The Ultimate Taboo

[Battery Guy]

Greetings Everyone,

I am about to admit to committing the ultimate battery taboo. Please bear with me and don't judge too quickly.

I just received my CL1909 bulbs from FM. The first mod I had planned for these bulbs was a 2D Mag with an 8 AA PowerGenix NiZn cell battery and an FM bifocal reflector. I have had a couple of weeks to do resistance measurements on the flashlight and performance testing of the batteries, and all of my results indicated that the instantaneous voltage of this fully charged battery in this flashlight body would be slightly below the insta-flash voltage of the bulb.

Well the moment of truth came today. After carefully cleaning the bulb and installing fully charged (and rested for 5 hours) NiZn cells, I turned on my new flashlight for the first time, and a beautiful and rare $16 bulb was destroyed.

I know that I must be only 100-200 mV above the flash voltage, and I also know that I have several options: soft start switch (probably the best option, but I am still waiting for AW to have more available), replace one of the NiZn cells with a dummy AA, add a resistor to the tail cap, etc... But all of these things require that I wait, or, as in the case of the dummy cell option, I sacrifice pack voltage and therefore light output.

So, in my impatience, I did something that I tell people to never do: I mixed cells with different chemistries. To be specific, I replaced one of the NiZn cells with an Elite 1700 mAh NiMH cell. I put in a new 1909 bulb, turned the light on, and this time no ! A quick voltage measurement indicates that I am running at ~12.7V under load on a fully charged battery, which is just under the flash voltage which is somewhere between 12.8V and 13.0V. In other words, perfect!

At first I felt very guilty for doing this. But then I started to think hard about it, and quite frankly I cannot think of a reason why I should not use 7 AA NiZn cells with 1 AA NiMH cell in series. The Elite 1700 cells are very close in capacity to the NiZn cells (about 1600 mAh for the Elite and 1500 mAh for the NiZn). Obviously, the cells need to be charged separately, but the NiZn are not supposed to be charged in series anyway (at least if you follow PowerGenix charging recommendations, but that is a topic for another thread ).

I know that the immediate reaction that most of you have is or :tsk: or :shakehead:. But I want you give this a little thought and let me know if and why this is a mistake, or if you think this is a viable solution. Ultimately, I will go with a soft start for this light, but this idea of "tuning" a battery pack voltage by mixing Elite 1700 and PowerGenix NiZn cells has me intrigued.

Looking forward to hearing your thoughts.

Cheers,
BG

 

[waddup]

a $16 bulb

why do you want to push your bulb so hard 99.6%?

why not get a brighter bulb and run it at 88% for the same amount of lumens and longer bulb life?


destroying things (cars, bodies,bulbs, machines) by driving them too hard never makes any sense to me?

 

[Mr Happy]

I always cringe inside when I instaflash a bulb. I did two bulbs recently: they are 6 V xenon bulbs that I have previously tested beyond 8 V without failing. I have been running one of them in a Mag powered by 6 eneloops but the light is a little soft. So I thought, what if I try 6 NiZn cells? On my first try everything seemed fine, but it was a frosted bulb. So I put a clear bulb in, switched it on and flash, dead bulb. Hmmm, I thought, maybe the bulb was faulty, so I tried another one. Flash, same thing. Here I decided to give up and not destroy any more bulbs.

But, what I could do is gradually replace the eneloops one by one with NiZn until the bulb reaches a nice brightness. I don't know why I didn't try that at the time.

Really, there is nothing wrong with this as long as you know what you are doing. If the batteries are fully charged and are reasonably matched for internal resistance and output current the setup will work.

Where problems might occur is when the batteries run down. Near the end of discharge you are likely going to reverse polarize one of the cells. So don't do that. Be conservative and recharge early.

 

[Battery Guy]

I always cringe inside when I instaflash a bulb.

I'm right there with you. I feel sick that I destroyed one of these custom made 1909 bulbs. I did not that bulb on a whim. I really need to go back and figure out what was wrong with my resistance measurements, because it looked like this was going to work.

Where problems might occur is when the batteries run down. Near the end of discharge you are likely going to reverse polarize one of the cells. So don't do that. Be conservative and recharge early.

This is very true for NiZn cells in general. I thought about this, and as long as the capacity of the cells are closely matched and I don't let the pack drain down too low, I think that the 7:1 NiZn:NiMH is a reasonably good solution. But somehow it just feels wrong.

Thanks for the input Mr. H!

Cheers,
BG

 

[45/70]

...... But somehow it just feels wrong.

Probably just because, it is.

There are a lot of things that you can do, that you really shouldn't. I suppose the worst thing that I do with flashlights, is run two unprotected LiCo cells in series. As in your situation though, as long as you are aware of the danger in doing so, it's acceptable, well OK, maybe not really, but......

I think where this can go that it shouldn't, is when one gets used to the idea of doing these kinds of things. In the case of mixing cells for example, eventually someone ends up mixing an unprotected LiCo cell with a bunch of A123 cells and !

I think maybe when we do this stuff, we just shouldn't mention it.

Dave

 

[MrGman]

Battery guy your logic is faulty. Just because 2 different battery chemistries state they have the same capacity that does not make it true, nor due they necessarily have the same capacity at a certain discharge rate. The Nickel Zinc may actually have more "apparent" capacity at the specific discharge rate to run the lamp than the NiMH. Or the other way around. You know that many batteries even of the same "type" but from different manufacturer's don't all discharge the same way.

Some like to discharge at low current levels and some do better at higher currents. So to say two different batteries both have as an example a 2600mAH capacity doesn't mean they really do if you discharge them at 2 amps versus only 500mA.

So the only way you will know for sure is to do a real load test run and find out, but you may find out the hard way when the batteries go and not just a lamp.

Now if you were really hell bent on doing this the best way would be to have the cells out in the open going into the lamp as the load or a closely equivalent resistor load for the nominal current draw of what the lamp would be. Next would be to have a voltmeter across the NiMH battery and another one across a Nickel Zinc. If you could keep the one on the NiMH and cycle the other meter across each of the NiZn to see what they are doing as they discharge. Let them all discharge and see which battery has its voltage collapse first, then open the circuit, that might tell you something usefull. Just running the flashlight with the mixed batteries because that's the voltage that doesn't blow the lamp isn't sound safe logic, might make your flashlight do this when you least expect it.

You know you cannot say with any certainty that it can't or won't happen.

At the very least you could just use one of each type of battery and set up a resistive load close to what the lamp will draw from the total stack and measured the voltage drop across those 2 batteries and see which one collapses first, repeat the test with other fresh batteries and change the order they are in. Some will say it doesn't matter what order they are in the current is in a loop. Do it anyway for independent and unbiased test results.

Be safe. G.

 

[LuxLuthor]

I think you have shown the appropriate contrition to the God of Batteries for committing this mortal sin, and now with 3 Hail Mary's and 4 Our Fathers, your soul will be cleansed of the demons that possessed you.

Seriously, I also don't have much problem with you doing this consciously, and that it is using Nickel chemistry cells, which don't have the risks of Lithium Cobalt.

You already know that they have different internal resistances and will discharge at different rates, and as others have already said, just don't push it to the drainied 'bloody' end, and you should be fine.

 

[Battery Guy]

I think where this can go that it shouldn't, is when one gets used to the idea of doing these kinds of things. In the case of mixing cells for example, eventually someone ends up mixing an unprotected LiCo cell with a bunch of A123 cells and !

Agreed. That would not be a wise thing to do.

I think maybe when we do this stuff, we just shouldn't mention it.

Dave

I thought about that, but then again one of the great things about forums like this is that we can all share ideas: good, bad and other.

Cheers,
BG

 

[Battery Guy]

Battery guy your logic is faulty. Just because 2 different battery chemistries state they have the same capacity that does not make it true, nor due they necessarily have the same capacity at a certain discharge rate. The Nickel Zinc may actually have more "apparent" capacity at the specific discharge rate to run the lamp than the NiMH. Or the other way around. You know that many batteries even of the same "type" but from different manufacturer's don't all discharge the same way.

Some like to discharge at low current levels and some do better at higher currents. So to say two different batteries both have as an example a 2600mAH capacity doesn't mean they really do if you discharge them at 2 amps versus only 500mA.

So the only way you will know for sure is to do a real load test run and find out, but you may find out the hard way when the batteries go and not just a lamp.

Now if you were really hell bent on doing this the best way would be to have the cells out in the open going into the lamp as the load or a closely equivalent resistor load for the nominal current draw of what the lamp would be. Next would be to have a voltmeter across the NiMH battery and another one across a Nickel Zinc. If you could keep the one on the NiMH and cycle the other meter across each of the NiZn to see what they are doing as they discharge. Let them all discharge and see which battery has its voltage collapse first, then open the circuit, that might tell you something usefull. Just running the flashlight with the mixed batteries because that's the voltage that doesn't blow the lamp isn't sound safe logic, might make your flashlight do this when you least expect it.

You know you cannot say with any certainty that it can't or won't happen.

At the very least you could just use one of each type of battery and set up a resistive load close to what the lamp will draw from the total stack and measured the voltage drop across those 2 batteries and see which one collapses first, repeat the test with other fresh batteries and change the order they are in. Some will say it doesn't matter what order they are in the current is in a loop. Do it anyway for independent and unbiased test results.

Be safe. G.

All very good points (except for the one about my faulty logic ). I have done a fair amount of testing, and the NiZn cells and the Elite 1700 cells actually have similar power capabilities, and at these discharge rates (5-6 amps), both deliver nearly 100% of rated capacity. So for this particular match in this application, the discharge capacity of the cells are nearly identical.

Cheers,
BG

 

[Battery Guy]

You already know that they have different internal resistances and will discharge at different rates, and as others have already said, just don't push it to the drainied 'bloody' end, and you should be fine.

If they are in series, all of the cells will see the same discharge current, which will be determined by the pack voltage and the bulb. For this particular configuration, the current draw should be around 5.8 amps (based on your destructive testing :thumbsup, and both the NiZn and Elite 1700 cells have very similar capacity with that load.

So from that perspective, I think that it should be fine.

My biggest concern is of course polarity reversal, but then again this is already a big concern for the NiZn cells.

Cheers,
BG

 

[gtwace]

I don't see any problems with mixing cells when running them, except the higher voltage cells may damage the NiMH cell by over discharge it and cause it to become a dummy cell, but if you watch the discharge rate and don't over discharge, it should be ok.

THe only problem of mixing cells comes with storage, whereby the cells can't reach equilibrium with each other due to different chemistry, thus the charge discharge cycle continues in storage and damages the cells.

 

[Mr Happy]

THe only problem of mixing cells comes with storage, whereby the cells can't reach equilibrium with each other due to different chemistry, thus the charge discharge cycle continues in storage and damages the cells.

You might want to think this through a bit more. Cells can only reach equilibrium with each other if they are connected in parallel, and you should never connect different kinds of cell in parallel. That one is absolutely a case of don't do that, ever.

 

[Battery Guy]

You might want to think this through a bit more. Cells can only reach equilibrium with each other if they are connected in parallel, and you should never connect different kinds of cell in parallel. That one is absolutely a case of don't do that, ever.

Agreed. Very bad things can happen. Actually, one shouldn't even connect two cells with the same chemistry but different state-of-charge in parallel. Connecting a charged cell in parallel with a discharged cell can result in a huge peak current, similar to an external short circuit. And trust me, the cells can get HOT if they have a low internal resistance.

One thing that I did think about last night was the potentially large variation in self-discharge rate between the NiZn cells and the Elite 1700. Not a deal killer, but something to be aware of.

Cheers
BG

 

[45/70]

You might want to think this through a bit more. Cells can only reach equilibrium with each other if they are connected in parallel, and you should never connect different kinds of cell in parallel. That one is absolutely a case of don't do that, ever.

......Actually, one shouldn't even connect two cells with the same chemistry but different state-of-charge in parallel. Connecting a charged cell in parallel with a discharged cell can result in a huge peak current, similar to an external short circuit. And trust me, the cells can get HOT if they have a low internal resistance.

I believe this does apply to most cell chemistries. I'd like to point out however, that when dealing with Li-Ion cells, due to the different behavior of their chemistry, this isn't necessarily true, provided some guidelines are adhered to.

SilverFox did some testing charging Li-Ion cells in parallel. Not only is it possible to charge paralleled Li-Ion cells with a difference in voltage of 0.5 Volts or less, but it is also possible to charge Li-Ion cells of different sizes and capacities in parallel (from this thread).

Sorry for going pretty much OT BG, but thought I would bring this up, anyway.

Dave

 

[MrGman]

All very good points (except for the one about my faulty logic ). I have done a fair amount of testing, and the NiZn cells and the Elite 1700 cells actually have similar power capabilities, and at these discharge rates (5-6 amps), both deliver nearly 100% of rated capacity. So for this particular match in this application, the discharge capacity of the cells are nearly identical.

Cheers,
BG

Had I known you did so much testing of this already I might not have been so concerned. Would still be interested to see a repetitive series of tests of NiMH and NiZn in series to a heavy load and which one collapses first for say 10 sets of batteries, it would be interesting.

But on the other hand my real question at this point is that lets say this configuration actually works for you and you constantly use the collection of NiZn batteries with 1 NiMH. After say 100 recharge cycles, will the NiMH still hold the same high current draw rate and ampere hour capacity to keep up with the Nickel Zincs? What if say after 50 charge cycles the NiMH starts showing a substantial reduction of capacity and the NiZn's don't, is something like this a legitimate long term concern?

How would you know? Would you be keeping track of the amount of recharge the battery took? I don't have any chargers that tell me how many ampere hours worth of charge an AA battery takes. I used to have one for my NiCad battery packs for the bigger RC cars that told me that. I haven't even seen one for AA batteries. Do you keep such careful track of how much charge your batteries are taking?

 

[kramer5150]

Question for everyone... Could the differing cell resistance from one chemistry to the next play a factor and reduce safety?

It is risky, what you are doing... but it seems you have taken some extra safety precautions that most do not. The biggest risk is depleting the cells excessively... so definitely do not do that intentionally.

If it were me, I would stick with all NiZn cells and just under-charge them (equally mind you) a little. Thats what the Lithium-Ion and LiPo crowd does... and its proven to be an effective way. That will carry you over at least until you get the soft start.

Is not hard to fabricate a dummy 14500 sized cell. Its just a wood dowel and solid core copper wire or a nail. You dont even need a drill. Just clamp the dowel in a vise and cut a notch into it with a hack-saw. Take some solid core wire strip the last .2" from ends and stuff it into the crack.

Maybe its just me, but I think there are less risky ways to accomplish a lower voltage to the lamp.

(sorry if this is a repeat... I didn't read all the other replies besides your own too carefully).

 

[jhellwig]

Voltage is voltage. It doesn't matter where it comes from.

All you need to do it get an amp reading and then you can estimate fairly close to when you are close to having used up a cell. From the sounds of it you have that map reading or can get it. You know that at X amps the cell will go for this amount of time. Give yourself a margin of error and you are fine.

You can mix and match any cell in series. Even li-Ion with nimh nicd lead acid whatever. As long as you make sure that you keep each cell within it specs nothing bad will happen, most of the time. Accidents do happen. Mistakes can and will be made. Results may vary. Yada Yada Yada.


Ok Now you can yell at me for saying that.

 

[Battery Guy]

I believe this does apply to most cell chemistries. I'd like to point out however, that when dealing with Li-Ion cells, due to the different behavior of their chemistry, this isn't necessarily true, provided some guidelines are adhered to.

You are totally correct about charging lithium-ion cells in parallel. It is not a problem at all. I only meant that when they are initially connected in parallel, they should be a similar states-of-charge to avoid a high current spike between the two cells.

Cheers,
BG

 

[Battery Guy]

But on the other hand my real question at this point is that lets say this configuration actually works for you and you constantly use the collection of NiZn batteries with 1 NiMH. After say 100 recharge cycles, will the NiMH still hold the same high current draw rate and ampere hour capacity to keep up with the Nickel Zincs? What if say after 50 charge cycles the NiMH starts showing a substantial reduction of capacity and the NiZn's don't, is something like this a legitimate long term concern?

This is an excellent point. If the NiMH cell degraded faster, then I would be driving it into reversal and could potentially get leakage. However, I think that the opposite is going to be the case. I seriously do not expect to get 100 cycles out of the NiZn cells before they need replacing, at least at this high of a current drain. I suspect that the NiMH cell will hold up much better. I am hoping to get sufficient life (25-50 cycles) to bridge the gap to a soft start. At that point, I will replace all 8 cells with a new set of NiZn.

Cheers,
BG

 

[45/70]

......I only meant that when they are initially connected in parallel, they should be a similar states-of-charge to avoid a high current spike between the two cells.

Well, similar in that they should be within 0.5 Volt. This can be as much as an 80% difference in SOC, which I wouldn't say is "similar". Mind you I've never tried this myself. It's one of those things that "just seems wrong".

I found it interesting what SilverFox found though. The "spike" is of such a short duration with Li-Ion chemistry that, although the spike can involve a considerable amount of current, it is within the tolerances allowed, as it happens so fast. Again, this concerns Li-Ion cells only.

Sorry, I realize this thread is not about Li-Ion cells. Back to the original topic.

Dave