Battery considerations - Zebralight SC600

[ZebraLight]

Re: ZebraLight SC600

Most Li-ion batteries since early 90's (including the Panasonic CGRs) have Cobalt based positive electrodes. However, the Panasonic NCRs use a Nickel based positive electrode. NCR batteries get this higher capacity from higher energy density AND lower discharge voltage. There was a news release about this around 2006/2007 but I couldn't find it right now. When we were trying to determin the cut-off voltage for the H60 in 2008, we looked at that NCR18650 spec as well as specs from all other major manufactures, and concluded that it should leave the protection circuits to set the cut-off voltages based on the cells used.

 

[kwak]

Re: ZebraLight SC600

This hobby encompasses many different people with different wants and requirements, so i can sympathise that it must be extremely difficult to the point of impossible to please everyone all the time.

As i say i have been using LiPo and LiFe cells almost daily for around 12 years now, as i have a lot of money invested in many different types of cells i tend to be careful and look after my batteries to the point of being a bit anal.

Personally i would prefer a higher cut-off as i'm 100% certain from my experience that this will lengthen the life of my cells.
Others may prefer a lower cut-off and not be concerned about replacing cells every 2 - 3 years.

End of the day it's your company, your torch so your decision on what you think the majority of your customers and potential customers want.

As it is i'm extremely disappointed in ZL anyways, in how early adopters of the SC600 like myself had confidence enough in your products to pay for a product weeks if not months before it's released, but have been left with a torch that does not accept 80% of cells currently in production, and none i've yet found above 2400mah.

So the LVC is really not enough to put me off your products alone.



Cheers
Mark

 

[fnj]

Re: ZebraLight SC600

I suppose (with some trepidation to even suggest this) that the only way to please everybody would be to allow programming your own cutoff criteria - something better than a simple voltage cutoff regardless of current level would be the really correct way to do it - the new step down algorithm appears to be an excellent step forward. We can already (in theory) control how much we charge the darn batteries (4.2 volts being arguably too darn much for long life); we ought ideally to be able to control how much we discharge them. An online readout of voltage by blink code (a la LF2XT) would be a good start. But we're never going to have everybody completely pleased. The basic LiIon technology is severely problematic anyway - can't store fully charged cells on the shelf without ruining them, give me a break.

I'm darn happy with my SC600, but it would be wrong to say the state of that darn battery inside it isn't always gnawing at me.

 

[infinus]

Re: ZebraLight SC600

Your findings do not corroborate the basic chemistry of the cells nor the findings of thousands of Lithium cell uses, manufacturers or distributors over the last 20 years or so.


To substantiate my claim with regards to discharge rates i would like to recommend (very highly) an extremely through and well carried out test here.
http://lygte-info.dk/info/Batteries18650-2011 UK.html


As i say IMO there is absolutely no doubt that 2 things dramatically reduce the usable cell life of lithium cells.

1/ Storing fully charged
2/ Taking the voltage too low

LOL.... sorry dude, no disrespect meant, but again, I have to totally disagree with you. First off, I don't know the state of the batteries used in that test, but I'm assuming they were much newer than mine. My personal observations using my hobby charger to do controlled discharges on my AW 2900's (panasonic cells) shows that at 1 amp discharge I go from 2600mah to 2200mah just by changing the cutoff from 2.7 volts to 3 volts. If I went to 3.4 volts this would be considerably larger. Like I said, my cells are fairly aged and show more voltage depression than your normal cells.

Now, lets use the data you just presented, assuming slightly newer cells. I think the draw on this light is actually like 1.5 amps or so? I'm going to use the 2 amp curves for this discussion, although that might exacerbate the problem a touch more than real life.

Looking at the 2 amp curve, and the AW 2900 cell, you can clearly see that going to 2.7 volts yielding 2700mah and going to 3.4 volts yielded somewhere around 2150mah or so. So even on the cells in that test, it was a swing of 600 mah. My cells, being older, would show an even great swing.

Once again I repeat my point. Going to 3.4 volts would severely cut run time. I'm not trying to be an a$$ or anything but I think you are going far too conservative. To each his own though. I would suggest that instead of limiting all users they keep it as is and let people make a choice on when to recharge.

 

[Colonel Sanders]

Re: ZebraLight SC600

I think the 2.7v cutoff is a good compromise for the later 2900/3100 cells. They still have some power left in them below 3.3v in my testing. I don't mean a lot, but it's usable power.

If someone is REALLY concerned about cell life then do as I do, leave the house with a fresh charged cell everyday. I have yet to exhaust an 18650 in one day in a Zebralight during normal usage. You could also charge them to only 4v and then pull them every morning for another 4v charged battery. Maybe a little trouble but if cell life is really your concern, why discharge all the way to even 3.3-3.5v?

Myself, when the cells get old, they get replaced by whatever is the latest and greatest. That's part of the fun.

 

[kwak]

Re: ZebraLight SC600

I suppose (with some trepidation to even suggest this) that the only way to please everybody would be to allow programming your own cutoff criteria - something better than a simple voltage cutoff regardless of current level would be the really correct way to do it - the new step down algorithm appears to be an excellent step forward. We can already (in theory) control how much we charge the darn batteries (4.2 volts being arguably too darn much for long life); we ought ideally to be able to control how much we discharge them. An online readout of voltage by blink code (a la LF2XT) would be a good start. But we're never going to have everybody completely pleased. The basic LiIon technology is severely problematic anyway - can't store fully charged cells on the shelf without ruining them, give me a break.

I'm darn happy with my SC600, but it would be wrong to say the state of that darn battery inside it isn't always gnawing at me.

I agree this would be fantastic, it's what the vast majority of modern electronic speed controllers do, many of these used by kids and teenagers that in my experience pick up on Lithium technology and it's safety aspects a lot quicker than many old farts like me.


I should point out that we're talking about percentages of cell life here over 100+ charges.

If you rarely use your torch then we're probably talking 1 to years of use before you start to see the effect of low cut offs and storage at fully charged.

LOL.... sorry dude, no disrespect meant, but again, I have to totally disagree with you. First off, I don't know the state of the batteries used in that test, but I'm assuming they were much newer than mine. My personal observations using my hobby charger to do controlled discharges on my AW 2900's (panasonic cells) shows that at 1 amp discharge I go from 2600mah to 2200mah just by changing the cutoff from 2.7 volts to 3 volts. If I went to 3.4 volts this would be considerably larger. Like I said, my cells are fairly aged and show more voltage depression than your normal cells.

Now, lets use the data you just presented, assuming slightly newer cells. I think the draw on this light is actually like 1.5 amps or so? I'm going to use the 2 amp curves for this discussion, although that might exacerbate the problem a touch more than real life.

Looking at the 2 amp curve, and the AW 2900 cell, you can clearly see that going to 2.7 volts yielding 2700mah and going to 3.4 volts yielded somewhere around 2150mah or so. So even on the cells in that test, it was a swing of 600 mah. My cells, being older, would show an even great swing.

Once again I repeat my point. Going to 3.4 volts would severely cut run time. I'm not trying to be an a$$ or anything but I think you are going far too conservative. To each his own though. I would suggest that instead of limiting all users they keep it as is and let people make a choice on when to recharge.

I have to be honest as say that i trust the data in that test and my own experiences rather than your unsubstantiated claims.


It's an absolute unequivocal fact that Lithium chemistry cells discharge in a non linear way and that the lower the voltage the sharper the drop.

So either you are wrong or you have some very special cells that fly in the face of all known physics and chemistry

Your also using capacity when we are talking about time.

On high there is no way that you'd get more than say 1 min from 3.3v to 2.7v.
As i get just over 1 hour on a new cell in my SC600 that's less than 1/60th difference in total run time, in other words inconsequential.

 

[psychbeat]

2.7 is perfect IMHO

These lights are about performance and I'd probably upgrade my 18650 before its seriously degraded.

I mean a new unprotected cell is pretty cheap in the long term.

A new cell every year or two is no big deal.

A higher cutoff is equal to lowering the rev limit on a performance car.

If u are worried about it then don't redline it.
The 5min cut is bad enough- I personally don't think we need more electronic nannies in our flashlights.
This is why I'll probably just stick to dumb 2-3 mode p60 lights.
I do like the zebralight UI and have 2 of their headlamps.

Anyways , in an emergency I'd rather run my batt down to 2.7 then be in the dark knowing I'd get more cycles.

 

[kwak]

Re: ZebraLight SC600

2.7 is perfect IMHO

These lights are about performance and I'd probably upgrade my 18650 before its seriously degraded.

I mean a new unprotected cell is pretty cheap in the long term.

A new cell every year or two is no big deal.

A higher cutoff is equal to lowering the rev limit on a performance car.

If u are worried about it then don't redline it.
The 5min cut is bad enough- I personally don't think we need more electronic nannies in our flashlights.
This is why I'll probably just stick to dumb 2-3 mode p60 lights.
I do like the zebralight UI and have 2 of their headlamps.

Anyways , in an emergency I'd rather run my batt down to 2.7 then be in the dark knowing I'd get more cycles.

At the risk of getting tired of my own voice here.

We are talking about seconds difference in total run time an extremely small percentage, yet a large increase in battery life.


If it were to reduce the run time on high by say 10 mins, i'd agree 100%.
In practice though we are talking no noticeable difference by the VAST majority of torch users.

 

[psychbeat]

How much runtime on low then?

 

[Derek Dean]

Re: ZebraLight SC600

On high there is no way that you'd get more than say 1 min from 3.3v to 2.7v.
As i get just over 1 hour on a new cell in my SC600 that's less than 1/60th difference in total run time, in other words inconsequential.

As a new owner of an SC600 I've been following this discussion with interest.

I do feel Zebralight has made the right decision here with their choice of the low voltage cutoff value. It IS hard to please everybody, and in this matter they chose to go with longer runtime, and for the vast majority of users I'll bet that is a good choice.

kwak, with the new step regulation it appears the light will step down to a lower level before it gets to the cutoff level, so instead of 1 minute of extra time we could be talking about a LOT more, and in some situations that might not be inconsequential.

In my way of thinking it's better to err on the side of the safety of the user by providing maximum runtime rather than being overly concerned about the long term effects on the cell. Just my 2 cents.

 

[infinus]

Re: ZebraLight SC600

I have to be honest as say that i trust the data in that test and my own experiences rather than your unsubstantiated claims.


It's an absolute unequivocal fact that Lithium chemistry cells discharge in a non linear way and that the lower the voltage the sharper the drop.

So either you are wrong or you have some very special cells that fly in the face of all known physics and chemistry

Your also using capacity when we are talking about time.

On high there is no way that you'd get more than say 1 min from 3.3v to 2.7v.
As i get just over 1 hour on a new cell in my SC600 that's less than 1/60th difference in total run time, in other words inconsequential.

Dude, I used your own data and showed a difference of near or over 600mah? That's a "few seconds?".

Look at your own charts, 3.4 volts to 2.7 volts provides substantial energy. Much more than a few seconds. Another fact you are ignoring is that the higher the current draw, the more linear that curve gets.

 

[psychbeat]

AND- on the lower modes with less voltage sag its got to be more than a few seconds.

 

[infinus]

Re: ZebraLight SC600

Here are some stats, hot off the charger:

I set this up to be similar to real life results with the ZL. I used a discharge of 1.5 amps which is a good approximation of high. This is with my slightly older and heavily used 18650 cells.

Starting from 4.16 volts, I set a termination point of 3.4 volts. After termination the voltage rebounded to 3.6 volts.

So 58 minutes of run time and 1465 mah burned.

Second, I kept the same battery, and restarted the discharge with a new termination of 3 volts. After termination the voltage rebounded to 3.34 volts.

Another 40 minutes of run time, another 1000 mah burned.

Going to 2.7 volts would yield, from my experience, another 150 or 200 mah, so probably another 4-10 minutes.

I know you just want to dismiss my proof as "unsubstantiated claims" but I much prefer real world results vs ideal tests done using brand new cells.

No thanks to giving up 40+ minutes of run time.

 

[infinus]

Re: ZebraLight SC600

Last post before bed.....

Important data sheets for NEWER panasonic cells:
http://industrial.panasonic.com/www-data/pdf/ACA4000/ACA4000PE1.pdf - shows that normal termination should be at 2.5 volts per cell
http://www.panasonic.com/industrial/includes/pdf/ncr18650_datasheet.pdf - shows discharge curves going to 2.5 volts, which is how they get 2900 mah ratings
http://industrial.panasonic.com/www-data/pdf/ACA4000/ACA4000PE4.pdf - shows low voltage over discharge protection should be at 2.3 volts

Again, I'm really not trying to come across as a jerk. I'm sure you have tons of experience with Li-Ion as you said, but it appears its mostly with the older cells that didn't like being discharged so much as you've said. New Panasonic cells are clearly documented as being ok to run down to 2.5 volts. This is how they've achieved higher capacities. My tests and the Panasonic curves all show that a 3.4 volt termination leaves a lot of energy left in the cell. I see no reason why Zebralight should prevent people from utilizing this. We aren't always in ideal situations where using half the power of a cell is a good option. I prefer to have all the run time I can get in case I need it. Otherwise, swap your cells out! Why are you running them down until the light goes out anyways if you are so concerned about capacity.

 

[DM51]

Re: ZebraLight SC600

A cut-off at 2.5V can only lead to problems such as shorter cycle life if the application using it is a low-amperage one, where little or no no voltage rebound occurs once the load is removed, AND where the cell is not recharged promptly when empty. In one situation or the other, the cell should be undamaged, but if both coincide together, cell life will be affected.

The problem is not draining the cell to 2.5V - it is leaving it there instead of recharging it promptly.

 

[samgab]

Re: ZebraLight SC600

Oh, can I just add: it's great to have the company director popping in here to add to the discussion: Thanks for that George. And I really look forward to the release of the Q50!

Also: infinius and psychbeat, you're exactly right.

 

[fnj]

Re: ZebraLight SC600

A cut-off at 2.5V can only lead to problems such as shorter cycle life if the application using it is a low-amperage one, where little or no no voltage rebound occurs once the load is removed, AND where the cell is not recharged promptly when empty. In one situation or the other, the cell should be undamaged, but if both coincide together, cell life will be affected.

The problem is not draining the cell to 2.5V - it is leaving it there instead of recharging it promptly.

And when it takes 80 days to drop to 2.5 v on low, the chance of it NOT lingering at low voltage is nil.

 

[psychbeat]

I'd bet most people's dinner cost more than an unprotected 3100mah Panasonic.

My AW2600s have almost 2years of being drained to cutoff or near it and still seem to hold ~4.18 a day or twi after charging. They get used mostly in my lower 2.8a light.
I use the 2900 and 3100 in my 4.2a.
Although the AWs still run the 4.2 fine.

My point being even running em hard u can probably get a couple of years out of yer batts.

It's unlikely u will need 80days but IF u do, replacing an 18650 isn't that big of a deal. My bet is the cell would be fine after tho.

 

[infinus]

Re: ZebraLight SC600

Yeah, my AW 2900's have only been run in high drain lights (ie: M3C4 XML, Thrunite Catapult, Zebralight SC600, and Scorpion V2). They don't always get run down to cutoff, but it does happen. They've been through at least a hundred or more cycles, and still test out to 2650mah (drained to 3 volts @ 500 mah). At higher loads the test a tad less. I think new most of the 2900's test out around 2750 or 2800. These things definitely take high drain low voltage pretty well. I do think these Panasonic cells exhibit more voltage sag than other cells I have in my collection, and I'm not convinced it's simply because they've been heavily used. I think it's part of this chemistry to sag a bit more under high load, but I'd have to do some testing to verify that, and I'm not buying new cells at this moment.

 

[DM51]

Re: ZebraLight SC600

42 posts (#18 through #59 above) have been moved here from the main SC600 thread.