CBA III, discharge voltage questions

[jasonck08]

So I've had my CBA III for a few weeks now, and I've noticed something strange that happens both on the CBA III and on my RC charger Turnigy Accucell-6.

When my CBA III measures the open voltage of a battery (18650 LiCo cell for example) it matches the reading on my multimeter give or take .01v.

When the CBA III is measuring the voltage during discharge the voltage that the CBA III is measuring is usually between .1 and .2 lower than what my multimeter reports.

For example:

2600mAH Sanyo 18650 Fully charged and begin discharging at 1A. After a minute of discharging the CBA will report 4.00v battery voltage, but the mutlimeter will report 4.10v.

If the current is increased, the voltage reported by the CBA III is off by a larger margin - about .2v when the current is in the 3-6A range.

.2v is considerably off for a voltage reading. Both my CBA and my much cheaper Turnigy RC charger both are off by about the same amount.

What's up with that? I'm fairly cretin that my mutlimeter is correct and the other units are off because the open voltage with no current load on all 3 devices match.

Also I've been testing protection circuits with manufacuters specs that the IC is set to 2.8v cutoff. The cutoff only happens at a reported 2.6v on the CBA, but its about 2.8v according to the multimeter, which would be accurate according to the company that made the IC for the protection circuit.

Any clues? Anyone deal with the same issue on there chargers, dischargers, or battery analyzers?

 

[VidPro]

clue , the wires ?

Voltage___________________________________________voltage
Here______________________________________________Here
. . . . . . . . . ^ wire is tiny ammounts of resistance

anytime there is a flow of current from or to something, and any ammount of resistance, the voltage at the other end of the resistance will be slightly different depending on the ammount of resistance?

Big 3.6v___________________________________________battery
Load______________________________________________3.8v
. . . . . . . . ^ tiny voltage drop through wire


Big 4.1v___________________________________________battery
Charge____________________________________________3.9v
. . . . . . . . ^ tiny voltage drop through wire




Then the Actual Battery voltage
|--{====]---|
|---load-----|

when a load is applied to a battery a reading now is inclusive of both the Load and the battery
IF the device or microcontroller Stops for a few miliseconds to actually Check the battery voltage, then the actual battery voltage is not the same as the Measured voltage while the load is applied.

and Visa Versa
|--{====]---|
|---Charge---|

when a charge is applied to a battery a reading now is inclusive of both the Charge and the battery
IF the device or microcontroller Stops the charge for a Few miliseconds to actually Check the battery voltage, then the actual battery voltage is not the same as the Measured voltage while the charge is applied.

now you add the wire resistance and the load or charge, and then you know why it is nice to tell the methodology of the testing when displaying data.

 

[jasonck08]

I did think about the resistance, but I didn't think that would effect it this much.

I have about two 18" 18 gauge copper strand wire going to 2 brass 10 gauge connectors which make contact with the battery. Do you guys think that this type of a setup is enough to see that kind of voltage drop?

There is always going to be resistance in a charging or discharging type setup. And no discharger probably has the ability to scale the voltage at different discharge currents given a set resistance.

Can anyone help me calculate the resistance in 18" of 18AWG copper stranded wire? What would be an expected voltage drop at 5A.

EDIT: Found a resistance calculator, and wow looks like 18AWG copper wire has a lot more resistance than I thought.

Looks like I'll be making some changes to my rig and will try to shorten the wires to about 6" and will use higher gauge, and I may need to go with copper contacts. Ideally I would like to see no more then a .005 voltage drop at up to 5A or so.

 

[VidPro]

its like ZERO resistance, i have looked this stuff up on the charts , good copper wire with . . . . 1000feet 1ohm.
all useless info, because the voltage drop and probably the quality of the copper or silver or whatever varies. and we already know it varies depending on the amps going through.

try and take a mili-voltage reading across the wire itself. while the load is applied, because an Ohms reading wont show much. the voltage reading still wont be much.

i have had to use 4guage and 1 guage real copper wire on some things , and still get huge drops, but that was at some 40amps with distance. went crasy trying to figure out how to eliminate it without "Starter motor" wires
at some point you just get some good copper and roll with some slight variances.

the Math doesnt seem to show it as well as reality, that is what i am trying to say. and them all copper COLORED wires

another example, a 100' Cheap AC cable 115VAC RMS power at one end, 500W load, 105V AC RMS at the other end, copper wire inside, 200 foot round trip. chart says cheap "copper" extention cord is alloys

 

[Russel]

If you have the Pro version of the CBA software there is a dialog box to calibrate the voltage drop. You enter the current for the test and the voltage at the battery and the CBA under the same current load and the software will account for the voltage drop of your test leads.

 

[Meterman]

The method Russel has shown how to eliminate the difference is the only way to more exact measurements but very expensive and really should be integrated in the normal software!

To only find out the present voltage drop in your rig simply set up a discharge of your desired current and compare the voltage shown in the upper right corner of the CBA window to the voltage measured at the battery under test. Then click at the "Pause" button to have no current flowing and thus no voltage drop in the rig and again compare the values as before.

Best way to check this is to use a bench power supply instead of a battery because of the stable voltage with or w/o any load.

Wulf

 

[jasonck08]

I don't have the pro version because its so much more expensive just for a few extra features. I'll work on reducing the resistance of my charging leads and clamp.

 

[jasonck08]

Wires are now like 5" long and using like 10 guage copper wire. Seems to have a lot less voltage drop. It's only usually .002 or .003 off from what my multi-meter reads @ the battery.

 

[Meterman]

Hi jasonck08,

wow - these wires seem to be more than adequate now!

But what about all the contacts? They can produce substantial resistance.

My answer in post #6 seems so self-evident, that you might ignore it, but my CBA II is all but a precise instrument - and others have confirmed that with respect to their own CBA II.

After I found out that the voltages shown by the CBA are too high, I made some series of measurements including this latest setup:

I have soldered gold-plated sockets in parallel to the Anderson Powerpole Connectors mounted to the CBA to attach leads to a Tektronix DMM4040 (equal to Fluke 8845A) and so I can measure the true voltage at the input terminals of the CBA totally independent of any current.

The results do not designate the CBA II to be a precision measuring instrument in my opinion - for example precise 4.000xyV at the input are shown to be 4.10V by the CBA. That's 2.5% off, but this inaccuracy is not even consistent over the voltage range. From 10V up the voltage shown is not even constant (although the input is very stable), the values are jumping a bit.

You've got the CBA III, of which I nowhere could find any test - only repeatedly the self-adulation of the manufacturer. I'm interested in the exact setup that led you to the figures shown in your post, to get an impression if the CBA is improved now.

Wulf