need recommendations on large data acquisition project

Hey Everyone,

Got a proposal I am working on that is out of my league and need a little h elp. Basically I need to collect data on charge/discharge curves for lithi um ion batteries but the trick is I have to collect data on 200-300 batteri es at the same time. This project may only require data points every few s econds as the charge discharge will not be very fast so a switching system may be okay (Keithley 2750 or something similar). However, I have been ask ed to spec out a system that can do simultaneous data collection as well wh ich I think means a lot more $$$. Is there anything that can do simultaneo us data collection on 200-300 channels? More generally I think a +/- 10 V input range and 16 bit resolution should be fine.

This might open can of worms but who do you prefer: Keysight or Keithley? or is there third alternative?

Kelly

Reply to
kellysbradbury
Loading thread data ...

Some random Googling found this: There are probably plenty more similar systems.

I'm not sure how you plan to do the charge or discharge, but if you use a constant current discharge, you'll only need one measurement (voltage) per cell.

You might also want to consider how much power you're going to need to burn up when discharge testing. The typical 3.7V 1A-hr cell, when discharged at 1C will produce about 3.7 watts of heat for an hour. 300 cells * 3.7 watts = 1,100 watts that needs to be dissipated.

You also might want to talk to Cadex, which has built some rather large battery analyzers using one of their analyzer and a mux such as the aformentioned data collector:

Trivia: I've had a hell of time with crappy battery holders on 18650 size cells. I suggest some kind of contact compression system to insure a good connection to the cell:

Good luck.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

Do you have enough time to design custom electronics? One piece of advice is make a modest system, with say 32 or 64 channels, and replicate it "n" times. My Arduino MPX-16H board does 64 differential channels.

One commercial product that scales to *many* channels: Keysight's 34972A system with plug-ins and expanders.

--
 Thanks, 
    - Win
Reply to
Winfield Hill

Are the batteries connected to each other during the measurement or are they all isolated?

--
This email has not been checked by half-arsed antivirus software
Reply to
Jasen Betts

National Instruments. Not the cheapest one for sure. I've got one SCXI-based setup in storage that has 4*32 thermocouples + 4*32 voltage. I bought the setup from eBay, because of NI procing..

--
mikko
Reply to
Mikko OH2HVJ

Are you sure about 16 bit resolution?

In the time of analog instruments, 1% was considered precision, and it is 7 bits. To get anything worth 16 bits, your analog connections must be extra careful and electrically clean, taking e.g. thermal voltages into account.

--

-TV
Reply to
Tauno Voipio

Simultaneous acquisition in a cheapo digital camera amounts to 640 x 480 color pixels thirty times per second. So, there's certainly 'anything' that does simultaneous sampling. What a battery-charge-drain system needs, is not simultaneity, just timely.

Wiring to 200-300 batteries is going to be a BIG wiring harness; have you considered charging and discharging cells in series? The measurement of current can be done once, giving the correct value for all the series cells, using no hardware-per-channel at all.

Hundreds of 16-bit converters: not cheap. Hundreds of track/hold amplifiers: cheap. But, you'd have to custom-build a breakout module (for, say, 16-at-a-time sampling) after which you interrogate the held values by addressing modules sequentially. Generous sampling times (a full millisecond) and leisurely digitization (10 kHz @ 16 bits) isn't a hard problem.

Some high-end sampling systems are available, and usually can do programmed general purpose digital I/O to support breakout modules. Certainly the Keithley 27xx series would be workable.

Reply to
whit3rd

The sampling rate is so low that a single ADC could easily do it all. I'm with Win though--I'd build a simple x32 or x64 system using a tree of 74HCT4052s and a differential ADC such as an ADC161S626.

Cheers

Phil Hobbs

Reply to
pcdhobbs

Simultaneous? Now, if you were recording vibrations in an airframe, that may ne required. It makes NO SENSE for something as slow as battery measurement. But, it will increase the cost by a factor of, maybe, 100X!

16 bits? Do you really need that? For a unipolar ADC, that is 1.5 mV resolution. (If a bipolar ADC, then 3 mV, but there's no need for a bipolar range here.) I think a unipolar 12-bit ADC would be WAY more than sufficient (2.4 mV per quanta.)

And, of course, if measuring single Lithium cells, why a 10 V ADC range?

Jon

Reply to
Jon Elson

OOPs, 150 uV resolution for 16-bit unipolar ADC with10 V range.

Jon

Reply to
Jon Elson

Kelly - spec a "simultaniety" of 2 milliseconds (+-.01) & go from there.

Hul

snipped-for-privacy@gmail.com wrote:

Reply to
Hul Tytus

Hmmm, you could make some gizmo that only asserted it's voltage when triggered. How often does one need to sample a battery?

I was going to say a labjack, but you'd need some analog mux interface.

George H.

Reply to
George Herold

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.