Question for the EE's here..?

[Maine Sail]

Hi Guys,

I have been using a Victron BMV-602 and a car headlight bulb to test the capacity of my batteries but as the voltage drops the amp draw changes slightly. Battery is 125Ah and the headlight 75W 12VDC. It is usually right around 6.25A - 6.29A at 12.7V, but, I went through four or five bulbs to get that...

I tried a DC dimmer, PWM type, and an AC dimmer through an inverter but neither of them will hold as steady as I would like. I want more accuracey than +/- 5-10%...

Is there a way to build a device, or maybe it exists, that would apply a load to a battery and keep the load steady despite the dropping voltage? i'm sure there is an easy way and I am just overlooking it.

I am trying to come up with a more accurate way to test the Ah capacity than what I have now and there are no commercially available products that I can find that will do this test..

The BMV-600 series has a built in relay that I set to trigger at 10.5 volts. This then goes through a SPDT relay and disconnects the load from the battery. The Ah's removed, before the battery hit 10.5V, is captured, on the battery monitor and approximates the actual condition of the battery based on the 20 hour Ah capacity test..

This is all well and good but not as accurate as I would like due to the changes in voltage.

Thoughts.....??????? I am sure one of you guys will have the answer for my fixed load device...... I want it adjustable so I can test more than just one size battery. An adjustable load of a steady 2-15A, despite the voltage dropping from approx 12.7V+ to 10.5V would be great!!

Probably not cheap if doable at all...?

[cal40john]

Originally Posted by Maine Sail

Hi Guys,

I have been using a Victron BMV-602 and a car headlight bulb to test the capacity of my batteries but as the voltage drops the amp draw changes slightly. Battery is 125Ah and the headlight 75W 12VDC. It is usually right around 6.25A - 6.29A at 12.7V, but, I went through four or five bulbs to get that...

I tried a DC dimmer, PWM type, and an AC dimmer through an inverter but neither of them will hold as steady as I would like. I want more accuracey than +/- 5-10%...

Is there a way to build a device, or maybe it exists, that would apply a load to a battery and keep the load steady despite the dropping voltage? i'm sure there is an easy way and I am just overlooking it.

I am trying to come up with a more accurate way to test the Ah capacity than what I have now and there are no commercially available products that I can find that will do this test..

The BMV-600 series has a built in relay that I set to trigger at 10.5 volts. This then goes through a SPDT relay and disconnects the load from the battery. The Ah's removed, before the battery hit 10.5V, is captured, on the battery monitor and approximates the actual condition of the battery based on the 20 hour Ah capacity test..

This is all well and good but not as accurate as I would like due to the changes in voltage.

Thoughts.....??????? I am sure one of you guys will have the answer for my fixed load device...... I want it adjustable so I can test more than just one size battery. An adjustable load of a steady 2-15A, despite the voltage dropping from approx 12.7V+ to 10.5V would be great!!

Probably not cheap if doable at all...?

google constant current load

200 watt DIY, so 13 amps at 12 volts but he plans for 4 that can be paralleled.
200Watt Active Load Unit

also saw commerially made ones, no idea how much they cost

Also saw simpler ones, but lower current capability, probably could replace the power FET or transistor with one that can handle more current and add a bigger heat sink.

John

[btrayfors]

Maine,

Great minds :-)

I have the same problem when testing capacity, but have a cheap solution which has worked well for me over several years.

I built a breadboard load device consisting of six sockets, an ammeter, and a spdt switch. I can screw in any desired load using 12VDC bulbs.

I use a Fluke 189 to measure battery voltage, and a WattsUp device to keep track of energy consumed over time. I find the amperage figures fully consistent with independent measurements using a Fluke 337 clamp-on ammeter, thus the reported AH, KWH, etc. are pretty accurate. The voltage figures on the WattsUp differ somewhat from the terminal voltage measured separately, but that's not a problem.

While the load does vary somewhat as voltage drops during the load tests, it's not all that important since the WattsUp keeps track of AH, etc., and that's the key factor being tracked.

Here are the results of some tests I did earlier in the week on a couple of old T-105s that I've been trying to rejeuvenate, just for the fun of it!

FWIW.

Cheers,

Bill

[Maine Sail]

Originally Posted by btrayfors

Maine,

Great minds :-)

I have the same problem when testing capacity, but have a cheap solution which has worked well for me over several years.

I built a breadboard load device consisting of six sockets, an ammeter, and a spdt switch. I can screw in any desired load using 12VDC bulbs.

I use a Fluke 189 to measure battery voltage, and a WattsUp device to keep track of energy consumed over time. I find the amperage figures fully consistent with independent measurements using a Fluke 337 clamp-on ammeter, thus the reported AH, KWH, etc. are pretty accurate. The voltage figures on the WattsUp differ somewhat from the terminal voltage measured separately, but that's not a problem.

While the load does vary somewhat as voltage drops during the load tests, it's not all that important since the WattsUp keeps track of AH, etc., and that's the key factor being tracked.

Here are the results of some tests I did earlier in the week on a couple of old T-105s that I've been trying to rejeuvenate, just for the fun of it!

FWIW.

Cheers,

Bill

Attachment 35022

Attachment 35023

Bill,

I've done the light bulb thing but want more accuracy than I can attain doing that and I don't want to baby sit it for 20 hours turning stuff on and off to maintain the load.. I want to be able to do this for my customers with expensive banks to keep track of performance year to year.

For an accurate 20 hour Ah capacity test I want my load to be exactly what the manufacturer has rated it for.

Eg:100Ah battery / 20 hours = 5A load until the battery hits 10.5V
or 125Ah battery / 20 hours = 6.25A load until the battery hits 10.5V

In both of the above batteries, when new and broken it, they should last 20 hours before hitting 10.5V at 77-80F..

The Victron monitor gives you 90% of what you need and a positive cut off that can be set for 10.5v to stop the test. I also keep my various meters connected during discharge to monitor all parameters.

Any load above or below the 20 hour rated load will give pretty close results but will be skewed by that guy named Peukert, somewhat.. I am also testing them at 77F so I have the right temp. Hotter than that and you gain some capacity in Ah and colder you lose some..

I know this is picky but I feel it should be able to be done... I really can't believe no one makes a tester that will do this..? I've been researching for what I want and just can't find it..

[cal40john]

I can find one, I don't think it met the cheap part though.

Storage Battery System GL-1000: Constant Current Electronic DC Load Tester

This is the simple DIY one I found. It's designed for 1 amp, but if you replace the 1 ohm resistor with 0.1 ohm it would do 10 times the current. Problem is although the FET is rated for 12 amps its max watt rating is 50 watts. You need to find one that can handle more power or build 2 or 3 and run them in parallel. Don't forget the big heat sink needed.
EEVblog #102 – DIY Constant Current Dummy Load for Power Supply and Battery Testing | EEVblog - The Electronics Engineering Video Blog

Originally Posted by Maine Sail

Bill,

I've done the light bulb thing but want more accuracy than I can attain doing that and I don't want to baby sit it for 20 hours turning stuff on and off to maintain the load.. I want to be able to do this for my customers with expensive banks to keep track of performance year to year.

For an accurate 20 hour Ah capacity test I want my load to be exactly what the manufacturer has rated it for.

Eg:100Ah battery / 20 hours = 5A load until the battery hits 10.5V
or 125Ah battery / 20 hours = 6.25A load until the battery hits 10.5V

In both of the above batteries, when new and broken it, they should last 20 hours before hitting 10.5V at 77-80F..

The Victron monitor gives you 90% of what you need and a positive cut off that can be set for 10.5v to stop the test. I also keep my various meters connected during discharge to monitor all parameters.

Any load above or below the 20 hour rated load will give pretty close results but will be skewed by that guy named Peukert, somewhat.. I am also testing them at 77F so I have the right temp. Hotter than that and you gain some capacity in Ah and colder you lose some..

I know this is picky but I feel it should be able to be done... I really can't believe no one makes a tester that will do this..? I've been researching for what I want and just can't find it..

[Richard Harper]

You may be over thinking this. You need a repeatable setup. I don't think lightbulbs will give repeatable results. The resistance will change with temperature and time. I built a load bank using 5 ohm power resistors (Digi-Key FVT100-5.0-ND), switching in as many as needed for the load current. The current will vary with the voltage but this is only a minor effect on the AH rating.
I think the biggest source of error would be accurately changing the battery to the same state of charge each time.

[Rick]

I have had great performance doing battery testing with non-sine wave PWM inverters and a 120V light bank with switches to swith in various power light bulbs to get the appropriate load required. This is easy to do. Have had 1% accuracy. Pick a power inverter that delivers good efficiency over the input voltage range (10.5 to 13V). Also check that the inverter will not kick out from low voltage (10.5V) AT the total power that gives you the 20 hour current rating desired.


BTW: Doing 100% cycling testing does decrease the life of any lead-acid battery and is, therefore, for research data only. I would not 100% cycle a customer's battery or one for my own use. One never knows what subsequent cycle life may be for a particular battery tested in such a manner. One merely infers from one battery test results that another new one not tested may be expected to perform similarly. This is quite true for gel-cel and agm high quality batteries.

[btrayfors]

Maine,

Here you go: Model 8540 - 150 W DC Electronic Load - B&K Precision

Only $500. It handles up to 150 watts, so would be good for batteries up to about 225AH total.

Nice piece of gear. They make larger ones, too, but quite pricey.

See pp 7ff under Application Notes re: battery testing.

Cheers,

Bill

[jewel114]

Try a mechanical load rather than a resistive load as the temp will remain constant. Starter motor and drag...

[goboatingnow]

An electronic load is what's needed. Easy circuit to build just watch the junction temp of the power FET. Easy to mod as constant current, constant voltage, or constant resistance.

Dave