audio-talk

and now for something completely different.....

The scenario

I was going to rig up a device to measure the internal resistance of a lipo battery(max 4 cells 18v), using an atmel 8 bit which requires a divider to take the voltage down to the max adc level.
I was going to use the circuit shown in order to use the function:

Ri = (Vopen-Vload)/I

The question: does the panel think that the divider will affect the on load vs off load voltage for the calcs.

If anybody is curious why the atmel and not just use dmm then the thinking is that I can lend out the device and the whole thing relies on timing between the readings...and obviously a disconnect after the prescribed time.

12k vrs 2R is far enough apart I would have thought. I make it a 0.033% difference (though its constant in the on and off load cases so maybe even less important). The max resolution of tha 12bit ADC is 0.024% and in practise will be a bit less. You could include it in the calculations anyway if you wanted.

Not sure why time is important for internal resistance though.

thanks for that it confirms my original thinking.

accepted wisdom relating to lipos of the 2-10 Ah type seems to suggest that off load should be measured greater than 2 mins after on-load. On load should not be measure before 20 seconds after connection.

The crucial thing seems to be off load must be after on load and off load must be after lipo has had time to recover. I would like to test at something like 20-30 amps but that entails heat sinks etc, and a light bulb wont cut it at the varying voltages. I could just settle for measuring the individual cells but the full picture will tella tale with the whole pack.

there seem to be many meters on the market which don't adhere to timing and they're stupidly expensive at the same time. Can't speak for their accuracy but I thought I'd get rid of one of the gazillion Atmels in the box.....

edit:
I've just had second thoughts.....I could just measure the individual cells and add up all the internal resistances to approx the whole pack. Measuring individual cells does limit the amount of current I can draw because the balance leads are not very robust....should be good for 6 amps of so though.
measuring individual cells does mean I don't need the divider and would negate my question and the whole thread.....mmmmm

Oh, I see, that explains the time aspect, thanks.