Specify testing procedure for 36V lead batteries

[Roger44]

Hello

My son imports and re-sells a small mobile thing that can reach 15mph powered by a 36V lead battery driven motor. He would like to impose a testing procedure for the batteries so that they don't send him any old aged or sub-standard stuff, and has mailed me from China for help drawing up the testing procedure. Really I don't know where to begin on this one, the electronics board was much easier.

Can anybody with industial experience of buying batteries or lead battery experience help me out?

 

[Roger44]

More info'. One I've just got in front of me is marked 6-DZM-7, so it would be a standard sealed 12V 7A/h batt for motocycle start or electric bikes. The 9A/h model is also used. Obviously three are put in series.

Is it possible to change the title of the thread?

 

[MikeMl]

The ideal test is the same one we use to determine the continued airworthiness of lead-acid batteries in aircraft.

Basically:

1. Fully charge it (whole other discussion)
2. Establish a cutoff voltage during usage/discharing (another whole discussion)
3. Apply a constant-current discharge at the same average rate as scooter motor. (Needs a constant-current sink circuit)
4. Monitor the voltage as it discharges. Stop when low cutoff voltage is reached. (Needs a precision voltage comparator circuit)
5. The time it took to discharge is the figure-of-merit you are looking for.(Need a timer)

Look at the discharge vs current and discharge characteristics curves in this data sheet, for example.

 

[Roger44]

Mike, I think this test is not to difficult to put together, I'm familiar with constant currents procesures because of Ni/Cd charging, and there are IC's which can automate some of the switching, or simply a few op-amps. I could fix up a test rig so that the boys in his shops could test the batteries, on a one-out-of ten sampling procedure for example. I'll mail my son to know if we want to record a discharge graph on to hard drive (for proof). I've got a datalogger that sends over USB to PC and would do this fine.

Concerning "1. Fully charge it (whole other discussion)" I think I would first discharge it at a constant current down to an "empty, or 90% empty or something like that" voltage, then recharge at constant current for X time, even if it's not fully charged,

then 2 3 and 4.

Concerning battery technology, can we be reasonably confident that this test garantees that the battery can supply high surge currents. What I mean is, this test shows a battery can accept and deliver charge correctly, are there other defects limiting high current surges which could go undetected?

 

[MikeMl]

...
Concerning "1. Fully charge it (whole other discussion)" I think I would first discharge it at a constant current down to an "empty, or 90% empty or something like that" voltage, then recharge at constant current for X time, even if it's not fully charged,

Not necessary for lead-acid batteries, unless you want to measure the energy put-into the battery during charging.

...then 2 3 and 4.

Concerning battery technology, can we be reasonably confident that this test garantees that the battery can supply high surge currents. What I mean is, this test shows a battery can accept and deliver charge correctly, are there other defects limiting high current surges which could go undetected?

No. If you need confirmation that the battery can deliver high-current, short-duration pulses, then you will need a separate test, like a Cold-Cranking test for automotive starting batteries.

 

[Roger44]

Thanks Mike, I'll get back on this tomorrow after a phone call with my son to really know what he wants (but I don't think he knows what he wants either, rather relying on me to tell him)

 

[Roger44]

A last thought Mike. Rather than cold cranking a motor, what about connecting the fully charged battery to a "sufficient length" of 1,5 or 2,5mm standard electrical domestic wire such that the current is around 15A or so when you manually close the circuit with a 30A circuit breaker, and record the voltage going down for 10 sec or whatever is appropriate (to be dicussed later).

Would this test just by itself be enough to conclude that the batt is of good conception, and hasn't been lying on a cold shelf uncharged for the last two years? What do you think?
Problem is I know next to nothing about gel battery technology other than they exist.

 

[MikeMl]

Oh, you mean something like this: Load Tester.

 

[Roger44]

Thanks for the link. I've mailed it to my son, who anyway will be fowing this topic if he's got time... What do you think would be a suitable high current short burst dicharge rate for 7 and 9 A-h batteries knowing that we'ere not wanting to do distructive testing? Have you got a ball-park figure in mind?

 

[MikeMl]

If you read the data sheets for the PowerSonic SLA, they have a peak allowed discharge rate (short duration). I would base the test on an expected short duration peak current such as the starting current for the DC motor in the scooter.

 

[RODALCO]

On standby battery banks in substations for operating high voltage switch gear 11, 22 and 33 kV we use 110 Volts battery banks. SLA 9 monoblocks 12 Volts in series.
These are often 100 or 125 Ah and duplicated.
For 1 hour discharge testing the bank gets discharged at 60% of the Ah capacity.
e.g. a 100 Ah bank gets discharged at 60 Amps for one hour.
The total Voltage during the test of the 9 monoblocks has to remain above 80 Volts.
A discharge test set is used which keeps the discharge current constant and will plot a graph.
At 10 minute intervals Volt meter readings are taken to check that no monoblocks get excessive voltage drop.

For smaller batteries 7 or 9 Ah a one hour test is done with a 4 Amp load
The voltage needs to remain over 10.8 Volts.
Often a 12 Volt 50 Watt halogen lamp is adequate as load.

 

[Roger44]

Mike, the Power Sonic site is loaded with info', thanks for that link, and no thanks in a way because it makes you question what you thought was obvious. It's taking me time to digest it all.

What I learnt was that Power Sonic garantee their batteries at constant current values, maybe all batt suppliers do.. Now I think you'll all agree with me that fixing a constant current of a few amps is childs play, but 15A or more you run into difficulties because of heating of the current regulation system you use. So feedback loops, heat sinks, ventilation .....

So a new idea is to say "OK, you specify time vales at constant currents, but we want you to specify voltage droop with a constant resistance load". Not sure if they'll accept this approach. But it's nothing for them to do a few calculations, wheras it makes things much simpler for us.

Now what Rodalco said in the second part of his message for smaller batts "connect a 50W halogen lamp" seems to go this way. It's more or less a constant resistance load after a few seconds of heating, and just the voltage is checked that it doesn't droop too far. So simple. Rodalco, do your battery suppliers accept this simple test?

 

[MikeMl]

I have a home-brew computer-controlled battery tester that uses 1200W locomotive braking resistors as the load. This is for 14V aircraft and automotive starting batteries, where the load current is initially ~35A (440W). Since these resistors have a relatively constant resistance even when hot, the current drops off a little as the battery discharges from 12.6V to 11.0V, but it is easy to calculate the Amp*hours delivered as a function of the battery voltage because the resistance remains constant. For this reason, I would not use a lamp, whose resistance varies widely as a function of filament temperature.

btw-my computer-controlled tester logs the discharge curve, time to discharged, Ah, capacity, terminates the test when the cutoff voltage is reached, recharges the battery automatically.

 

[Roger44]

Hello everybody

This thread is a year old I know, I thought my son had dropped the 36V battery testing idea, but lo and behold he brought the subject up again last week.

I gave him two 100m rolls of 1.5mm section common household wires as a starter, if he measures 32V over this length that would mean 14A if I'm not mistaken. I guess I'll have to swat up on all the things you mentioned. i'll keep in touch.

 

[RCinFLA]

Most lead acid batteries have a date code melt branded into the case. You can find the code for most battery manufacturers on web, usually two digits for year, and a character for month. 'A' for January and so on. Sometimes 'G' and 'I' are not used. A lead batteries longevity starts from point electrolyte is put in cells. You don't want old batteries. For lead calcium alloy based batteries, like SLA and car batteries, six months of shelf life without refreshing charge means some sulfation has started. If vendor has done a top off charge every few months then they can be okay but you can not tell if they just did charge before you bought it. Sulfation of plates locks up some of the acid from electrolyte and lead of plate into hard crystals that are no longer part of battery activity so it has an effectively reduced capacity.

Simplest first test is rested open circuit voltage. Rested is no loading or charging for several hours. For a 36v battery you should see 37.5v minimum. Fully charged will be equal or greater then 38.1v Be sure you have a DVM will sufficient accuracy to make the measurement. Below 37.5v means the battery has been sitting around for a long time with no refreshing charge that means some sulfation has likely set in.

I have a little pocket DVM I calibrated for 12.5v. I would not buy any 12v lead acid battery that reads less then 12.5 vdc, preferably more like 12.6v or 12.7v.