*Update* *Pics* Building a 12-cell NiCd/NiMH pack

I presently operate an MR16-based 50W halogen light, powered by a 7Ah lead-acid gell-cell. While the battery is excellent for standby use and ready at a moment's notice (I have a charger designed to maintain it, not a dumb trickle charger) I want whiter light and a lighter pack, possibly at the expense of runtime (I can presently get a good solid hour out of it, 30 minutes would not be objectionable). I am considering using two 7.2V sub-C R/C race packs connected in series for a total of 14.4V (typ.) to power the light. I have a charger designed for charging R/C packs (can handle 4-8 cells, .5-6.5A variable charge rate) and have thought about using a switch to connect the two packs in parallel for charging, series for operation. Any thoughts on this? These will be high-quality race packs (containing matched cells) and I don't plan on taking it all the way down until it croaks. I don't think connecting the two packs in parallel would be a problem, but I thought I would ask anyway--it would make charging the entire unit a lot faster and easier (I should be able to increase the charge rate by at least 75%, assuming the two packs share the current evenly). Switching the packs in series/parallel would also make a simple way to lower light output (at the expense of it looking yellow) and significantly increase runtime as needed. Any input is appreciated! /ubbthreads/images/graemlins/smile.gif

Re: Building a 12-cell NiCd/NiMH pack

Hmm. I think that the only potential problem is the part about 'assuming the two packs share the current evenly.'

When you put two loads in parallel, say two resistors, the current _never_ balances exactly evenly. But if the two loads are reasonably matched, then the current will balance well enough. The real question, however, is what happens when there is _imbalance_ in the current flow.

In the case of things like lamps, if you take two lamps that nominally have the same resistance and put them in parallel, and one lamp actually hogs the current, then it will heat up more than the other one. But as the lamp filament heats up its resistance goes up, and the system quickly reaches a stable equilibrium with pretty good sharing of current.

But NiCd and NiMH cells show a voltage _drop_ when they are heated. This is where the voltage drop detected by 'peak chargers' comes from. If one set of cells consumes more than its fair share of current, then it will heat up more than the other set, and its voltage will drop, and it will draw even more current. I could see a situation where one set of cells charges first, and then just sits there staying hot and cooking, while the other set of cells doesn't charge.

Given that you are starting with matched cells, if you do something to maintain the current balance, or do something to keep the temperatures matched, then the system has a reasonable chance of working...but there is a known failure mode (as described above) that IMHO makes it a poor idea.

-Jon

Re: Building a 12-cell NiCd/NiMH pack

Jonathan! Welcome back. You have been keeping a low profile lately. CPF has been a poorer place lately without your clear, thoughtful, patient, and technically accurate contributions.

Re: Building a 12-cell NiCd/NiMH pack

Wow. Thanks for the input Jonathan. I was concerned about some of the same things myself, but I thought I would ask.

I got the parallel idea from the fact that the C. Crane charger charges individual cells (not packs) in parallel, and that it might be scalable. I guess I could do some testing, watching with a current probe to see how well the two packs share the current.

How, then, could two batteries in parallel discharge evenly?

Re: Building a 12-cell NiCd/NiMH pack

While Jon post is correct. I see another alternative. The packs are reasonably matched, this much has been said already.
Facts:
1) During parallel charging, pack voltage gradually increases.
2) The pack with the lower voltage will take more of the current.

The more fully charged pack's voltage will be slightly higher than the other, thereby slowing it's rate of charge acceptance. Eg... the system will reach equilibrium quickly an easily. I used to charge these exact same r/c racing packs the same way myself. Also, a LOT of these questions go away depending on charge rate. Even at 6.5A wide open charge, that's only 3.25 each. Most good r/c packs can stand a little overcharge if that even ever happens.

The thing that has not been mentioned is that the charger will heat up a good bit more. The parallel packs show 1/2 resistance compared to a normal pack. This makes the charger work harder to regulate current. They also will charge for twice as long as a regular pack, which blah blah blah...

Bottom line, it's probably fine. For the first few times, monitor the packs and charger temp/etc. I would start by trickling both packs separately, and repeating this if the packs are ever drained too much. Were trying to keep them in balance here.

Re: Building a 12-cell NiCd/NiMH pack

The charger I have uses a PWM output to vary charge current, through .235 ohms (for current sensing). I was thinking along the same lines as well, about the more-charged pack taking less current due to its higher voltage. It's fan-cooled and uses a switching power supply (which is also fan cooled, with a thermal cut-out) so I don't think I'll run into any temperature issues with the charger.

Initially I would peak charge both packs independently, and cycle them a few times to break them in. After that I would buss them together and closely monitor each pack, measuring current (now I need to find some precision current shunts) and pack temperature. I probably won't charge the whole thing at the 6.5A capacity of the charger, but a 4A (2A per pack) rate would probably be more than acceptable. At this rate it would take a little over an hour and a half for 3300mAh packs. Again, I'll experiment with it and see what works best. Thanks for the input everyone!

Re: Building a 12-cell NiCd/NiMH pack

I use the "power pole" connectors for all my r/c and low voltage dc stuff. They make it really easy to hookup stuff in funky ways. You should be fine with what you're gonna do. I would worry more about putting too many volts into that bulb. R/C packs tend to run a little high on voltage, especially when fast charged. I bet initial load voltage is WELL over 15 volts. As far as shunts go, just grab a decent DVM/DMM... lots easier.

Re: Building a 12-cell NiCd/NiMH pack

I use a Duratrax IntelliPeak charger (the deluxe model--with the LCD display) for my R/C stuff. I have been using the standard Tamiya connector, but I'll probably solder the pack leads to the series/parallel switch and use some kind of Molex connector for charging and a second (of a different type) for output.

I have a couple cheapie 1500mAh packs that I was doing some testing with, here is the info so far:

50W 12V MR16 halogen lamp

-14.5V across the lamp terminals, under load (this is after lead resistance and the like)

-4.5-4.6A sustained current

This is after peak charging both packs separately. A total of 66W isn't to terribly overdriven, I could see this in a worst-case scenereo in track lighting. Even if the higher-quality packs do provide more voltage, the lamp life is specified at 6000(!) hours. I probably woudln't accumulate 200 hours over the period of a year, and even replacing on a yearly basis is still not too bad. The fixture is rated for 85W, so I should be safe there--and even if I do blow up the halogen lamp, there is a glass cover shield in front of the reflector integral to the lamp and a nice glass window in the fixture.

I have a great DMM, but I wanted to have the same resistance in series with both packs so I wouldn't set up a current imbalance. Now, if I had two DMMs...wait, that would be the same as having two shunts. /ubbthreads/images/graemlins/smile.gif

What are "power pole" connectors?

Re: Building a 12-cell NiCd/NiMH pack

I use the power pole connectors for all my rc stuff because they are so versatile as well as having a low resistance (something similar to two 14 gauge wires directly soldered to each other...) You can get them at most hobby shops or tower hobbies http://www2.towerhobbies.com/cgi-bin/wti0001p?&I=LXD176&P=7

Re: Building a 12-cell NiCd/NiMH pack

Thanks for the link. I might get some of these, if nothing more than to play around with.

I have purchased the components for this project. It's not quite what I was thinking originally, but the changes will better suit my application.

- (2) Radio Shack 3000mAh 7.2V NiMH R/C battery packs (23-431)
- (1) Female 4-pin Molex Connector (274-234)
- (2) Male 4-pin Molex connector (274-224)
- (1) Male/Female pair, 2-pin Molex connector (274-222)
- (2) 7.2V battery pack/RC car repair kit (23-444)

I used the Molex connectors and repair kits to make three adapters. I decided not to modify the battery packs in case I wanted to use a different set of packs, or use these for their originally-intended purpose. The first adapter has two Tamiya plugs (which mate with the battery packs) on one end, and a 4-pin female Molex on the other. The second has a 4-pin male Molex on one end, and a 2-pin female Molex on the other (the 14.4V output), wired in such a fashion so that the two packs are connected in series when this adapter is plugged into the first. The third adapter has a 4-pin male Molex connector on one end, and a battery-end type Tamiya plug on the other (this connects to the charger). This adapter is wired in a method that connects the two batteries in parallel for charging. Avoiding the useage of a project box keeps the entire thing small, in fact the two batteries are pocketable, side by side.

After charging both packs separately at a 1 amp rate, I checked the open-circuit voltage differential as 100mV. This is after one pack has been sitting idle for 3 hours, and the other fresh off the charger. As a test I then connected the packs in parallel through my ammeter...1.8mA. This would probably drop to 0mA once the "hot-off-the-charger" (in reality it was barely warm to the touch) pack settled down. (5 minutes after writing the above, the OC differential was down to 50mV.)



This is a picture of the whole shebang, meter showing OC pack voltage. Note the lamp fixture at the bottom left.



Here are the adapters. #3 (top) is connected to the charger's output (not shown), #1 is connected to the two battery packs(lower-left), and #2 is connected to #1 (lower-right), meter leads still probing the output connector.


My next step is to test the packs in series driving the lamp, and I'll watch the voltage output of both to see how far out of whack they might get.

(~15 minutes into the first test) One pack shows 7.31V, the other 7.29V. This is the very first discharge cycle, with no conditioning of the packs whatsoever. The lamp's base temperature is a sizzling 296 degrees celsius (565 degrees farenheit)...54 degrees C below the rated limit. /ubbthreads/images/graemlins/smile.gif Temperature rise (in still air, ambient room temperature) is ~100C (212F). The plastic "handle" remains cool to the touch however.

(~30 minutes into the first test) One pack is slightly warmer to the touch than the other. Its voltage is now 6.6V (1.1V/cell), my cutoff for this test. All cells in the pack are evenly warm, so I don't think I'm reverse-charging any here. The other pack is still 6.90V. It took slightly more charge than the first. Perhaps some cycling will even things out. The two packs recover to 7.16V and 7.44V. If I buss the two packs together at this stage, a 300mA current passes from one to the other, but this value steadily drops as the two packs equalize. I seriously doubt that any hazardous currents would be present after normal operation. Still, I will charge the packs separately one more time.

I use an almost identical scheme to externally power a small low power (5 watts) ham radio transceiver. Originally used the RS NiCD racer packs, then later switched over to NiMHs when they were on sale last year. These charge nicely using a MAHA C777 charger, or I can use the cheaper, simple RS charger for these and the 9.6 8AA packs (which are, incidentially, very similar to the internal battery pack that this particular radio (Yaesu FT-817) uses. 14.4VDC is a tad high for this radio, I use a 3A diode to drop the 0.6VDC and make it a nice safe 13.8VDC. Also prevents any polarity reversal accidents, which are difficult to happen due to the Molex connectors, but then, Murphy strikes when least expected.
A reverse-connected LED just stays dark, a reversed transceiver becomes a very expensive smoke generator.

/ed B in NH

[ QUOTE ]
Hemingray said:A reverse-connected LED just stays dark, a reversed transceiver becomes a very expensive smoke generator.

[/ QUOTE ]

A nasty smoke generator, at that! /ubbthreads/images/graemlins/ooo.gif To add to your statement, an incandescent light doesn't care how you hook it up. /ubbthreads/images/graemlins/smile.gif The Molex-lookalike connectors on the batteries are polarized, and the "real" Molex connectors I used for the rest of the project are very polarized. I checked, double-checked, and then triple-checked my wiring, because if I reversed the leads of one pack while parallel charging it, something's going to go boom! /ubbthreads/images/graemlins/icon15.gif

So far the packs seem to be breaking in nicely, one measures 3104mAh, the other 3199mAh...at room temperature right after charging and cool-down, discharging to 6.0V (1V/cell) at a rate of 5-6A, which is higher than the lamp's draw of ~4.6A. The voltage starts falling off quite rapidly around 6.15-6.20V, and at at this point the lamp is actually at rated voltage so I should get Xtreme Performance out of this thing up to the very end. /ubbthreads/images/graemlins/crazy.gif I probably should use some kind of low-battery detector circuit so I don't push any cells in reverse, given that the light won't really start to get yellow and dim until this happens. On that note, the packs get mildly warm toward the end of their discharge cycle, but no one cell gets noticeably warmer than the others. I'm going to see how well they parallel charge after another independent charge test. I'll do this by parallel-charging the two packs and then independently measuring stored capacity. Of course, I'll closely monitor the temperature of each, and immediately stop the charging process if one starts to get significantly warmer than the other.