Home-brew battery measuring station

[lincpimp]

Originally Posted by BrianG

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

Ah, but "knowing" you will indicate that the previous statement is not the case.

In any case there will be people who discredit you due to the type of equipment you use, or the color of underwear you have on on Thursday. I am looking forward to seeing the results, and would like you to test some used cells just for my own interests. I just like to see what happens after hard use.

[himalaya]

Great thinking Brian.

I am worrying about the MMM's input decoupling capacitors, you are pushing them really hard by this app.

IMO a programable(or just use a potentiometer to adjust) linear current source would be just fine for this application. just get a few power MOSFETs mounted on a HUGE fan-cooled heatsink, drive them with an Opamp connected as a current source, that's it.

You know what a stable and constant current means for a battery test platform, right?

[hoober]

I've offered to test batteries on a few forums before up to 500 watts. I have one cc-400 and 1 cba that is paid for already. It produces nice repeatable graphs , but no-one seemed interested enough to send me any packs.

I know you like to invent/engineer/build stuff , but my point/question is how much are you looking to spend on labor/equipment? This set-up seems available for about $1 per watt, maybe $500 total and can add multiple cc-400's at about the same $ per watt.

I figured it this way: I had already paid for the test equipment and didn't want to buy batteries. I thought perhaps users may send battery samples, but shipping twice really comes close to some of the battery packs I'd be testing. So then I'm left with simply buying the packs myself (packs I don't need)

We've talked about standards too, I can't remember which forum. I kinda like testing single cell for direct comparison , but doing "packs" seemed to be the choice of many users even though it requires 2,3,4,5,6 times the power.

Looking forward to the results here for sure. Someone send me a "test" pack, so I can pull the equipment out of the box ,lol. I don't think there will be any credibility issues with your testing/graphs/data/info Brian.

[Arct1k]

http://www.powerohm.com/catalog/GRCAT00.pdf

RH350?

[glassdoctor]

That does bring up the issue of testing packs with wiring and connectors vs. bare cells with direct hookup. Those two methods will need to be clearly noted.

Linc... will have to search for the old posts. Is there still access to the old forum server?

Originally Posted by BrianG View Post
Being a "senior" member of this forum, people may claim the test results are skewed in favor of whatever Mike carries.

Besides "knowing" BG here on the forum, I know him in person also. I can back him up 100% for what that's worth. Trusting his test results shouldn't be an issue.... but there's always a few, ya know?

[BrianG]

Originally Posted by himalaya

Great thinking Brian.

I am worrying about the MMM's input decoupling capacitors, you are pushing them really hard by this app.

IMO a programable(or just use a potentiometer to adjust) linear current source would be just fine for this application. just get a few power MOSFETs mounted on a HUGE fan-cooled heatsink, drive them with an Opamp connected as a current source, that's it.

You know what a stable and constant current means for a battery test platform, right?

Yeah, I was thinking about the caps too, and was going to add several more (probably another 2000uF or more). Being a resistive load though won't have that constant ripple which heats them up, at least not as much as a motor's inductive load. I do still have to take the MMM's PWM switching into consideration though.

From some research on using FETs as a pass transistor for a high-A power supply, it takes a special type of FET (read: expensive) to operate well in the linear region (as opposed to switching). FETs would be easier to parallel though. So, that leaves paralleling a bunch of transistors - ugg. I'd need a "few" to say the least!

Originally Posted by hoober

I've offered to test batteries on a few forums before up to 500 watts. I have one cc-400 and 1 cba that is paid for already. It produces nice repeatable graphs , but no-one seemed interested enough to send me any packs.

I know you like to invent/engineer/build stuff , but my point/question is how much are you looking to spend on labor/equipment? This set-up seems available for about $1 per watt, maybe $500 total and can add multiple cc-400's at about the same $ per watt.

I figured it this way: I had already paid for the test equipment and didn't want to buy batteries. I thought perhaps users may send battery samples, but shipping twice really comes close to some of the battery packs I'd be testing. So then I'm left with simply buying the packs myself (packs I don't need)

We've talked about standards too, I can't remember which forum. I kinda like testing single cell for direct comparison , but doing "packs" seemed to be the choice of many users even though it requires 2,3,4,5,6 times the power.

Looking forward to the results here for sure. Someone send me a "test" pack, so I can pull the equipment out of the box ,lol. I don't think there will be any credibility issues with your testing/graphs/data/info Brian.

That type of device would be the ideal platform since that is what it was designed for. My idea was just based off things I have or can get fairly cheaply. If testing single cells, 1 or 2 of those would be sufficient power-wise.

I too am concerned about getting enough interest to justify the cost/bother of assembling this (although I could use it for my own uses too ). What we really need is to set up a company that does this "right": manufacturers send in sample packs, pay a certain fee, and we provide them a "birth certificate" (similar to some car audio amplifiers) outlining its performance over a variety of tests that they can use in their advertising. Voila! A third party unbiased testing result!

Originally Posted by Arct1k

http://www.powerohm.com/catalog/GRCAT00.pdf

RH350?

I just sent in a quote request for the 350, but I would speculate that just the resistor is well over what I want to spend on this project.

For all the parts and components, I can see about $500 if I can find good deals. I have some stuff already. I have a watch on an eBay auction of 240 Dale 50w power resistors (3 ohm).

Thanks guys! Keep up the feedback. Maybe there is a better and cost-effective way to get us the data we all want (but few realize they want it).

[BrianG]

Originally Posted by glassdoctor

That does bring up the issue of testing packs with wiring and connectors vs. bare cells with direct hookup. Those two methods will need to be clearly noted.

Direct cell hookup would be the best test for the cell performance, but since we use these in packs in the "real world", I think the testing should be done that way too. Heck, even a very good cell can appear crappy if using poor assembly techniques and substandard wiring. Testing a pack would account for all of that. Of course, discrete cell results will be better than pack results, but testing packs would give people expected results. If they get cell tests, then take eagletree graphs of the pack, they may note the discrepancy and not trust the tests. Once credibilit is destroyed, then all that work/money is down the drain.

Originally Posted by glassdoctor

Besides "knowing" BG here on the forum, I know him in person also. I can back him up 100% for what that's worth. Trusting his test results shouldn't be an issue.... but there's always a few, ya know?

Awww, shucks!

[hoober]

How about doing "pre-made packs" since that's the way most are sold now. Not many users buy cells and make their own anymore. That way the lead and the cell construction would be in the picture so to speak. If I was to test packs , I think I'd like to limit to 2S or at most 3S and use bare leads only. That way the connector of choice would not influence the testing and it would be easier/cheaper to test them. It would still be a good enough sample to interpolate data for 3-8S packs without requiring very high power capacity.

[BrianG]

Originally Posted by Arct1k

http://www.powerohm.com/catalog/GRCAT00.pdf

RH350?

Well, I got the reply back from the Iowa distributer (pretty quick too) and the cost is....

drumroll please.....

$749.00 (plus freight)



Yeah, I knew it was gonna be pricey, but man! For a resistor? Granted, it is an 11,000w resistor, and nothing that beefy is gonna be cheap. Wow. Looks like I'll be making a resistor bank!

Originally Posted by hoober

How about doing "pre-made packs" since that's the way most are sold now. Not many users buy cells and make their own anymore. That way the lead and the cell construction would be in the picture so to speak. If I was to test packs , I think I'd like to limit to 2S or at most 3S and use bare leads only. That way the connector of choice would not influence the testing and it would be easier/cheaper to test them. It would still be a good enough sample to interpolate data for 3-8S packs without requiring very high power capacity.

That's kinda what I was thinking anyway. I just figured the ESC was good up to 6s, so why not? At least the option for higher power will be available. Still, to get say, 200A, out of 2s, I would need 0.037 ohms. To get that, I'll need a bank of 81 resistors at 3 ohms each. I'm gonna need to build an addition to the house.

I agree; the bare leads idea is probably the best. I'll probably make a bus bar setup where I solder the bare leads to the bars for the least amount of contact resistance. No alligator leads for this thing!

[Arct1k]

before going any further i'd confirm with pdc that nothing funky in esc elecs that would stop it driving a resistor vs a motor...

an oil bath maybe easier than fans...

[Metallover]

You could find a way to hook up an OLD space heater to act as a resistor... I have an old one that's so inefficient.. It's 1500 watts, tripped a breaker a couple times.

[BrianG]

Originally Posted by Arct1k

before going any further i'd confirm with pdc that nothing funky in esc elecs that would stop it driving a resistor vs a motor...

an oil bath maybe easier than fans...

If anything, it would work better. No inductance, no phase angles, no "current leading voltage", etc. Purely resistive loads are always easier to drive. Which is why some devices use caps to cancel out inductance for power factor correction to make the load appear more resistive. I suppose it wouldn't hurt to ask him though, if he answers his PMs...

Originally Posted by Metallover

You could find a way to hook up an OLD space heater to act as a resistor... I have an old one that's so inefficient.. It's 1500 watts, tripped a breaker a couple times.

That would work if I was testing 120v batteries. That heater's element is around 10ohms. I would need at least 3s just to get 1A out of it. Thanks for the suggestion though!

[Arct1k]

That was my concern though - I'd bet castle uses some of that feedback to tweak operation of the motor...

[Metallover]

Originally Posted by BrianG

That would work if I was testing 120v batteries. That heater's element is around 10ohms. I would need at least 3s just to get 1A out of it. Thanks for the suggestion though!

I was thinking it would be possible to make a "reverse power supply" fairly easily. Convert the voltage from the battery to 110v for the space heater. The space heater may even have a power supply to step down the voltage a little?

Or, instead of a resistor, you could hook itup to a bigger motor driving a fan or something....

[Arct1k]

Can you put the power back into the grid... Be cool to get your meter going backwards....

Charger the batteries at work - discharge at home - tee hee