Battery testing data

The key factor in battery life is depth of discharge, which we are controlling with the terminating Voltage. I didn’t immediately find a relation on discharge rate. Any of our battery experts want to chime in?

A deep discharge is a cycle use battery’s worst enemy.

Enyersys says their 12v AGM batteries are good for

1200 cycles with 30% Depth of Discharge
450 cycles with 50% DOD
175 cycles with 100% DOD

While there is conflicting information on the exact open circuit voltage and SOC some commonly found numbers.

70% SOC 12.3v
50% SOC 12.0v
10% SOC 11.5v
0% SOC 10.5V

With a long slow .9a discharge you will get very close to that 0% SOC and definitely under 10% as the voltage is not going to recover signifcantly after that .9a load.

With that 50a load for 15 min yes you will certainly create more heat in the battery but it will recover a lot of voltage immediately and more over time.

To really know which is going to do more damage to the battery you’ll need to measure both 30 min after the load is removed.

However I was specifically talking about the RC-300 which applies pulses of 40a loads and analyzes how the battery responds both during time it is loaded and resting and takes under a minute, so no significant discharge of the battery.

3 Likes

It sounds like I should change my testing to stop at 11.5 Volts, at least at the 12 Amp rate! That should get me comparative data AND weed out dead cells.

I’ll do a test with a 30 minute recovery time and see what I get.

I applied an “a miracle occurred” step and re-routed the air to blow over all the fins… ugh.

Did a test on the same battery, 12 Amp rate, 10.5Volt stop voltage. The recovery Voltage was 11.829 Volts. @Mr_V should I interpret that as between 10% and 50% DOD?

The one thing that most of the charts I’ve seen agree on is that 11.8v or 11.81v is a 30% SOC. 40% ranges from 11.9v to 11.95v . So I’d say it is safe to assume the test was to a DOD of about 70% to a SOC ~30%.

How many AH did you get with the 10.5v cut off?

More data…
12 Amp rate, 12.1 Amp hours, 11.83 Recovery Voltage
30% SOC, 70% DOD Some life impact
50 Amp rate, 8.6 Amp hours, 12.31 Recovery Voltage
70% SOC, 30% DOD VERY little life impact

Also, here’s a 100 Amp 10 second load test on a -good- battery. The one tested above.

2 Likes

Got some proper old-school lamacoid label action!

2 Likes

Great data. Not only does the 50a draw result in a lower DOD it is more like the load it will see in a robot.

I’ve been rereading this thread and trying to decide what direction to go, as my team definitely needs a battery tester. As I see it, keeping within an approximate $250ish budget, the options are:

  1. CBA - the “default” choice, well proven, more costly but a known quantity. Works out of the box. Limited discharge rate (with all the tradeoffs discussed above); the add on amplifier/extra load is beyond what I can justify.

  2. BD250 - from comments above, seems like the “cheap and cheerful” solution though overall it’s a little more unknown. Seems like it was designed for the RC market; I can’t find anything outside this thread about use for FRC. The lower cost and higher discharge rate are appealing though.

  3. DIY - if @Weldingrod1 is willing to share a little more design detail, I can see duplicating what he’s done. This could be a good little electrical-team project. But it’s also a “science project” compared to either of the other options that Just Work™. I do like the idea of being able to do quick, high-load tests to weed out failing batteries. Cost for parts is probably somewhere between the other two, depending on how the load is done, I think.

Has anybody else tried the BD250?

@Weldingrod1 do you have a parts list/schematic that you could post? Or details on how you run your tests with it? I’m a little unclear on whether the meter controls the test cycle or merely provides readings from when the load is connected.

Open to hearing any other advice or if there’s another option in this price range. Thanks all!

I would check your local area. Beyond a Battery Beak to ensure you avoid process failures, the best battery tester is like the best truck or boat: your friend’s that you can borrow when you want.

The money you can save may well buy you more batteries, which is the solution for many battery woes.

There is merit to this view and we’ve done that in the past, actually. We borrowed a CBA which let us make a pass through what we then had and characterize them, which is why I see the value in the tool.

The counterargument is that a tool you have at the ready in your shop is one you will use. I’m thinking of all the times that someone brought a battery in from a drive practice and claimed it “acted weird” or “might have a problem”. It would be very nice to be able to determine right away that yes, it does, and should be flagged/retired, or no, it’s fine and something else must have been going on. Batteries being as critical as they are, the one-time investment in a tester may pay off the first time it saves a match…

2 Likes

The thing that REALLY brought this to a head was having two batteries less than a year old drop to 10% capacity! My takeaway was “we MUST test our batteries before EVERY competition”. Two tests in a 2 hour meeting weren’t going to solve this. Thus, the 50 Amp nuclear tester.

Before our first competition I had tested all our batteries over a span of weeks with the bare circuit board tester. Which works very well BTW. It’s called a DL24, 180 Watt battery tester. $99 on Amazon. Electronic Load Tester 180W Upgrade Version 4-Wire Color Screen DC Tester Electronic Adjustment Constant Load Meter Lithium Battery Capacity Monitor Discharge Charging Power Meter Checker https://a.co/d/fNxuiO2

The BD250 will give you a bit more current for $85 on Amazon. Much cleaner packaging, professional, simple.

The difference: the DL24 is all singing all dancing, but a bare board. The BD250 has very simple controls: stop voltage and current. Bot tell you Amp hours. The DL24 will tell you internal resistance and Watt hours. Both can do data output to a computer. The BD250 has to be set -every- time. The DL24 requires a trip into menus to reset the totals.

The CBA has two BIG problems: it’s sloooow and it -requires- a computer. Both of the cheap options work happily stand-alone.

For a “just buy it” the BD250 is the best choice. Label it with instructions and settings!!!

A 100 Amp car battery tester with a SB50 connector on it will identify a bad cell in TEN SECONDS. Very good price to performance ratio, and you can test ALL your batteries in an hour. Rember to let it cool between tests! It WON’T tell you which is your-best- battery, but it will tell you “don’t compete with THIS one!” I think its a very good investment of $23 + SB50 connector. This tool is GREAT for “this battery was weird” questions.

I’ll sketch up my schematic. There is a flea-bay module that does all the brain work: start/stop test, limit settings, measurements. A big relay drops the load on the battery. I picked this one for easy kid operation: one button to start, auto stop, critical settings are hidden, easy reset. All the rest failed on one or more of those…

I’ll stop now :wink:

4 Likes

Thanks, this is all super helpful! The 100A tester is a great idea; can’t beat the price on that for sure, so it seems like a must-have. I suspect we might find a few surprises among all our batteries with that alone. :astonished:

The DL24 looks intriguing but I’m not sure how well a bare board will hold up long term.

Looking forward to seeing your design as well, think we’ll hold off on one of the other bigger ones until we take a look at that and see if it would be a good team build.

I’d go BD250 if I was starting again… cheaper, easier to use, and professionally packaged.

There is a thingiverse design for a DL24 case, but it’s not awesome.

So, fun story that underscores my counterpoint: at Electric City 2022, we had an instance where a battery got knocked off our robot cart and onto the floor. (Well, first my foot–but then the floor.) With no sign of leakage, we immediately put a band of blue tape around it–our signal that a battery is only for pit usage.

We didn’t buy our CBA until after the regional, and even if we did have one available it would remain unavailable to us for at least half a day (charge battery, rest battery, test on CBA, read results, rest battery, charge again–all while trying to put out the many other fires we were fighting at the event).

But since we had more than enough batteries to get through the event, we had the luxury of leaving the blue tape on until we got home to test at our leisure. (Sure enough, tested great and it returned to the fleet for Championship and SCRAP.)

A few years ago, we locked up the team trailer and prepared to depart for an offseason demo. We had packed lightly-- only the robot and the cart, a TV to play our demo reel, and a couple small items. Just before we closed the shop, we second-guessed the number of batteries we needed. “It’s going to be a long day, why don’t you grab one more battery and throw it in my truck,” I said to a student. He did so.

I made it a mile down the road before I realized that the slamming sounds I was hearing during accelerations and decelerations was a battery sliding up and down the pickup bed.

That’s trucked up.

1 Like

I’ve spent some thinking time on the 50 Amp tester, and I’ve come up with a cheaper and simpler way to do it, now that I know more about how it ends up being used.

The brain is capable of driving the relay directly and has an internal protection diode, which simplifies things. The load is a 600 Watt 12 Volt water heater element in the side of a bucket with a cable on it. That makes the heat rejection issue MUCH simpler; just change the hot water! I suggest NOT boiling the water.

I suggest building up a project box with two SB50s mounted on it. One goes to the battery and one goes to the load.

In use the bucket gets at least 1/2 gallon of water in it, then gets plugged into the tester. The battery gets plugged in, the Amp Hours gets re-set, then you push the “on” button. The relay turns on and the water heats up by about 45 degrees in 10 minutes.

Components:

About $100 plus wire and enclosure.

The circuit diagram ends up being -surprisingly- simple! MUCH simpler than the one I built…

I’d love to support a team assembling one of these and taking some pictures! And, of course, sharing some data!

I don’t think its a good substitute for the water heater element, but you -could- buy a 12V DC air heater for $37. I think these are PTC heaters, so they are likely to be kind of complex loads. There are two fans in there. You need to watch the polarity if you go with this load.

6 Likes

Liquid cooling can remove so much more heat energy than air cooling.

Some years ago, the local electric car club visited a startup developing motor controllers for converting delivery vans like those used by UPS to electric propulsion. Their prototype inverter had the transistors mounted on a cooling plate that had a pair of garden hoses leading to an above ground swimming pool outside the loading bay.

1 Like

I like your idea with the water heater element.

However:

bucket of water +
automotive voltage power source +
kids assembling &

no fuse in diagram?

Might be worth a bit of caution there.

Also, and you do mention this, hot water but nothing to stop things if the water is very hot or someone forgets the water completely.

I am going to add: I have something setup in my home using the cheapest car radiators I could buy and a circulator pump both to cool some computer projects and to add heat to my home in the winter using that waste heat. The heat exchanger to the home heating system is literally a water tight box with 2 radiators in it so the liquids never circulate into each other. My water is from a well and it’s quite high in iron content so my home’s boiler gets extra water from the house, but the heat transfer system is using distilled water with various antifreeze compounds I’ve used over time (including vodka which I did as an experiment).