Invenfion of fhe day: The AAA AA Aligator fnip bite

[RoGeorge]

Tadaaam!!!  Proudly presenting ... [drumroll] ...  The AAA AA A ligator fnip bite! 



Wood block + half-snipped alligator + sideways wood screw fix = battery test socket

[magic]

Hmm, what's the contact resistance?

(Load with 1A, measure with DMM mV between battery and the gator.)

[RoGeorge]

The contact resistance was pretty bad.  Those are alligators bought from LIDL, cheap and not particularly good.  A few were even heaving small spots of rust on them.  Now, that you made me reach the box with alligators, I realize they are sold to be used for max 1A and max 24V.



Wiping the contact points with isopropyl alcohol makes the contact resistance reasonably OK.  Measured the contact resistance already, but don't recall the value.  Will measure again tomorrow.

Anyway, the intent was to make DYI battery Kelvin probes, by soldering a small piece of double sided PCB at the tip of the alligator, with a gold plated pogo-pin on each side of the PCB.  Just that instead of building yet another ADC to properly measure the voltage, I've just used the I and V measured by the power supply itself (logged by SCPI) and didn't bother adding the pogo pins anymore.

The power supply (a 3 channels Rigol DP832) has only internal compensation for V dropped on wires, but no external connectors for compensation wires.  So if I were to use a Kelvin connection, will have to either modify the power supply, or build an external voltage logger with 3 channels.

[RoGeorge]

Measured mV at 1A
  - at the positive terminal, between 3.5 and 7mV mostly, would say less than 10 milliohms.
  - at the negative terminal however, it is much bigger, about 50 milliohm

Varies a lot from one battery to another, or between reinserting the same battery, or by wiggling/rotating an already inserted battery.  The alligator's moving jaw alone has about 1.1mV drop along its length.

I would say less than 100 m\$\Omega\$ contact resistance per battery.

[timeandfrequency]

I would say less than 100 m contact resistance per battery.

To be compared to the internal resistance of a fresh AA primary cell : about 70 to 120 m .
The gators will have a not so negligible influence on some of the measured values.

[temperance]

Can I ask you why you are doing this? PCB mount battery holders for AA and AAA batteries cost nearly nothing. I have some leftover AA battery holders if you like. Somehow there must be an other reason why you went into this adventure.

[RoGeorge]

Casual battery holders are dirt cheap, but not very good.  Tried a few no name models and didn't like them.  I'm talking about holders like these:



Kelvin battery holders are not cheap.  Ideally would be to have about 10 testing channels, so 10 of them.  The tester is a one off, for my own use, which made me try different ideas of DIY battery holders.  This one improvised from alligators is just one of the many.  It was meant to be just a quick try.  Turned out it was good enough for about a week of measuring the mAh of about 50-100 recheargeables.

As usual, any temporary setup is meant to become permanent.  And each time I was connecting another battery, the snipped jaw of the alligator was reminding me of the picture of that dumb dinosaur, which was very funny. 

But the real reason, I guess, is that I like more to tinker and improvise, than to order and wait.

[temperance]

Now I'm curious about what you're trying to measure because AAA batteries and high currents don't go well together.

Edit: when I needed kelvin sensing on those low cost battery holder I just soldered a wire onto the spring itself.

[RoGeorge]

Same idea of alligator battery socket can be used for AA too, or for any other round models:  C, D for 18650, etc, wich might be of higher currents.  Kelvin connectors would be nice to have more for precision than for max current, to eliminate the contact resistance.

Things I want to measure would be the mAh, and the charge/discharge curves (already did), other nice to know would be the internal resistance and the max I short-circuit current, things like these.

[magic]

Truth be told, unless you plan to solder your batteries, ignoring contact resistance will not give you realistic results

10mΩ is not too bad at low currents. Maybe you should use those clips on the negative too.

[RoGeorge]

Both ends of the batteries are connected the same, with a sniped alligator each, just that the minus of the battery is usually more covered with dirt, and is more oxidized, that's why the big difference in measured contact resistance.

Overall it was not that bad, typical measure for C is made at a discharging current of 0.2C, so at less than 0.5A.  This means probably no more than, say, 100mV voltage drop including other wires and connectors.

100mV at a nominal 1.2V would be less than +/-5% errors.

Room temperature, battery temperature (air circulation speed), the charging method (fast/slow), or the charge termination algorithm, each of these may introduce variations of more more than 5%, I guess.  Anyway, the goal was to sort out which batteries worth keeping from a pile of NiMH and NiCd accumulated in the last 20 years or so.

Some were not able to sustain at all a current of 0.2C, others were still usable just that they were not good with high demanding loads (e.g. not good for the laser pointer).  In terms of C, even the 10-20 years old NiMH were still able to store about 50-70% of their nominal mAh.

But yes, the initial plan was to make battery holders with Kelvin connectors at each side, then to use a couple of ADS1115 I already have (16 bits ADC w 4 channels and programmable gain input) to build the ultimate charger/discharge ever made!   

Except all that would have been unrealistic for a weekend build, so I've end up measuring with this instead:


 

- the instrument in the background is the Rigol DP832 triple power supply, measures the I and V on each of its 3 channels.  It also regulates the discharging current, and turns off each channel when the respective discharging battery drops under the minimum allowed voltage (I've used 0.5V as discharging stop voltage)

- the PCB with radiator is from a former ATX power supply, now turned into three "power Zener" like, of 5V and max 3A (with a TL431 voltage regulator and a BD680 darlington each).  The 5V power Zeners in series with each channel turn the power supply into an active load, to discharge the batteries (DP832 can only source, can not sink current).  The radiator for those 3 power Zeners needs a fan, to dissipate the difference between 5V and 1.2V (the fan on the right of the plastic tray is a 12V ATX fan, under powered with only 5V to keep it quiet)

- then, all the wires and the "fnip bite" battery holder you already seen

- and all the I and V, sampled at each 10 seconds from the three batteries, are logged from the DP832 in a CSV file (over LAN)

- then, from the CSV log file, a GNUplot script displays live the chart of I and V, and the mAh discharged so far, while each battery is discharging under its constant current set, charts like these:


This chart is for 3 of AA Tronic 2300mAh batteries in use for about 10 years so far.  Still usable, but not in a very good shape.



This is how some 5+ years of usage Varta 2100mA discharging chart looks like, much better.



And this is for brand new Tronic 2500mAh batteries manufactured this summer, at their first discharge.

[temperance]

Interesting, you even recycled the ATX power supply board.

[rteodor]

That adhoc terminal block is also something to see.
Too bad there aren't such connectors with 2,54mm pitch at the general store. Pheonix and Molex are robbing me.

[voltsandjolts]

Can you please share your gnuplot script?

[RoGeorge]

you even recycled the ATX power supply board.

All I wanted from that ATX PCB was its TL431 regulators (that board was heaving 3 of them!), and most of all, its radiator.  But to extract the radiator I had to unscrew the original transistors/diodes from it.  And to make room for the screwdriver, I had to desolder most of the other big components around the radiator,  and the 2 transformers, which were the most difficult to remove.

Once the big parts were removed, it became clear that would be much faster to just strip away any other components from that PCB, than to re-anchor the radiator on another protoboard. 

After all, to make a 5V power Zener, all was needed was a TL431 with 3 resistors and a power transistor.  It was a single side PCB, so it was easy to find enough empty islands to build 3 isolated power Zeners, one for each DP832 channel.  Schematic is from the TL431 datasheet, except I've used a Darlington BD680 instead of a normal PNP, because it happened to already have a few BD680 in stock.

For a 5V drop between CE, the R1, R2 are of 2k7 each, and the R between BE is 220\$\Omega\$

[RoGeorge]

Can you please share your gnuplot script?

Added inside the attached zip from this other topic:
https://www.eevblog.com/forum/projects/cc-cv-power-supply-as-a-contolled-load-for-batteries-(i-e-rigol-dp832)/msg5744693/#msg5744693

At the end of that post there is an 'Example' section for how to use the unziped files.