in today and I measured their resistance. 7.9, 7.7, 8.3, 8.2 Ohms so combined it is 2.006 Ohms. This would put a 100ah 51.2V battery at around 0.25C (22A - 28A) discharge. Perfect for a test. Most of my batteries are at that or below in production.

Let the fun begin.