Will what I'm planning to do ruin my Prius battery or my life?

July 19, 2016 12:01 PM

I am replacing a single cell in my 2005 gen II Prius' battery pack. In order to equalize the voltage in all of the cells I plan to temporarily wire all of them in parallel. Is there some reason not to do this?

I'm not asking about the advisability of performing this repair myself. I know the dangers and am confident in my precautions. Besides, it's too late. It's disassembled in my driveway.

So, I took this thing apart and measured the voltages of all 28 cells in the Prius battery pack. One is significantly lower than the others, indicating its failure.

My understanding of what causes a cell to fail is this: Over time and charge/discharge cycles, voltages start to vary between different cells. Eventually, when one or more of them is enough lower than the rest and the system's charge drops low enough, the under-voltage cell(s) become reverse-charged, ruining them.

Therefore, the most important factor to prevent near-future failure of my repair is ensuring that all of the cells' charges are the same. However, according to my multimeter, the charges of the remaining good cells varies fairly significantly, though within operating bounds.

Many sources say it's best to run each cell through several overcharge/discharge cycles to accomplish this, but that requires special equipment and each cell takes about three days to run its cycles. I don't want to buy special equipment or wait that long, and other sources I've read say that this is done in order to restore capacity somewhat, but I'm not overly worried about that this time around.

So what I plan to do is rig up a wiring harness to connect all of the batteries' positive terminals to each other, and then all of their negative terminals to each other and let them equalize overnight then, of course, put the whole thing back together the way it was before.

The forums are full of conflicting information and confident proclamations on either side of almost any issue, so I turn here.

This seems like a simple and straightforward thing to do to me. The batteries are not connected to any kind of circuitry or anything at this point, so there's nothing for me to damage. Batteries connected in such a way will move toward equal voltages.

Right? Am I not thinking of something?

posted by cmoj to Technology (15 answers total)

Best answer: Are you doing anything to limit the current?

Yes, if cell A is 7.60V and cell B is 2.00V and you connect them in parallel, the higher-voltage cell will try to charge the lower-voltage one. But the amount of current that flows is limited only by the wiring you use and the internal resistance of the cells. If that current exceeds about C/10, your batteries will be sad, and may express that sadness in ways that lead to you yourself being sad, some smallish number of milliseconds later.

The Gen 2 has NiMH batteries, which are best charged using pulse charging. Your proposed method does not accomplish this feat.

You will die in the woods covered in poo in your driveway covered in burning plastic. Do not do this.
posted by sourcequench at 12:31 PM on July 19, 2016 [15 favorites]

The difference in voltage between the cells will cause very large currents to flow between cells (I = V/R and R is quite low when you connect cells directly). Large currents will damage the cells.

You need specialized equipment to do this.
posted by ssg at 12:35 PM on July 19, 2016 [1 favorite]

Yeah, I agree with sourcequench that this seems like a Very Bad Idea^TM and a good way to end up with a potential hazmat situation at best. Effective charging of modern rechargeable batteries is not just a "shove watts into them until they're charged" situation.
posted by jferg at 12:36 PM on July 19, 2016 [2 favorites]

Measuring the open loop voltage of a battery doesn't tell you much. If you put a resistor (50 or 100 ohms?) across the battery, you will get a more accurate reading of the battery voltage.
posted by H21 at 1:32 PM on July 19, 2016

Response by poster: I realize now that I left out an important detail. The borked cell is borked. I'm replacing it with a new one that is already charged, so if it makes any difference, the maximum difference between any two cells in the system would be .07V.

Also, for whatever it's worth, I didn't dream this up all by myself. A youtube guy claims that this is what he does when he repairs them. I realize "some youtube guy" is not the greatest authority on anything, but that's basically what exists on this subject.

Anyway, it seems at the very least this wouldn't be very good for the batteries, so barring any new input, I'll just slap the new one in there and see how long it goes until I have to do it again or break down and buy a whole unit.
posted by cmoj at 2:19 PM on July 19, 2016

When batteries of this kind are charged, as they approach full charge an increasing percentage of the energy being pumped into them becomes heat instead of being stored. Chargers for NiMH and Lithium go into a mode near the end called "topping off" where the charger backflows current for a second or two, and then waits for several seconds to let the battery cool before applying current again for a second or two. This is the "pulse charging" referred to above. If you don't do this, the battery can overheat and explode.

And if I may express an opinion, your youtube guy is an idiot.
posted by Chocolate Pickle at 4:59 PM on July 19, 2016

Response by poster: I can't tell if everyone is getting that I'm not applying any new current to this system. The plan (yes, currently scrapped) is to wire all of the positives to each other and all of the negatives to each other. You could say that the higher voltage cells are charging the lower ones, but nothing will reach full charge, as no external electricity would be applied to the system.
posted by cmoj at 5:52 PM on July 19, 2016

Best answer: Assuming that all the cells are disconnected from each other (i.e., no longer in one or more series strings as I believe they would be in the car), and no cell is more than 70mV different than any other cell, then I don't think putting them in parallel will have as dire consequences as some of the other respondents, but I also don't think you'll accomplish anything useful.

Suppose that you successfully get all the cells connected and the voltage across the parallel pack is close to equal (which it should be, once the charge distributes itself across the cells, so that current no longer flows in the wires you've added, and there are no consequent voltage drops.)

Now, you disconnect the cells and begin to wire them back up in series. And they are free to drift apart again, based on temperature and self-discharge differences. So you haven't really accomplished anything.

I think what you really need to do is run a few charge/discharge cycles using a smart charger, because you're looking for cells that behave differently under load and under charging. If you can't (or don't want to) do that, I don't think you're going to get much benefit from further fiddling with the pack, compared to just wiring it back up and driving away.

Please be careful, whatever you do. High-voltage and high-current battery packs are quite dangerous: you have the typical electricity dangers, plus you have the fire & explosion dangers if you mistreat the batteries.
posted by spacewrench at 6:35 PM on July 19, 2016 [1 favorite]

Best answer: Disclaimer: I know very little about Prius batteries. I do know a little about analog electronics. My instinct is to make sure that all of the voltages and currents you're working with are super small. For voltages, your only protection is making absolutely sure you don't accidentally have cells in series. For currents, you need resistors and fuses.

If you short two batteries with a difference of .07 V and an internal resistance of (quick Google) .46 Ohms, you end up with a current of ~75 mA, which is pretty high for no overheat/over-current protection.

I'd put in a beefy resistor (less than 100 Ohms) and a very thin fuse (less than 1 mA? I'm not sure they go this low). Make sure each cell is protected separately.

I think the main danger is accidentally hooking up a cell or several backwards, ending up with a 2.4 V or more unprotected dead short. I imagine all of the cells are installed in some sort of series zig-zag, so making sure that everything is disconnected, all of the negatives are bound together, and all of the positives are bound with a resistor and a fuse to the bus is non-trivial.

I can kind of see the idea behind wanting to do this, but I'm unconvinced it's worth the hassle. I can't imagine the car's controller is letting the whole array, never mind individual cells, get so close to reverse charge that it makes a difference.
posted by lozierj at 7:12 PM on July 19, 2016

I can't tell if everyone is getting that I'm not applying any new current to this system. The plan (yes, currently scrapped) is to wire all of the positives to each other and all of the negatives to each other. You could say that the higher voltage cells are charging the lower ones, but nothing will reach full charge, as no external electricity would be applied to the system.

Yes, I understand perfectly what you're thinking.

Don't do it.
posted by Chocolate Pickle at 9:22 PM on July 19, 2016

This seems like a simple and straightforward thing to do to me. The batteries are not connected to any kind of circuitry or anything at this point, so there's nothing for me to damage. Batteries connected in such a way will move toward equal voltages.

Batteries disconnected from electronics are also disconnected from safety protections. And there is still a battery pack left to damage.

What if the faily cell shorts during this process and damages the rest of the cells? What if you have misidentified the failing cell, it's still in the pack, and by wiring it in parallel you force it to take a charge it cannot hold?
posted by zippy at 10:31 PM on July 19, 2016

My comment above should say "more than 100 Ohms."

(I noticed the greater-than sign had been interpreted as some sort of HTML right as the edit window was closing. I replaced it with words, but the wrong words.)
posted by lozierj at 10:53 PM on July 19, 2016

Best answer: Side note: "fixing" a series pack this way almost always leads to another cell failing in very short order. (I work in the off-grid solar industry and we see this with various battery chemistries in non-managed systems.) Inserting a single new cell in an older pack will mean it is the odd man out: (depending on chemistry) its series resistance will be slightly lower, meaning it'll take less of the series voltage, meaning that the highest R_s cell in the pack will now get even more over-charged than before, accelerating its death. Depending on the pack chemistry and charger, this can be mitigated, so caveat emptor.

Having said that I'll also say that hand-equalizing big NiMH can be dangerously exciting. (Read: extremely hazardous if you make a mistake.) 8 volts doesn't seem particularly daunting but it's hard to imagine just how exciting 500+ amp pulses can be.
posted by introp at 9:50 AM on July 20, 2016

Response by poster: It seems like, "not worth the hassle" is the encapsulation here. I could wire up 54 sets of resistors and fuses, giving myself orders of magnitude more opportunities to physically make a mistake for definitely inferior benefit to running the cells individually through charge cycles, or I can just put it back together with all (currently) good cells and use my bought time to save up for an all new unit. I'm convinced, and I'll go with the latter.
posted by cmoj at 9:53 AM on July 20, 2016

Batteries connected in such a way will move toward equal voltages.

While this statement holds true as long as the batteries remain connected in parallel, it's not the whole truth.

Batteries are not like little tanks of water that once balanced out, will stay that way. The reason for the differing voltages is the state of the internal chemistry of each battery. You're not going to change that in any meaningful way by a temporary parallel connection.
posted by HiroProtagonist at 8:49 PM on July 21, 2016