Battery capacity testing

There are some nifty hall-effect sensors with almost zero drop. Used one in a solar/wind system with a shunt regulator. Wind generators don't like being unloaded by a series regulator.

I've been

I'd like to hear more about your desulfation successes. High voltage didn't help. Other crazy ideas I'd read about, like AC at various frequencies to 'ring' the plates and shake off sulfation, etc. Got absolutely nowhere.

How stable is the zero reading? This is the best DC Hall effect ammeter I've seen and its zero drifts by several mA per minute if held still. When moved the Earth's magnetic field throws it way off.

I've fixed several "dead" U1R lawn tractor batteries and used them for several years. The symptom was not accepting more than a few milliAmps from a commercial charger, although the resting voltage seemed OK. The fix was using the variable power supply to force 16V to 17V which caused the current to very slowly increase and then the required charging voltage to drop, an unstable condition that requires current limiting, such as with a low end lab-type supply like these if you can't rig up your own.

After the battery had accepted enough charge to raise the electrolyte level I checked specific gravity and found one or two low cells. Charging at a current that didn't make the others bubble excessively, around 0.5A, eventually brought up the low cells. The resurrected U1R in the tractor now can put 150A into my HF carbon pile load tester.

I've read that salvaging batteries this way can take up to a week. I saw progress with salvageable flooded batteries in a few hours but haven't had much luck with AGMs.

The neighbors who give me these batteries know I can fix them, and that they will need frequent attention afterwards. I have to top up the charge at least monthly or their internal resistance will rise again.

I don't know if the cause is literally lead sulfate recrystallization or not but it's a handy suspect to blame. There's a theory that it's a thin oxide barrier between the grid and the active material. Whatever the cause, the effect is a very high internal resistance. Automatic chargers see the voltage rise as though the battery was fully charged and shut off. Some people have reported success using a Harbor Freight manual charger on a Variac as the variable supply.

-jsw

Ahh, OK.

OK, so your -fridge- demands 0.5A? It's hard for me to conceptualize a half amp draw at 12v, as few things operate at that level.

I'm fairly fond of 18650s, having used them in the LED flashlights for awhile now. I love that Lithiums don't self-destruct, and don't suck their own juices from themselves while they sit on a shelf. Nicads and LAs have always bothered me with that nasty characteristic.

Amazing, but good news.

Notalotta, but it's a bummer when they do.

Are you trying to (or not to) tell me about a box full of 18650s you're putting together to run your fridge when the power's out? Testing those with a 6w drain just might be the connection my mind continues to search for with your bloody half amp thing.

Maybe I should have repeated that I was discussing a different battery closer to the mention of the fridge? I post all this to practice clear, concise technical writing and then try to determine why it's so often misinterpreted.

The fridge takes 120W initially, slowly dropping to 80W, as measured with a Kill-A-Watt on wall AC. When the power goes out it runs from an APC1400 SmartUPS pure sine inverter with two external 12V Group 31 batteries. Discharge tests on them could end in the middle of the night and let my fridge warm up, so I bought equipment to automate the capacity measurement with safe loads.

I'm checking out that equipment on less valuable batteries that don't take nearly as long to discharge. I've been posting a summary of significant observations, not a daily diary of my experiments, so I haven't always clearly indicated which battery or test load they came from. The 0.5A test was to measure the C/20 capacity of a smaller AGM battery from a 350W UPS.

I found a source of tabbed Li-ion 18650s for $2 each and hope to restore old laptop batteries etc with them. They appear to have been in storage long enough that some dropped below their low voltage safety disconnect, but still accept a charge applied directly to the tabs.

-jsw

Hmm.

You added water to an AGM? How did you do that? And how had the AGM lost water?

I've seen soldered IC, resistor, capacitor forever encased in clear RTV which, while not purty, was fully functional.

OK, I think I now have the full picture.

So "dead" batteries go into meltdown mode once they do finally start accepting a charge? I see why most people have no success at it. Even with the "save your battery" goo scam.

I wish I'd grabbed the cheap Variac which was offered to me long ago.

For most people, plug-n-play is the only way. For Makers like those of us on RCM, fiddlin' is the only way.

Interesting. Remember seeing the difference between OK after a tornado and elitists in NY after a hurricane? Everyone came out to help in OK, while elitists sat surrounded by mess and complained to the Press about their blocks, all while not one single person came out to work to clean it up. Absolute night & day differences, wot?

NO: Save the city in a hole! Don't fill in the ground and bring it above levee level so it never happens again, just put in bigger pumps. Crom, those folks are smart...not. I wonder what genetic strains will come from those folks now living where thousands of fracking ponds (full of 600 different hazardous chemicals) were strewn by Katrina. I tend to not buy anything from LA nowadays.

Were I there, I'd have months worth of supplies for everything.

New Atlas had an article recently on doubling the density of lithiums. I hope Tesla Gigafactory takes advantage of it. That should drop the price considerably.

Interesting. It looks like I'm going to have to take their "course" to get more up to speed. Until now, I had never heard how CCAs were measured.

Indeed. Some equipment cost is measured by X years of debt and probe valued at X months of wages. =:-0

Data is far easier to get scattered on Usenet, where we potentially zip through hundreds of items with dozens of topics daily. But, yeah.

My 2002 Frigidaire 18cf fridge uses only 135w with the compressor running. I didn't track it while running, so mine may have tapered off once the system normalized. That figure really surprised me, it was so low. The '80s model before it took a lot more power to run. Progress is good.

Yes, I finally came to that realization yesterday after having some fun with you. So solly.

Z-tabbed for spotwelding, or other? I'm not sure if the cheapies I bought have the protection cell, so maybe you're talking about bypassing it on your genuine cells. Maybe I'll rip one apart and see if it's done properly. (Judging by Youtube vids, they don't have the protection cell.)

I found a source of (supposedly) new 18650s on eBay for a buck apiece with wild claims of up to 6000ma capacity. I figure 1,500 to be the norm, and they're worth a buck. AC and DC chargers go for a buck a pop now and then, too, so I stock up. Bright LED (XLM-T6 and Q5) flashlights and headlamps go for $2-5 each so I got sets for the whole family as stocking stuffers one Christmas.

The 18650/T6 combo lasts a couple hours on high, and my normal use is under 3 minutes (forgotten mail, raccoon in tree) so I'm happy. The zoomable waterproof lights are great, but I seldom use medium or low beam, or the flasher or SOS modes.

The zoom function works well on the bike, but I'm not crazy enough to ride at night. It allows you to set your own beam width.

I pried off the cover plate over the rubber caps, which pull off easily. I don't know why some of the cells were dry. It's a replacement from Batteries Plus I bought in 2009 to revive a dead UPS. When new it delivered 82 Watt-hours; 12V times 12 Amp-hours is 144 Watt-hours, though only at or below the C/20 rate.

I measured computer run times on it a few times and plugged it in occasionally to recharge. It wasn't on continuously because I had changed from desktops to laptops that consume about 1/4 as much power and serve as their own UPS, and after I found the free APC1400 and rigged two external 105A-h batteries to it the small UPSs weren't needed. In 2015 I noticed that its run time was very short and began testing suggestions on how to salvage the battery.

I suspect that pulse desulfators are a simple way to apply a high enough voltage to break through sulfation while reliably limiting current with capacitive and inductive impedance instead of active control. Personally I prefer DC because an ammeter that reads milliAmps shows how much the battery is improving, or not. My own results and advice I've read suggest that a battery which needs over

17-18V applied before it accepts current can't be saved.

Dumb transformer + rectifier battery chargers have a moderate output impedance that makes the current decrease as the voltage rises and conversely increase as it drops, but not as sharply as a regulated power supply.

The charger I was using then was an old manual 6 Amp Schauer with a small 3 Amp Powerstat added to adjust for the current I wanted. The current didn't change all that much as the voltage rose or fell, so the battery didn't actually run away.

A tightly regulated power supply could possibly let a battery run away unless its current was limited by circuitry, its transformer or the solar panel source such as the HF kit which is a good match to slow battery reconditioning.

The power supply I built from a Variac and an arc welding transformer has a relatively high output impedance to give the arc its constant current characteristic, enough that the rectifier/cap output ripple at

20A is roughly a 1V sine wave. It puts out over 50V no load and drops rapidly to around 35V as the current increases, with 120V AC in. The output voltage then holds steadier up to around 25A out, its experimentally determined 100% duty cycle limit. It will briefly exceed 50A which is useful for testing components.

As a 20A 24V battery charger it holds its output current quite steady during the bulk charge phase but then becomes dangerous because it won't automatically decrease the current enough when the battery voltage rises to the gassing level. The simple fix is to pass the output through a P20L or similar cheap solar controller set to reduce the current when the voltage per battery reaches 13.6V. Light fizzing that varies a little between cells becomes visible around 13.8V.

My two series-connected batteries, of different ages and sometimes topped off to different levels, automatically self-balance to 13.6V each on the APC1400's float charging current. Each has its own voltmeter to watch that, also if one discharges faster.

Cheap Variacs may need new brushes, which are far from cheap. I bought a brass bar and some larger carbon brushes and machined my own replacements because the right ones were out of stock. The original brush geometry was too tricky to mill and assemble so I made the brushes thicker and thus less fragile and will have to trim the end contact width as they wear.

They use silver epoxy to glue in replacement carbons instead of pressing tightly fitted ones in like the originals, and after pressing mine in I understand why. I had better luck tediously cutting them to size with a razor saw and files than milling them because they are so brittle.

I think two weeks is probably enough in an area where people and local governments have the tools to recover, though not in cities where there's no place to run a generator. Germany and Japan kept their societies functioning pretty normally until the very end.

I assume my biggest need will be roof repair to prevent further damage. When a fallen tree top punctured my roof in over a dozen places I quickly covered the holes with sheet metal shoved under the shingles at the upper end. I lost the shiny finish on my 6061 aluminum but saved the house from water. Plywood covered with a large tarp would also have worked, unless everyone else had the same idea and the stores were empty. I happen to have the machines to work sheetmetal and the need to make electronic enclosures.

When Jay Leno asked New Yorker actress Lea Michelle how she handled TS Sandy she told him she didn't even own a flashlight. In my Mensa experience that's fairly typical of the complete dependency they accept as natural, can't change and don't think about. Let the good times roll.

-jsw

Don't be. I've intentionally chosen proofreaders who would take everything the wrong way when I was ambiguous. That's why I wrote the "press any key" routine that faked a crash on Alt or Shift.

-jsw

The HF carbon pile performs the test as they specify, with a 15 sec timer to warn you to turn the current down before it overheats. So far it seems to be a good and relatively inexpensive tool to measure the margin your battery has over the starting current your vehicle requires.

Here is a good, simple overview of battery charging:

I use simple dumb chargers for the high current bulk phase and smaller, more efficient and better controlled ones for the final ones. It costs nothing to leave a battery on a solar powered charger. The big chargers draw idle power, transformer magnetizing current, that may be more than goes into the battery. I've added an input tap after the rectifier to let a charger operate on solar if available or grid power if not.

"The charging parameters discussed here are applicable to ?ooded lead acid batteries. Be aware that some available smart chargers may not be suitable for other applications."

However the makers post their products' parameters, which aren't hard to meet with a voltage-adjustable power supply with meters. Some of my AGMs have the charging conditions printed on the cases.

-jsw

I don't recall seeing covers on AGM cells, just small perfs.

I got about 7 minutes from my desktop UPS (450VA IIRC). The rest have failed at the worst moments, NOT keeping my comp running when the power went off. I've bought maybe 3x the number of batteries than I have UPSes so far.

That's a lotta volts.

Ayup. That's safer.

I like the current limiting in the cheapie 0-30.0v 0-3.00A Chiwanese lab supply, and suppose I could get finer adjustment with a milliammeter inline and the fine tune pot.

Nice and beefy!

So, output from the PS to solar input terminals on the controller? Unusual, but I guess it'd work, wouldn't it? Cool.

Hooked up in series or parallel to charge? I don't know APC1400 specs or setup.

I nearly -ran- out of an electric shop once after getting a quote on brushes. I guess they hate competition.

They _are_, aren't they? Nasty stuff to work with.

Yeah, IF the local gov'ts do recover. What if the area is larger, like the whole east coast? Prepping is truly cheap, compared to the alternative.

That's why I took most of my trees down. I'd rather my house not be a target, thanks.

Luckily, nature accepts the survival only of the fittest. But the rich folks' armed bodyguards might take things from others if others aren't careful.

They had to be assembled somehow.

Yep, works fine as long as I keep the voltage from spiking above the safe input limit when the current is interrupted. The welder transformer has that nasty habit and so do poorly regulated generators. Testing this stuff drove me to buy the (older) HF inverter generator, which holds a very steady 120V from no to full load. The Coleman I had been using need to be set close to 140V at no load to be able to supply barely 100V at full load. The APC rejected its output and stayed on battery until I found how to reduce the line quality sensitivity.

The APC1400 needs 24V, originally from two internal 12V 18Ah AGMs in series. Their heavy cables run to a rear panel Anderson plug, in a Y shape with the battery on one leg, the inverter on another and a safety disconnect shorting plug on the third, normally at the rear panel. Moving the short to the internal battery position makes the rear panel plug the input.

I ran fused twisted pairs from each battery to a voltmeter and switch panel under my Harbor Freight solar controller. They connect to a DPDT center off selector switch that can tie the HF controller to either battery. Both wires in the pairs are fused to protect against any switch failure, since the batteries may still be connected in series.

I lived in the center of the track of Hurricane Carol. The eye passed right over us and was quite spectacular, bright sun lighting an enormous dark ring wall.

Afterwards my father found gasoline at a station that had rigged a Jeep to run one pump and drove around to assess the damage, as far as we could since fallen trees blocked many roads. Back then few people had the chainsaws or generator so widely available now but we managed to recover without too much trouble.

Do you belong to a church or club with an experienced leadership structure, that knows members' abilities and how to organize events?

-jsw

Have you seen these?

Cute, but just a one-time sight gag unless you can write a routine that detects and responds to them.

I've used Dial-A-Prayer for that sort of prank. It's an ominous warning when -they- call -you-, right? (transfer the call)

Yeah, it's only good for people walking by your keyboard and finding the infamous ANY key. I wonder how many they sell at $26 a pop!

Eek! (running away) I had a nice church lady spittin' nails a while back when she read the sign on my front window. I didn't have to answer the door, and I could hear her through the window and see her through the translucent white curtain. She wanted to spew religion at me and that sign made her show her true colors. I'll bet the man with her never went out with her again. He left with a smile on his face, following behind her.