diy solar

diy solar

Overkill BMS question about over current load

2Big2B

Free Wheeler
Joined
May 7, 2021
Messages
73
Location
Chesapeake, Virginia
I am building an 8S 24v 100AH battery pack with an Overkill 8S 24v 100AH BMS for a "Bounder" model power wheelchair. The pack will replace a pair of 100AH 12v Gel cells in series.

The power wheelchair uses two massive 30A (?) electric motors. I just found a mention in the wheelchair's service manual that it can draw up to a 140A load from a stall (motionless) state to a Start (putting it into motion). Otherwise at full throttle the load will be under 60A.

Obviously the original 24v 100A SLA Gel Cell battery configuration didn't need a BMS - just circuit breaker/fuse protection.

My question is, if I am pulling a momentary 140A load through the Overkill 100A BMS, will it see that as a dead short and shut down? Will that be a problem I had not considered? If so, what would be the best work around - bypass the BMS and draw the load directly from the battery pack? Get a bigger BMS?

Naturally I have emailed this question into Overkill support - which is probably overwhelmed, so I am asking here too. I hope is not just a newbie stupid question. So what does a BMS do with a current greater than it is spec'd for - and/or is there a programable parameter I can set to define the maximum current draw as an over ride.

With the power switch on but the chair not moving, the chair should draw less than 120mA. Next, drive the chair on flat level hard terrain. The BOUNDER should draw between 12A and 20A from the batteries at full speed. Chairs geared for slower speeds and with lighter users will tend towards the lower readings. When accelerating, climbing hills or on soft surfaces such as carpets, the power consumption will increase. At stall, the current from the batteries can exceed 150A. When going downhill, current should be less than 10A except when the joystick is released, currents can increase as the chair slows to a stop. When turning on a hard level surface, less than 15A should be required, and less than 25A on indoor/outdoor carpets
 
Last edited:
My question is, if I am pulling a momentary 140A load through the Overkill 100A BMS, will it see that as a dead short and shut down?
If the 140A surge is only for a couple seconds, you should be fine. The overkill has two separate over current settings. One that allows a surge for a few seconds before it cuts out and a second that will shut down almost immediately if it sees a short (Huge spike).

If you have not already purchased, I would advise getting the 120A version just to be a bit more on the safe side. In my opinion, the price difference is small enough that I would always buy the 120A instead of the 100A version...... but that is just me. Other folks look at it different.

BTW: What is the normal run-time current for the motor?
 
Thank you.

I wasn't aware that Overkill made anything greater than 100A for an 8S 24v BMS. I put my order in on Sunday, paid up front, and it is being processed (in backorder). I emailed same questions to their Tech Support (Steve, no doubt) and am hoping for a reply. I would upgrade to a bigger BMS if one is offered.

I am uncertain exactly what the specs are for the motors - tbd - but I am guessing 30A each, so that The BOUNDER should draw between 12A and 20A from the batteries at full speed. This power wheelchair is built like a tank, and built for speed (9+ mph), but I doubt I would ever be running any continuous 20A draw anyway unless I was on a racetrack!

... and with the new LiFePo4 battery pack at less than 50lbs it will reduce the chair's gross weight by at least 100 lbs!
 
Last edited:
Well firstly I don't have the wheelchair yet. It is a new old stock auction item w/o batteries. It will be delivered by motor freight on Thursday. So far I haven't had eyes on it.

Secondly I haven't invested in a DC clamp meter yet and probably would never need one after I get this all set up. The wheelchair has its own high end onboard controller with all the safety features that takes care of everything, but designed around a 24v gel battery pack. I will add an ammeter w/hall sensor to use as an accurate "fuel gauge", as the onboard voltage metering will be useless to monitor remaining capacity with LifePo4 chemistry. I will mount it next to the joystick control.

As I continue to read up on the subject through this forum I am coming to an opinion that a 140A load spike will not shut down the BMS. The issue seems more about thermal load on the BMS - which I doubt will be an issue in this application. I think that this 24v 100A BMS working a prolonged 12A to 20A maximum load - in the unlikely event that I were to run full speed continuously for more than a few minutes at a time - should probably not cause the BMS heat sinks to overheat.

Meanwhile I have my eye on a HP 6286a desktop Power Supply on an auction going cheap which I might use for tweaking. I probably won't use it much after I set things up though, but if it goes cheap ...?

So I will probably be fine with this overbuilt Overkill 100a BMS.

What do you think?
 

Attachments

  • ammeter.jpg
    ammeter.jpg
    119.6 KB · Views: 7
  • HP power supply.png
    HP power supply.png
    196 KB · Views: 7
Last edited:
So I will probably be fine with this Overkill/overbuilt 100a BMS.
Yes, you are probably fine.

You mentioned that in your reading you discovered the motor will draw 150 on startup surge. We can work backward from this.
The rule of thumb is that the locked rotor current (start surge) is 4-5 times the normal run-time. So... the runtime would be 30A-38A. This is higher than your 12-20A estimate but still well within the capability of the Overkill.
The wheelchair has its own high end onboard controller with all the safety features that takes care of everything, but designed around a 24v gel battery pack.
The LiFePO4 will run at a slightly higher full-charge voltage than the LA Gel.... but it should be close enough that it would not cause an issue.

as the onboard voltage metering will be useless to monitor remaining capacity with LifePo4 chemistry.
Do you happen to know the AH rating of the original Gel pack?

I will add an ammeter w/hall sensor to use as an accurate "fuel gauge",
You need a battery capacity meter. The picture you linked appears to have battery capacity on the display so you should be OK with it.
(Could you provide a link to the product?)

Note: A battery capacity meter does an integration of current in and out over time to determine the total capacity in the battery.
 
Meanwhile I have my eye on a HP 6286a desktop Power Supply on an auction going cheap which I might use for tweaking. I probably won't use it much after I set things up though, but if it goes cheap ...?
Just looking at the controls and displays, it looks like it tops out at 24V. It would probably work fine for top-balancing the cells but would be too low of a voltage for trying to charge the 24V pack.

Having said that... HP made some great product so if you can get it cheap, grab it.
 
Good Stuff!

...and I agree with you about the miscalculation of estimated running amps. According to the Owner's manual the original batteries are a pair of Group 27 12v 105A Gel batteries. Being SLA their voltage range is a wide swing from about 11.5 to about 14.v I think. Twice would be in the 23v - 28v range. The LiFePo4 pack will be 8 Liitko 3.2v 100A cells in a happy series to supply a tight nominal 25.6 v with BMS protecting against 17.6 or 32v - if the bms sees it that way (?). The main load is against what I believe to be 30A 4 brush (?) heavy duty electric motors. Per the manufacturer's email reply to me about it:

"Your chair actually has the faster ST3 motors for speed. It is currently set up at 9 mph".

So they were a special order and may not even be available now as an option. I will know more on Thursday when I get eyeballs all over it.

There is a Youtube channel called Totally Normal. Dan, its owner, is all about building custom power wheelchairs from parts he collects. Then he rides around in them with his camera gear and takes his audience on interesting wheelchair adventures around the Portland, OR area. His favorite steedi is his 21st Century Scientific Inc "Bounder" power chair set up for off road. His has a different from mine with a maneuverable seating platform, chain drive probably driven with a pair of 25A electric motors and is powered by on a LifePo4 25.6V 100A factory installed battery pack that fits into the sliding battery tray under the seat.

He has a video worth watching:

MEGA VLOG 402: inside a LITHIUM bounder battery


Inside Bounder LiFePo4 pack.png

In a chat he told me that 10A is the maximum size charger that can be used through the port on the chair. 21st Century's tech support says if I swap in a lithium power pack the battery level indicator won't work properly - it just sticks, thanks to the tight LiFePo4 discharge cycle. Hence Dan uses that ammeter gauge.

Charge-Discharge Monitor, DROK 0-90V 100A DC Ammeter Voltmeter, Battery Capacity Amp-Hour Watt-Hour Power Time Multimeter, LCD Display Digital Voltage Current AH WH Watt Meter Tester​

with Hall Sensorammeter.jpg

Visit the DROK Store

I ordered that ammeter power meter that the Dan guy uses on Amazon for about $30.
I also found ordered an 8S 24v 10A charger for it on AliExpress for about $50.

... and I won the Goodwill auction on the HP 6286a desktop Power Supply for $66 + $15 shipping. That will be handy for top balancing. When I am done with it I will offer it for sale to someone who can really use it.
 

Attachments

  • Battery tray.png
    Battery tray.png
    982.3 KB · Views: 9
  • Circuit breaker diagram closeup.png
    Circuit breaker diagram closeup.png
    598.5 KB · Views: 8
  • 8s 24v 10 charger.png
    8s 24v 10 charger.png
    525 KB · Views: 7
Last edited:
I would assume the 150 amp draw would be if you popped on full throttle to do a wheelie or some stunt like a 16yo boy messing around.
If you ramp the throttle like a normal person I doubt you will have any issues.

On second though it might be OK to have this throttle limiter ;)
 
The 140A spike apparently would occur as power is applied to the motors and the brake solenoids are released. Otherwise, according to the Owner's manual the amp draw would range from 12A -20A, which seems like an underestimate to me. Perhaps the manual is assuming smaller motors.

This thing will be even more frisky without having to carry the 100lbs of lead in the batteries this LiFePo4 pack is replacing!
 
The LiFePo4 pack will be 8 Liitko 3.2v 100A cells in a happy series to supply a tight nominal 25.6 v
Depending on how high you charge or low you discharge, LiFePo will run ~ 3v to 3.6V per cell. That is 24V to 28.8V in an 8S pack. You can charge them to a higher voltage, but the additional storage is negligent. Typical Gel would not get above 28.4.

the original batteries are a pair of Group 27 12v 105A Gel batteries.
Lead Acid can only be discharged to ~50% so your pack will no only be a 1/2 the weight, but it will have about 2x the effective capacity.... Nice.

I could not enlarge the pics of the charger, but from the thumbnail it looks like one of the many ebike chargers on AliX. Those units have a charge target of 29.2V (3.65V/cell) That is higher than I like and certainly higher than needed. Once you get to 3.6V/cell, you are not going to get any appreciable additional storage. I think you will find that the BMS will cut out before the charger does. You could adjust the BMS to a higher max cell voltage but once you get that high, one of the cells is probably going to run away so the BMS max cell voltage would have to be pretty high to keep the BMS from cutting out.

Note: A lot of the 'Dumb' BMSs have their max cell voltage fixed at a really high voltage, and the systems work just fine...... I just like to keep it lower.

1621405257223.png
 
Here's more detail on that charger:

Depending on how high you charge or low you discharge, LiFePo will run ~ 3v to 3.6V per cell. That is 24V to 28.8V in an 8S pack. You can charge them to a higher voltage, but the additional storage is negligent. Typical Gel would not get above 28.4.


Lead Acid can only be discharged to ~50% so your pack will no only be a 1/2 the weight, but it will have about 2x the effective capacity.... Nice.

I could not enlarge the pics of the charger, but from the thumbnail it looks like one of the many ebike chargers on AliX. Those units have a charge target of 29.2V (3.65V/cell) That is higher than I like and certainly higher than needed. Once you get to 3.6V/cell, you are not going to get any appreciable additional storage. I think you will find that the BMS will cut out before the charger does. You could adjust the BMS to a higher max cell voltage but once you get that high, one of the cells is probably going to run away so the BMS max cell voltage would have to be pretty high to keep the BMS from cutting out.

Note: A lot of the 'Dumb' BMSs have their max cell voltage fixed at a really high voltage, and the systems work just fine...... I just like to keep it lower.

View attachment 49582
Here's some better detail on the charger:

29.2v Charger.png


I wonder if the charger's voltage could be adjusted down a volt?
 

Attachments

  • charger guts.png
    charger guts.png
    754 KB · Views: 7
I wonder if the charger's voltage could be adjusted down a volt?

I can't tell from the ad on Aliexpress. The ad lists different voltages for different battery types. It also talks about adjusting for when it goes into float..... but it does not say anything about how the chemistry is selected or the 'start float' adjustment is made. When/if you find out I would be interested in knowing.
 
I can't tell from the ad on Aliexpress. The ad lists different voltages for different battery types. It also talks about adjusting for when it goes into float..... but it does not say anything about how the chemistry is selected or the 'start float' adjustment is made. When/if you find out I would be interested in knowing.
I am sort of taking it on faith that it is specifically a 8S 29.2v wheelchair charger. It sounds perfect...

8s 24v 10 charger.png
I hope it is, anyway. If the ammeter is accurate enough I should be able to unplug it at whatever point I like to charge it to 80% or whatnot. We will see. As a layman I don't know if I could just tweak the charge voltage with a bridging resistor or swapping out a transistor somewhere on the board. Or maybe I could devise some kind of an automatic shut off circuit using a shunt. Or is that just fantasy?

In the meantime, while I am waiting on a slow boat from China I plan on swapping in my good pair of 12v 55Ah gel batteries from my current Golden Compass GP620 "house chair". It will give me about half capacity of what the Bounder is designed for, but will be sufficient temporarily I think, if I keep my wanderings with it local. I will swap back in the pair of the fading batteries into the house chair - which I have been maintaining with a "Battery Minder". That way the house chair will still run fine, but with ever smaller range for around my apartment. hopefully I won't burn out those newer 55ah batteries that I am going to put into the Bounder temporarily so that I can swap them back. I might eventually make another LiFePo4 pack for later when I need new batteries for the house chair. Perhaps by then I will want to upgrade the Bounder to a higher pack. Like computer memory, prices will only go down (I hope) as the new wave of technology replaces the obsolete SLA technology...

My next part of the project - how I am going to best package my LiFePo4 battery pack for mobile use, and if I need to devise some cute adjustable wooden crate spring loaded battery box using long screw bolts, like some DYI builds I have seen, is still on the mental drawing board...

Something like this:

crate package.png
 
Last edited:
10 amps isn't a very high charge rate. .1C. If you don't discharge very far then it won't take too long to complete a charge. But a deep discharge is going to take a while.

There are plenty of charger/converter units you can buy that would have a higher charge rate. The converter I use in my RV is rated for 55 amps. That's too high for you, but great for my 560 Ah battery bank.
 
10 amps isn't a very high charge rate. .1C. If you don't discharge very far then it won't take too long to complete a charge. But a deep discharge is going to take a while.

There are plenty of charger/converter units you can buy that would have a higher charge rate. The converter I use in my RV is rated for 55 amps. That's too high for you, but great for my 560 Ah battery bank.
I had been looking into that. I just bought a HP 6286a Power Supply that I will use for individual cell top balancing on an auction, cheap. If I were to plug in a charger into the chair's connector greater than 10A it might overload the wiring and/or its control module. If I were to bypass that and charge directly to battery I would then need to have to let the battery sit for awhile to let the BMS to re-balance the pack. I don't know if I would attempt that by charging high amps through the BMS. Maybe. But I figure I will just go with the individual cell top charge, then parallel top charge the pack, then set it up and install - then plug it in through the connector @10A and let the BMS rebalance. I doubt I will ever really come anyway near a full discharge in using the chair. An overnight charge @ 10A should be fine. Its not like I am commuting with it anywhere.

I suppose I could try using my car battery charger very carefully to get a head start, but I doubt it is worth the risk. I think it has a 12/24v setting with 2A, 10A and 50A settings. i wouldn't trust that dirty power supply though.

HP power supply.png

The Keysight 6286A family of convection cooled, 5 1/2 inch, half rack, constant-voltage, constant-current power supplies is available in 100 Watt and 200 Watt power ranges. All models have individual 10-turn voltage and current controls.

  • Remote sensing and programming
  • Auto-series, auto-parallel, and auto-tracking
  • Front and rear output terminals
  • Floating output to serve as positive or negative source
  • Output ratings (0 - 55 °C): 0-20 V, 0-10 A
  • Ripple and noise (20 Hz to 20 MHz): CV: 500 uV rms, 5 mV peak-to-peak; CC: 5 mA rms
  • Line regulation: 0.01% + 1 mV, 0.05% + 1 mA
  • Load regulation: 0.01% + 1 mV, 0.05% + 1 mA
  • Physical Specifications: Weight: 11.3 kg (25 lb); Dimensions: 210 mm W x 131 mm H x 435 mm D (8.25 in x 5.16 in x 17.13 in)
(I won this for $66 + $15 + $5 tax on Goodwill's auction last night. I will probably list it for sale up here on the forum for cost when I am done, if anyone is interested. I can see it being damned handy.)
 
Last edited:
When I top balanced my cells I put all 8 cells in parallel and hooked up the benchtop power supply. I didn't do it on a cell-by-cell basis. Nothing wrong with that approach, I just didn't want to fiddle around with it eight times.

You don't leave the battery alone to let it rebalance. Balancing typically happens during the upper stages of charging. The BMS does the rebalancing.

There are a lot of posts saying that letting the cells sit in parallel for a while will not rebalance the cells. They just don't work that way. It takes charging them to get the process moving.
 
I have watched countless videos of balancing the way you do, and also read about doing it cell by cell which is more trouble to do. I figure I will test each cell with my Fluke VOM and individually charge the cells to 80% with the HP power supply, then parallel balance like you, let it sit, retest, then reconfigure for 8S 25.6v 10A and finish charge balancing through the Overkill BMS I have on order. That should be the best way. I am also wondering if I should first try to do a discharge to near the lower limit first but not sure exactly what I would use for the load. I have a 300 watt car type inverter I might try hooking up to 4 cells at a time with a light bulb. Not sure what I would use otherwise...
 
Back
Top