Home-brew battery measuring station
 

With all the overinflated battery ratings floating around, it would be nice to have an impartial, third party, and repeatable test procedure performed on various cells. This pipe dream has been discussed a number of times, but never goes anywhere and just fizzles away into oblivion. Getting a company to perform this service is probably not gonna happen, so it's up to the R/C community to get the ball rolling. Unfortunately, the proper test equipment is rather expensive for most of us, so, I thought I'd come up with a home-brew solution made from fairly inexpensive parts.

Warning: No offense, but some of this will make no sense to many people, so don't feel bad. Some of the techy stuff is for electrically-minded people to contribute. :smile:

So, here is what I have come up with:

http://scriptasylum.com/forumpics/batt_tester.gif

The heart of this is a MMM ESC configured in high power brushed mode. This will be used to effectively vary the load to compensate for various pack voltages and rated capacity. In high-power brushed mode, the MMM could theoretically handle 360A of current. And the LVC can be used to prevent damage from "overzealous" testing.

The control circuit (labeled 556) is a basic timer circuit utilizing a 556 IC (dual 555's in one package) that will produce the throttle signal. Half of it will be configured in astable mode to get a 50Hz frequency. The other half of the chip is configured in monostable mode to get the 1ms to 2ms pulses to emulate a radio signal. This circuit could be (and has been) done with a single 555, but ESCs tend to be a little more finniky than servos, and this would provide better and more consistent control. The pot is a multi-turn version for fine tuning. The control circuit is powered via the MMM BEC, so it's pretty convenient.

The load (labeled 0.1 ohm) could probably be lower, but is currently sized to draw up to 74A @ 2s and 222A @ 6s (:oh:). Due to using the ESC, 2s would be the minimum value voltage-wise. To dissipate the power, many resistors would be placed in parallel, heatsinked, and fan cooled (yes, you heard right... a fan!).

Plottable measurements will be a bit tough. An eagletree with the high power module would be ideal, but the circuit above shows a voltmeter and ammeter (actually just a voltmeter across a very low value shunt) as an alternative. Obviously, using meters would be more difficult to measure pulses and true values, but it's an option.

This is something I'm seriously planning on putting together in the near future. Obviously, the MMM would be the most expensive part, as well as the batteries to test. I guess we'll just have to see. Thoughts, ideas, comments, and suggestions are welcome...

Great idea! I wonder if this could be the basis for a RC-M battery vault?

In regards to varying the current level - couldn't you just use a lower resistance or add/remove resistors as needed, and then vary the throttle input to achieve a desired current draw? Or would the MMM not like the switching losses and thus you have chosen for it to operate in a "full throttle" orientation?

It's definitely a very simple setup that can be extremely effective. I think other Eagle Tree recorder products use a shunt to monitor/record amp draw, so if they offer one with a high enough current rating maybe that would be the way to go?

A guy name access from rcu did some great testing - you might be able to get some ideas from him - http://www.rcuniverse.com/forum/m_8310589/tm.htm

Looks like he just hooked up some resistors and got some info on a screen. He did have a fan when he used four resistors. I want to see this testing go down.

A sort of battery vault is the ultimate goal. It would also be important that all tests be consistent and repeatable. And then, there is the question of credibility; will people trust test data from a possibly biased forum member?

Yeah, I suppose you could have banks of resistors to get in the ballpark of the current you want to test. I was envisioning several such banks all wired in parallel with CPU heatsinks and fans on them. I want to eliminate as much contact/wire resistance as possible, because at resistances less than 1 ohm, any little bit can make a big difference! Don't forget, I am talking about measuring 100's of amps here...

Not pictured, but I was also thinking of adding caps to the ouput. Since resistors are a resistive load (duh), the caps would smooth out the PWM pulses to provide a more consistent draw. Priobably use a spare car audio cap for that, but would be limited to ~5s because of the voltage rating.

Eagletree would be the way to go, but I was shooting for cheap and effective. The teade-off being lack of sampling/plottable measurements.

I see where you are going Brian. I do not mind sending some 2s packs made to your spec of the various cells I have laying around. As long as you either send them back, or test them to failure (would be nice to see the top end of what they can do...) I would not mind providing what I can. Some method of recording temp at the cells would be spiffy too, as some cells just seem to run hotter than others, at least in my "paltry" experience.

Thanks for the offer linc, I may take you up on that! I have to put this project together first. And actually, I was hoping for test subjects from Mike, MA, etc as well. That way, the cells are new and from a known source. I would send them back of course... providing they survive

Hopefully, there is enough feedback so I can improve the initial design before I start building it. Finding the resistors is gonna be a challenge though I think. 6s @ 200A is almost 4500w! Even if I limit the test duration, that's a LOT of heat. Maybe I'll save testing for next winter so this can also be used to heat the house. :lol:

Even if an eagletree had to be used, I would love to see REAL graphs for all the packs out there. That might quiet the maxamps debate, quiet the hobbycity actavists, and help hundreds of people make the right battery choices. If such a battery info bank was compiled, I think it would be a better online tool then the speed calculator.

lol, I just realized I made a mistake in my original diagram. IIRC, the three motor leads are tied to one end of the load, but the other end of the load is tied to the battery + lead. Oops. No biggie, the idea is the same.

Metallover: Yeah, I'm hoping it is useful. I don't want to spend the time and money on something no one will pay attention to...

LOL

I posted data a couple years ago on my maxamps packs... showing much better numbers than what was being talked about here. The reaction was that I must have something wrong with my eagletree or something because they wanted to believe the crappy numbers someone else got with their packs.

Trust me... if MA packs get good #'s then people here will take issue with it, and continue to march with their pitchforks and torches.

I have never, and still don't... make exaggerated claims about maxamps packs. When I posted data in the past, it showed they were not even the best packs I had at the time. My 20c kokam had the best #'s. I've also posted in the past that my Core 5000 pack had better #'s than a similar MA pack in the same test.

wow that'll be cool!

Hopefully any data I release will be accepted more readily as fair and unbiased. That's one of my fears; that people will think I'm biased because I am here...

+1

btw, feel free to delete my posts here, to keep the thread clean if you like... I won't flame you LMAO

How can you be biased when posting raw info? As long as it's all correct Info I can't see anyone thinking you're biased.

Being a "senior" member of this forum, people may claim the test results are skewed in favor of whatever Mike carries.

I do remember seeing your numbers. From what I can remember they were similar to the results I had with my maxamps packs.

Performance was not really my issue with the dozen or so packs I had. It was longevity. I had them puff just sitting, and had to weed out the bad cells to make good packs. Once I had done that the packs were decent, and I even sold them. Some are still running today, so my "backyard matching process" (trial by fire so to speak) apparently worked.

The packs I had, 3000, 4000, 5000 and 10000 cells, just did not have enough headroom to handle the setups I had going at the time. To be fair I was trying to push a MT on a 4s setup, and we all know that 4s in a heavy vehicle is hard on lipos. The fact that my trakpower packs outperformed the ma 5k packs and were cheaper kinda sealed the deal for me. Then there was the Jason incident, and that opened my eyes to the fact that money overcomes truth for some people...

Please post a link to your numbers GD, I would like to see them again, now that I have a better appreciation for lipos.