Cell gap - how much is too much?

[donorcycle]

I have four 6s Lipo packs I want to run 12s2p.

Two of the packs never have a cell gap over .010V whether charging at a 20 amp rate, or discharging on the charger (Hyperion 1420)

Two of the packs have a maximum cell gap briefly of .012V on discharge and .016V on charge (at 10 amp charge rate; its closer to .030 at 20 amp charge rate)

Should I put the similar packs together in parallel? Or, should I put good pack with worse pack in each parallel string?

Or, should I return the packs that have a cell gap over .010V? My charger will do a 20 amp charge total, in which case, if I am understanding it correctly, each pack will be receiving a 5 amp charge rate as I will be charging all simultaneously. Oddly, at the 5 amp charge rate, none of the single packs go over a .010 cell gap ever, so should I worry?

Thanks

 

[neptronix]

Depends entirely on where you are measuring the gap.

4.2 - 4.1v; there will be a notable gap.
4.0 - 3.7v; The gap will be extremely minimal.
3.6v - 3.0v; The gap will widen at an increasingly rapid pace. One cell being short or having an extra 50mAH will make the voltage vary as wide as 0.4v.

Where are you measuring the gap?

 

[donorcycle]

I have not discharged a pack to less than 3.8V.

The charger discharge load is only around 3.5 amps maybe 4 at max (which is annoying but maybe it has to go through the small balancer wires). The largest gap happens briefly just below 4.1V or 96-97% of discharge IIRC. My concern is, even though the cell gap is only .012 @ 3.5 amps discharge rate, will that be exagerated at 30 amps under the motor's load? Should I test down to 3.0 Volts using the charger? That seems a little dangerous, but my LVC will be at 3.0V, so it might be a good real world test.

The bigger cell gap disparity is when charging especially at 20 amps. It happens more consistently over a longer range around 3.85 to 4.00 V and is .016 Volts max at a 10 amp charger rate. It then goes down to under .010 and stays there for the rest of the charge. I may not need to be concerned with that since the charge amperage will be spread out over 4 packs rather than one. Whenever I have charged one of the more questionable packs @ 5 amps there has never been a problem, so I'm guessing charging 4 packs at 20 amps there should not be a problem.

 

[EBJ]

my Hyperion came out of the box w/ a 3.3V stop.
I just let it discharge until it stopped itself, then charged back up to storage voltage.
If you plug it into your computer you can watch the discharging happening real-time. And you can be there to pull the plug if it goes below 3.3V,
which it shouldn't since I didn't change anything on my charger and that is the lvc.

 

[donorcycle]

I swear when I select the "discharge" function on my Hyperion, it is defaulting to 3.8 something which is not very low. I definitely do not remember anything in the 3.3 range because that would make me a little nerveous. I guess I could set it lower, but should I? I do have it hooked up to the computer. I can't seem to get the cycle function to work from the computer though for some reason.

So now it appears I have one pack that peaks at a .016 cell gap on charge. Is that too much? I have done about 15-20 cycles on it, and it appears it is not getting better. I can probably return, unless it is nothing to worry about.

 

[EBJ]

I swear when I select the "discharge" function on my Hyperion, it is defaulting to 3.8 something which is not very low. I definitely do not remember anything in the 3.3 range because that would make me a little nerveous. I guess I could set it lower, but should I? I do have it hooked up to the computer. I can't seem to get the cycle function to work from the computer though for some reason.

So now it appears I have one pack that peaks at a .016 cell gap on charge. Is that too much? I have done about 15-20 cycles on it, and it appears it is not getting better. I can probably return, unless it is nothing to worry about.

I'm sorry, I gave you incorrect info.
I just double-checked and my lipo-graphs tell me my Hyperion was set from the factory to 3.5V (not 3.3V like I said earlier)

The "cycle" function didn't work for me either.
So I just have to manually have it charge, then discharge.
.015V difference during the cycle doesn't sound like a lot.

I suggest you set your Hyperion to lvc of 3.5v, then discharge. If all of the cells stay within a normal range just before hitting the "cliff", then I'd say it's a perfectly good battery.
When it hits the cliff, you should see just how "far off" one cell is. If it starts to cliff before 3.7 then you need to keep that in mind when you are setting the lvc on your bike.
Most of mine started dropping around 3.65V, so I'm gonna start my LVC at 3.7v

 

[donorcycle]

I just checked my charger and it's set at 3.8 for discharge as default from factory. I guess they are getting more conservative.

Yes I guess .016 is not that bad, and like I said, the packs do not go beyond .010 when I charge at 5 amps. I guess I am concerned because the other two packs I have never went out of .010 even charging each at 20 amps right out of the box. Which makes me wonder, should I put the two packs that fluctuate more in parallel with each other? Or should I put a stable and less stable pack together in parallel?

 

[neptronix]

Balance charge and then discharge each pack seperately and graph the results.
The hyperion comes with software to do this. It will help you find dud packs, if there are any.

What you want to look for is a cell that hits LVC far quicker than the other cells.
Some cells do indeed discharge and charge a little quicker, that is because they have lower internal resistance.

20A is a good rate for 4 packs as that's only a 1C charge rate.

 

[donorcycle]

OK, I'm sending my first one over the cliff. Thank you for the graph, Neptronix, it is very helpful

 

[EBJ]

OK, I'm sending my first one over the cliff. Thank you for the graph, Neptronix, it is very helpful

yeah, that 3.8v isn't low enough to even start to see the cliff that happens around 3.65.
Take 'em down to 3.5 like my Hyperion was set for and you can see them dive.
non of my packs (9 of em) hit evenly. There is always one cell that will trip that 3.5V before the other ones, you just want to make sure it does so after all the other ones already started diving.
and I would suggest you parallel a "good" one with an "ok" one, as it was mentioned earlier that a good cell can help out an ok cell if they are paralleled. But a bad cell (doesn't sound like you have one a this point) can easily bring down a good pack if it is paralleled with it.

 

[donorcycle]

Thanks EBJ & Neptronix. This is my first build and it has been challenging. It has been a month so far (nights & weekends), and I did not think it would take this long.

Anyway, I took 'em all down to 3.5 V which was very interesting. It got the blood pumping a bit. The worst pack I have , had a cell gap of .095 @ 3.5 V. That scares me a bit. Can something be done to rehabilitate the pack? Or should I even be concerned? I will be using the Goodrum Fechter LVC which cuts off at 3.0V under load and I will also be using a Cycle Analyst to monitor the entire pack. However, this jittery pack scares me. Should I be scared?

 

[EBJ]

Thanks EBJ & Neptronix. This is my first build and it has been challenging. It has been a month so far (nights & weekends), and I did not think it would take this long.

Anyway, I took 'em all down to 3.5 V which was very interesting. It got the blood pumping a bit. The worst pack I have , had a cell gap of .095 @ 3.5 V. That scares me a bit. Can something be done to rehabilitate the pack? Or should I even be concerned? I will be using the Goodrum Fechter LVC which cuts off at 3.0V under load and I will also be using a Cycle Analyst to monitor the entire pack. However, this jittery pack scares me. Should I be scared?

can you post a pic of the graph? I think you have to zoom in on the "cliff" and right click "save as". I know cell gaps widen a lot once they dive off the cliff. And then they align again as soon as you charge them.
The important thing is that the cells are aligned up until that "cliff" point. If you have a cell that isn't as strong as the others, it will be the first to trip your LVC. I can't remember exactly, but I think mine started getting gaps of .060 - .080 when I ran 'em all the way down to 3.5V. Unless you can see a noticeable cell-gap throughout the graph, I still don't think you have much to be concerned about, as cells become way out of whack below the 3.65 mark. (hence a lot of ppl going to 3.7)

 

[999zip999]

So what is a gap that you can live with ?

 

[theRealFury]

So what is a gap that you can live with ?

Well the gap the OP had of 0.016v or 16mv is really nothing to worry about, slight variations in internal resistances and even miniscule inaccuracies in the voltage measurement circuits on the charger for each cell can produce differences of this on even the very best lipo packs. Id say firstly dont rely blindly on the voltage readout from your charger as i have found my icharger to be as much as 0.020v out (reading 4.15v after balancing but really being 4.13v registered on a calibrated VM) i tend to worry if i see inaccuraciies suddenly open up during the middle of the charge ie 3.8-4.0v range as this can signify issues.

If you know that your measurement instrument is 100% accurate then a difference of 0.020v and above should be a point to start investigating cell imblances if you ask me.

 

[999zip999]

So the worry point is after 3.8v and 4.0 diference or .20v ? And not .020v ?

 

[donorcycle]

EBJ, I'll try to post a pick of the graph tonight if I can find it and figure out how to get it on here. I'm still new at the Hyperion

I can tell you the gap started going out of control at 3.66V but was decent up to that point. As you know, the "Cell gap" box on the Hyperion GUI turns red at .031, so I was using that as a "runaway" point of measurement. The cells were getting a little warm. My other packs fared better finishing at 3.5V with cell gaps between .040-.056 and the "runaway" point was lower at around 3.64 V if I remember correctly.

I did notice too, the charger had to work extra hard to get the "less good" pack balance charged after running it down to 3.5V. The charge graph looked more active than I have seen.

 

[theRealFury]

So the worry point is after 3.8v and 4.0 diference or .20v ? And not .020v ?

Sorry i think you misunderstood my post. Any more than 0.020v difference when the cells are of a voltage between 3.8v and 4.0v (so for instance if you had one cell reading 3.88v and another reading 3.93v) should be investigated but at the bottom of the discharge and at the top of the charge for instance when discharging below 3.6v you can expect cell gaps of way in excess of 0.020v and when charging beyond 4.10v you can see cells differences start to get a little wider apart as the weak cells reach full first.

Hope that explains it better?