Honda Insight Forum banner

LTO Testing

24K views 277 replies 10 participants last post by  jime 
#1 · (Edited)
Lv3 = Create your own new profile from scratch .. when modifying existing is not enough .. An extension/expansion of Lv2 just taken a bit further , to allow one to escape/change some of the 'safety' settings that might be part of Lv2 above , and are more limiting than you might want to be.
Completed Lv3 style PL8 LTO preset.
My thanks to those who provided me with programming assistance.

To those who are interested .. Unzip the attached file.

Any SoC to start .. Charge .. Discharge .. Cycles .. Data Log .. etc.

The PL8 firmware has to be updated to at least no older than version 3.35 .. this is available as a button in the free (with PL8) Charge Control Software , with Internet access.

Discharge should be done using regenerative discharge mode .. For lower voltage .. like one cell at a time , it has to be done in regenerative discharge mode.

You should use the cell level voltage node connection .. in additional to the main power banana plug connection .. even for just 1 cell .. But you can do 1,2,3,4,5,6,7, or 8 series set of cells at a time .. allows for cell level balancing through the 8 cell node.

IF the 8 being tested are part of a larger series string .. like my 48s LiFePO4 pack .. and IF you are using multiple PL8s at once .. Like I use 3 .. then each PL8 must have it's own (don't share one source among multiple PL8s) Source Battery to put regenerative discharge energy into and pull charge energy out of.
 

Attachments

See less See more
#2 ·
Completed Lv3 style PL8 LTO preset.
My thanks to those who provided me with programming assistance.
Great news. I for one am grateful for your work. I think this may turn out to quite valuable.

Just as soon as I can get the bugs worked out of my installation, I plan to do some testing on my unused 4 packs, so this is coming at a really good time. My mancave is going to be very busy on those winter snow days:D

I think you might want to split this off as a separate thread where you, I and probably others will have some valuable posts. I'll set that up if you want me to do so.
 
#3 ·
Welcome .. Happy to help :D

An 'LTO Testing' like thread .. is probably not a bad idea.
If you wish to split it off as such .. so such discussion doesn't clutter up this LTO conversion too much.

I do already have lots of the parts (batteries, cell fins, VIP, etc) , effort , etc .. for my already in progress A123 LiFePO4 ~8kwh v2-Li-IMA .. that will take me well into 2019 .. Soo , I probably won't be doing allot with LTO for a while yet .. I might tinker a little , but not large amounts in the near future.
 
#4 · (Edited)
@Ian

Since the "design" discussion has died down some, I'm getting motivated to start doing some testing of my 4 remaining LTO packs.

I have a question. If I use my PL8 to cycle my packs, 8 cells at a time, will I be able to leave the voltage of ALL cells at some selected top storage level? Put another way, will I be allowed to specify and achieve an accurate and balanced finishing voltage for all cells, even though the pack is done in segments?

Obviously, I need to buy some more harness pieces from Revlectrix:)
 
#5 · (Edited)
Yes you can.

The PL8 has the ability to do either top balancing or bottom balancing.

You make the physical connection with the balancing wires.

You make the software selection in the charge control software.

If you are going to do lots of the same .. I would recommend just setting up the profile how you want it .. to do what you want it to do .. and then it's just saved that way in the PL8 itself .. so even if later your PC isn't connected it just runs that profile the same way you have it set up to do before .. The PC connection is only needed to tweak the profiles, or to log the data.

----

One note about a little balancing quirk on the PL8.

The PL8 does balancing by bypassing the amps one cell sees .. up to either the balancing Amp rate you set .. or .. up to about ~99% .. works fine most of the time .. but over the numerous times I've used mine I have found a quick in this method.

If one if trying to bottom balance .. you are taking energy out of the cells ... if the balancing Amp rate is set for say 1A .. well it will then bypass up to that 1A setting or .. about ~99% of that .. but would still be pulling a small minimum around ~10mA from that lowest cell (~1% of ~1A) .. If I were to start the bottom balance discharge from a good SoC away from the bottom .. there is plenty of time for the ~1A balancing rate to correct any balance issue before reaching the bottom .. or if there is only a small difference in balance to fix , also no issue .. but .. if start the bottom balancing very near the bottom and there is also a very large difference between cells .. there might not be enough 'head room' left and it will get some balancing , but might not have enough 'room' to do 100% of the balancing before that minimum ~1% rate pulls the one lowest cell down to the cut off , and the PL8 just times out before being 100% done.

Conceptual example .. 1A on highest voltage cell and 10mA on lowest voltage cell .. is a 990mA rate of balance correction .. A ~990mAh out of balance difference might be fixed in a minimum of ~1Hour , to move those two cells closer to each other .. 3A on highest voltage cell and 30mA on lowest voltage cell .. is a ~2.97A rate of balance correction .. a ~2.97Ah out of balance difference might be fixed in a minimum of about ~1Hour .. but in both cases, how far apart are the highest and lowest cells .. and how far is the lowest cell from the bottom.
 
#6 ·
The PL8 firmware has to be updated to at least no older than version 3.35 .. this is available as a button in the free [with PL8] Charge Control Software , with Internet access.
More accurately .. I tested it on mine with the 3.35 version of the firmware .. screen shot attached .. I do not know the differences between firmware versions , I don't know how much it matters (if at all).
 

Attachments

#7 · (Edited)
Just a quick FYI for anyone ever wishing to test LTO cells. The iCharger 206b and probably similar but higher-end models, at minimum, plus clones like the Reaktor or 'Charsoon', would be able to test LTOs. You can adjust voltage thresholds to fit the LTO's profile in the Lithium cycling program. In other words, the voltage thresholds for the standalone chemistry profiles on these chargers have discharge and charge voltage limits that don't match the LTO. But the cycling program allows you to adjust voltage to a much finer level. For instance, the 'LiFe' cycling program can discharge down to 2V, which I think is low enough for even the high energy Scib cells. I know that threshold works for the high power Scibs (those are near fully discharged at about 2.3V at 20 amps)... And then you can set the charge high voltage cutoff to anywhere below the max voltage for the particular chemistry profile you're using, such as I think 3.4V for the LiFe - so you could set it to say 2.7V for the LTOs...

hmm, on the other hand, 20 amps is the current limit for the 300W iCharger 206b and similar, and that might not be high enough for a large, 20 amp-hour cell... You might want to go for one of the higher power models, such as the 1000-Watter...
 
#20 ·
The iCharger 206b and probably similar but higher-end models, at minimum, plus clones like the Reaktor or 'Charsoon', would be able to test LTOs.
Looked at some ProgressiveRC chargers and you are probably right on those. Could save a bunch of money if one could be found refurbished.

I looked some at the 206b and the 306b. The 306b allows for 30A discharge in a regenerative mode, so at a quick glance, that charger looks like it would be useful. Just a first impression. Didn't look at the others.

Ian has developed an adjustable "Generic LTO profile" for the PL8 and that seems tailor made of testing those cells, but no actual experience yet.
 
#8 ·
I think most LTO testers are going to want to test multiple cells at once with a final precision balanced charge to get all the cells top balanced. (Multiple cells because there are so darn many in a pack.) The PL8 will in theory do that, but non of the stock profiles were anywhere near the operating range of the cells. Plus high current testing.

Plus, it is desirable to get out close to the full range of the cells(1.5-2.7V), so that there is a meaningful measurement of capacity. And, don't forget graphing capability so that there is a historic record.

There may be other chargers which will do all that. I'm just not familiar with them well enough to know, plus I had a good adviser;)

Maybe these aren't the essential testing requirements, but seem right to me.
 
#9 ·
Plus, it is desirable to get out close to the full range of the cells(1.5-2.7V), so that there is a meaningful measurement of capacity....
Looking at a graph I have of the 20 amp-hour Scibs, looks like voltage falls off a cliff at about 2V at 0.2 to 1C rate, maybe just a hair lower at a 3C rate. So whether you go to 2V or 1.5V shouldn't matter...
 
#10 ·
Although I'm sure other testers can be used .. one should just check because LTO is not as common as other types of batteries.

----
Because some people have expressed difficulty .. I'll try a bit more step by step approach.

Reminder .. thanks to those who helped me with the programming of this custom LTO battery preset .. this was not all me by myself .. my solo effort was the previous more 'lv2' approach of just modifying one of the existing PL8 presets .. and that worked , but had it's limitations .. this 'lv3' custom LTO preset works far better.

The Insight Central Site won't let me upload *.ps8 file types .. I previously tried to get around it by zipping the file first .. which will require the other person to unzip before using .. but I also uploaded the unzipped *.ps8 LTO battery preset to my google drive .. can be directly down loaded:
LTO Preset Link

Although I don't know for certain if firmware version 3.35 is needed .. For those having trouble with getting it .. I think this will work as an executable .. I'm not a software guru , nor have I done testing on every conceivable combination of computer hardware and software .. so, 'use at your own risk' .. also uploaded that to my google drive as well:
Firmware 3.35 Link

I also recorded a short desktop screen capture video walk through as well .. Uploaded to my YouTube page.

 
#11 · (Edited)
Some graphs of LTO testing (mine and others):
There is not much left to be had discharging bellow 1.80v per cell .. and not much to be gained for charging above 2.60v per cell.
 

Attachments

#21 ·
I think those last two graphs are of an LTO-LMO cell, so they may not be representative of what to expect from the 20Ah high-energy Scib... Actually, looking at the graph you made they are not: the 1C discharge for the LTO-LMO tanks at about 2.3V, whereas your graph does the same at about 2.05V...

Here's some Toshiba graphs of a pack made up of what I think are the same cells you guys are using. If you eye-ball the voltages and divide by 12 they look to be about the same as what you got in your own graph. For instance, 1C discharge, ~24V/12 cells = 2V per cell...

 
#12 · (Edited)
Voila!!!

Ian, I was able to load up your LTO profile by following the directions in your video. For some reason, I could never get through the process with Firefox browser, but had success with Microsoft Edge.

Now, just need to pick your brain a little bit more - probably won't hurt much;)

I have a question on driving the balance port. Currently the PL8 ships with something called a "Single Port Safe Parallel adapter (SPA) and FMA Wiring mode vs. XH/EH Wiring mode" adapter. There is a nice 9 wire (8 white, 1 black) harness with the adapter. I'm wondering if I should just "butcher" this harness to get 9 balance wires, OR if I should try to use the adapter. Any guiding words??

Have to shop for a deep cycle golf cart battery on Monday. Can't continue to use one of my packs for these experiments.
 
#13 · (Edited)
:D

I have a question on driving the balance port. Currently the PL8 ships with something called a "Single Port Safe Parallel adapter (SPA) and FMA Wiring mode vs. XH/EH Wiring mode" adapter. There is a nice 9 wire (8 white, 1 black) harness with the adapter. I'm wondering if I should just "butcher" this harness to get 9 balance wires, OR if I should try to use the adapter. Any guiding words??
The end that plugs into the PL8 itself is the most useful part .. For my own use .. I just cut the ends of the wire and put whatever connector I want onto it .. ring terminals , alligator clip , anderson connector .. etc .. whatever you want.

I used single cell for testing .. so I don't have the car module you have to look at .. but and ideal .. would probably be cutting and splicing together the wiring harness that the OEM car already used to get to each individual cell , for the OEM BMS .. or if that isn't readily available .. however you want to reach the + and - of each cell , up to 8 can be done in series at one time.

You can select which balance wiring approach you prefer in the PC PL8 charge control software under the 'options' tab .. under 'node'.

There select either 'FMA' mode or 'XH' mode .. examples of each are shown in the PL8 manual .. copy attached .. pages 61 to 71.

Although the PL8 itself will let you select up to 3A to the balancing wires .. I only set mine to use up to 1A with the 22Ga wire that came with my PL8s .. maybe I'm just a bit cautious .. but I'd probably make a slightly heavier gauge wire version if I wanted to run continuously at the full 3A through the balancing port.

Have to shop for a deep cycle golf cart battery on Monday
I do like the regenerative discharge feature .. but it isn't a 100% efficient cycle .. so be sure you have enough room in whatever source battery you use .. for test battery discharge wh to charge into that source battery .. and for test battery charge wh to come from discharging out of that source battery .. if you are going to do lots of testing .. it can be useful to have a CV power supply to trickle replace what is lost in the less than 100% round trip .. just leave room/capacity at top and bottom for both charging and discharging .. The PL8 does have 'supply' tab , so you can setup/configure whatever source battery details you like.
 

Attachments

#17 · (Edited)
Still in Trouble

Well, I spoke too soon. I thought I had downloaded the profile, but no. It did load something to slot 24, but it has no "LTO" in the name and I can't load it to PL8.

I'm getting really frustrated, so gotta give this a break for sanity sake. It probably has to do with some "protection" I have set up on this Win10 laptop, but I don't know where to look and am getting pretty down with the trying. Gotta give it a break.:(

Gonna go off and build the Lee-Hart Bat-Bridge. I think I have the parts for that.
 
#18 ·
Well, I spoke too soon. I thought I had downloaded the profile, but no. It did load something to slot 24, but it has no "LTO" in the name and I can't load it to PL8.

I'm getting really frustrated, so gotta give this a break for sanity sake. It probably has to do with some "protection" I have set up on this Win10 laptop, but I don't know where to look and am getting pretty down with the trying. Gotta give it a break.:(

Gonna go off and build the Lee-Hart Bat-Bridge. I think I have the parts for that.
Sorry to hear about your trouble .. It was all pretty smooth for me.

Worst case .. I'll update your PL8 from my laptop at the next InsightFest.
 
#19 ·
Think I have the kinks ironed out. Problem was Norton Internet Security Anti-Virus which kept deleting the firmware upgrade to LTO. Finally caught a small message in bottom right which said basically the file was "dangerous" and was automatically deleted. Disabling anti-virus briefly allowed the LTO firmware to be downloaded.

Thanks Ian for your considerable help and patience:)
 
#23 · (Edited)
Thanks for adds eq1
:D

---

I didn't get a full battery pack from a salvage yard .. like those building and LTO IMA pack .. I only bought 1 sample cell for some basic testing and to use for configuring the PL8 preset for LTO use .. so , I don't know how representative that one cell will be to all others other people use .. but it was a starting point .. The PL8 preset was left open enough , so we can tweak it , as we get more real world testing in from people.
 
#24 ·
^ If the cells people are getting in the Fit packs are anything like the new, smaller LTO cells I got, then there should be minuscule variation from cell to cell. Or put another way, any variation you see, or others see when testing Fit LTO cells, is likely a function of their usage history, age, etc... I've tested something like 70 cells (1 extensively, the rest just a couple cycles each, plus some more cycles on some of them at a later date), and there's no variation that rises above the 'noise' in my testing methods... The consistency is truly phenomenal - particularly relative to the hodge-podge one sees when testing used Insight sticks and cells...
 
#26 ·
Good to hear .. Truly is a good thing.
:D

---

The only 'caution' comment I feel compelled to post .. just so people are aware of such a thing.

Even if all the cells are perfectly identical .. all the way down to the each individual atom .. even if we assume that ideal case (as unlikely as it is they are that perfectly matched).

And .. LTO are more tolerant of cold weather usage .. they are diminished less from being cold .. which is also a good thing for this application.

..
However
..

During use in the car .. there will be a thermal gradient .. in use in the car .. not all the cells will be exactly the same temperature all the time .. no matter how match the cells themselves are .. that thermal variation is just a reality of the heat flow path.

And although they do do 'better' in the cold .. they are not 100%'immune' to the cold .. For those in places (like here in RI) that have night time winter lows into the single digit F temperatures .. even the best LTO cell at like 5F will not have identical properties and perform the exactly the same as when it is 50F , etc.

--
Because of that .. even if we knew all the cells were 100% perfectly identical to start with .. I would still recommend:

#1> Safety margin top and bottom
Avoiding using the very top 10% .. avoiding using the very bottom 10%.


#2> Do at least a minimum of a once a year check .. top and bottom .. do they all reach the top exactly the same .. and do they all reach the bottom exactly the same.

This can be done more often , and even faster with things like the half pack compare LED .. or with a very high MTBF cell level BMS .. but a simple volt meter can do it also.

If they all reach the top exactly the same .. are top balanced .. but get a little out toward the bottom .. you might be ok with it .. but that might mean you want to go to increase the distance you keep from the bottom .. like from 10% to 15% .. etc .. and vise versa .. if they all each the bottom exactly the same .. are bottom balanced .. but get a little out toward the top .. you might be ok with it .. but that might mean you want to go to increase the distance you keep from the top .. like from 10% to 15% .. etc.

---

I admit .. maybe that is just me .. and maybe others are comfortable not doing such things .. and maybe because of it , they only get 4,000 cycles instead of 20,000 .. that 4,000 would still be over 10 years of use .. sooo .. to each their own .. but like I said .. I did feel 'compelled' to at least write such.
 
#25 ·
Hopefully when I get up and running with the PL8, still some minor problems with 8s testing, I'll be able to provide some actual graphs. For now, the discussion is interesting:)
 
#28 · (Edited)
On 'minuscule variation,' I'm mostly thinking about amp-hour capacity, charge and discharge curves, and to a lessor extent IR and temperature change during charge and discharge... I didn't do much quote 'testing'; rather, just enough to get a sense of how the cells performed, the voltage profile, how consistent they were with one another. Actually, I think I started, thinking my data would be used to match, but the cells were so similar that I quickly realized any variation could easily be due to my methods and/or equipment...

On all but one cell this is what I did:

-checked voltage with multimeter, wrote it down, measured IR with Reaktor's push-button IR measurement thing,
-discharged at 1C to near empty (2.2V for my cells)
-charged to full at 2C (2.7V with typical CV to input current /10)
-discharged to 2.2V at 2C
-charged to about half full at 2C (2.5V with CV to input current /10)
-measured IR...

That's it. Loaded next cell and did the same thing...

I re-did this sequence on about 4 or 5 cells about 10 months later to compare...
One cell I did a bunch of cycles, about 26 or so, at different rates of charge and discharge to develop a map of how voltage changes at different rates. The changes were pretty consistent, predictable, and temperature hardly changed at all, all the way up to 6.9C, the most I could do... Keep in mind these are only nominally 2.9Ah cells - so it's a lot easier to deal with than what Fit LTO'ers would have to do, having 20Ah cells... Also, they're 'high power' cells and apparently have a slightly different voltage profile than the 'high energy' which Fit LTOs are. They have a nominal voltage of 2.4V vs. the Fit LTO 2.3V. The charge termination looks the same/similar, but stuff toward the bottom doesn't. These are near empty at about 2.2-2.3V; looks like the Fits are like 2-2.1V...

I used the 300W Reaktor, original iCharger 206b clone (not the later model that changed and lost functionality). Charge and discharge termination voltages can be adjusted appropriately in the lithium cycle programs, for example, the lithium iron cycle program allows a range of 2V to something like 3.2V - and I only need 2.2V to 2.7 or 2.8V.

edit: One thing that does surprise me, glancing at some of my data, is that self discharge seems higher than what I would have expected. The 4 cells I re-tested after sitting 8 1/2 months - at the half-charged state - lost about 28% of their charge. I don't know, I don't have much experience with lithium and self discharge, just seems a bit on the high side...
 
#29 ·
Thank you eq1. I appreciate the account, the equipment notes, and the results summary.

I have been trying to test with the Powerlabs PL8, but it has been a struggle. I was motivated to try that because it will do 8 cells at once and do balance charging at the same time. Unfortunately, there was initially no LTO profile so Ian got some assistance and created one. It works well on charge, and balances, but it seems to stop short on discharge. The PL8 has some parameter called "fuel" and the special LTO fuel wizard doesn't seem to work properly.

Anyway, the second hit is that the PL8 quit so now I have to return it for service.

Thanks again for the summary.
 
#31 ·
The PL8 has some parameter called "fuel" and the special LTO fuel wizard doesn't seem to work properly.
Sorry you continue to have trouble .. I had no trouble discharging the one cell I did .. what did it do ? .. or .. what error did it give ?

The PL8 fuel table is just a set of correlations .. Voltage 1 = SoC1 .. Voltage 2 = Soc2... etc .. it is a away for the PL8 to be able to give you a 'guess' at what the cell being tested is as far as % SoC .. sense you could be starting a charge or discharge of a cell that is at any SoC point .. You can use the default values already in the preset .. or you can manually set them .. or (usually) you can use the 'generate fuel table' function.. but even if (for some reason) you can't use the auto generate fuel table function .. you can still use the default values , or change them manually yourself as you see fit .. however .. these are just guesses for the SoC% .. I've never myself had any of my PL8's stop a discharge based on the fuel table values .. it stops based on other settings .. Discharge voltage cut off .. time cut off .. if supply battery unable to absorbs the energy from the discharge battery .. if the 8 cell balance voltage detection is not attached .. etc.

---

edit: One thing that does surprise me, glancing at some of my data, is that self discharge seems higher than what I would have expected. The 4 cells I re-tested after sitting 8 1/2 months - at the half-charged state - lost about 28% of their charge. I don't know, I don't have much experience with lithium and self discharge, just seems a bit on the high side...
That seems very odd to me .. almost seems to suggest some level of tiny internal short in the cell.

Most Lithium batteries I've tested are under 5% Self Discharge per year themselves .. 28% lost in just over 8 months , seems unusually high.

Unless there is a BMS attached to them .. and if that BMS is constantly pulling some mA from them .. then it might not be the battery cell itself.

28% of 2.9Ah = ~812mAh in ~8.5 Months = ~95mAh per month / ~30days = ~3mAh per day / 24 hrs = ~0.132mA leakage rate.
 
#30 ·
I don't know for sure, but seems like this kit might work: https://hobbyking.com/en_us/turnigy-reaktor-1000w-30a-balance-charger.html?___store=en_us

30 amp limit might be a bit low, but you'd be able to do basic cycling stuff. Can do up to 10 cells in balance mode, 12 in non-balance for LiFe (which voltage is more similar to LTOs than other lithium cells, so that would apply to LTOs)... I Imagine you'd be able to adjust the voltage range in the cycling mode with this 1000W charger like it is with the older 300W...
 
#34 ·
I studied that unit a bit after you mentioned it. It looks interesting, but the balance spec is pretty loose for the idea of baselining and running the LTO cells without a BCM.

The Turnigy has a spec of <35mV unbalance.

The PL8 has a spec of <78 microvolt.

If one is to run the pack without BCM, it is highly desirable to start with all cells showing the same voltage to the millivolt. That makes monitoring much simpler. For my current pack I initially balanced the cells manually.
 
#32 · (Edited)
^ It'd be 28% of about 1500mAh since they were only half charged. Nothing attached to them, roughly same values on all 4 cells I looked at so doesn't appear anomalous. I did notice that the charge and discharge totals were about 50-100mAh lower relative to the charge and discharge totals I got before they sat. Maybe there's some 'shake-out' that happens when the cells are new and then cycled a bit and/or after sitting for months. My recollection is that most lithium cells lose more capacity in the first bit of use or whatever, relative to later use/cycles...

Later...
Here are actual numbers (average of 3 cells):
D0, C1, D2, C1.5
NA, 2967, 2934, 1704 (before)
1248, 2908, 2861, 1557 (after)

So based on the initial charge and discharge I should expect to pull out 2934/2967=0.989 from the C1.5 charge, which equals 1704 X 0.989=1685mAh, that is, if self discharge were zero. After sitting 9 months I pull out 1248mAh, so the cells lost 1685 minus 1248=437mAh, and 437/1685=26%... You can see how values for subsequent charge, discharge and half charge are all lower than they were initially - the charge is about 60mAh lower, the discharge is 73mAh lower, and the half charge is 147 mAh lower... So, unless there's something funky going on with my equipment, the cells are different after the sitting than they were before it, and maybe that factors into the self discharge results I got. But it'd still be a high value based on what you say - 5% vs. this, what?, at minimum maybe 17-18%?...

Not sure what to think, it's possible my charger is acting up in some way and not reproducing the operations exactly as it did before...
 
#35 ·
Not sure what to think, it's possible my charger is acting up in some way and not reproducing the operations exactly as it did before...
Possible .. but you should be able to test to see if you can rule that out or not .. amp and volt meter you feel confident in trusting it's accuracy / precision .. then , compare that to the tester.

Was there a temperature difference ? .. that can also effect things.

Was there a rate of charge or rate of discharge difference ? .. that can also effect things.

I am not certain .. maybe it is the specific Li you are working with .. are in some way different in that way from the other Li types I've had experience with myself.

Also .. might be worth mentioning .. as you touched on .. battery degradation is not the same as self discharge .. self discharge is stored energy lost, but capacity is still the same .. although both might combine / mix some in the results you get.

Although I do agree with having also seen some 'shake out' phase in the beginning .. that should level off pretty quick .. although high SoC and/or high temperatures can accelerate the yearly rate of degradation.

I've done a little bit of arithmetic to try to determine the appropriateness of the supply battery.
That is a good place to start .. but .. you can do better than just guess .. the PL8 can display and can log the supply batteries volts / amps / etc .. so you can see what the supply battery voltage was and how it changed during the time you were feeding test battery discharge energy into it.

Maybe I get another and hook in series to have a 24V supply. OR maybe I just use one of the "off" LTO packs to test another LTO pack???? That actually seems simpler and more appropriate.)
Either is an option .. just let the PL8 know the new supply battery information .. top , bottom voltage , etc.

If the cell charge/discharge cycle starts with the supply battery full, then it should certainly be able to absorb the cell discharge after charging. (One potential problem might be the starting SOC of the 8 LTO cells.)
yup .. it isn't max capacity .. but .. available capacity.

At the time of charging the test LTO battery .. does the supply battery have enough available capacity to do it .. if yes , you're good.

At the time of discharging the test LTO battery .. does the supply battery have enough room in it to store all of that .. if yes , you're good.

If .. for example .. you take 100wh from supply then later try to put back in 300wh .. that might be an issue.

The "fuel" table question: From the manual, p55, para 3, "If you use Auto mode (Auto 1C, 2C, or 3C) for simplified charging, and your battery is anything other than a generic 15C to 20C LiPo, you probably need a custom fuel table."
yup .. that's why I did a basic / generic 'custom' fuel table in the preset for LTO already.

The one I put there might not be as refined or fine tuned as you might want .. so .. you / I / we can then change it / tweak it / fine tune it .. if there is an issue with the PL8 doing it for you/us automatically .. then we just enter the values manually ourselves to fine tune it.

I've attached a screen shot of the 'default' / 'generic' .. LTO fuel table that I already included into the LTO preset .. it is just used (before you reach top or bottom) for the PL8 to try and guess what the % SoC is .. but the top point to stop charging and the bottom point to stop discharging are different values .. charging under the charge tab , discharging under the discharge tab.

One more mystery, I don't understand at all why the unit would need to time out, which I did see at least once. Maybe that is some sort of built in protection for small capacity cells, but it is going to be very difficult to work around the unit timing out, regardless of the designer's technical motivations.
It is just another optional feature .. it to can be adjusted.

If someone .. didn't want to 100% fuly charge a battery .. but only wanted to put some known amount of Ah into a battery .. or pull some known amount of Ah out .. you can set the charge amp rate .. set the charge time out point .. and poof .. controllable Ah charged .. same can be done on discharge .. Or maybe someone just wants the time out charge or discharge as one more level of safety.

It isn't hard to 'work around' .. the charge time out allows you to go out to 16 hours for 1 charge of a battery .. ie .. if you haven't finished charging the battery after 16 hours .. just stop .. but it can also be set down to lower values .. if you only want 2 hours , ok .. etc.

It isn't a maximum 'test' time .. it is a maximum for that one step .. ie what is the longest 1 charge should take ? ... or .. what is the longest 1 discharge should take.

At up to 40Amp Charge Rate ( per PL8 ) .. for up to 16 hours .. that's 640 Ah .. what crazy big battery are you planning to put in your car ?
;)

For example .. you could set it for 16 hours for charge .. then 1 hour rest .. then 12 hours for discharge and to repeat for 20 cycles :

Ln1 = up to 16 hours at up to 40Amps charging
Ln2 = 1 hour rest
Ln3 = up to 12 hours at 40 Amps discharge
Ln4 = 1 hour rest
Ln5 = Repeat/Loop to Ln1 until 20.

1 such full cycle could take up to 16 + 1 + 12 + 1 = 30 Hours x 20 such cycles = a maximum of 600 hours of continuous testing / operation .. @ 24 hours per day = 25 days .. at which point , you can just save the data and start another 20 cycle run .. of course each step can be done in less time .. it can be done in 5 minutes instead of 16 hours .. but the time out is a configurable option to set a maximum.
 

Attachments

#33 · (Edited)
I am using a deep cycle 12V Marine/RV battery as a supply. I suppose that the battery at its SOC cannot absorb the discharge energy of the 8 LTO cells, or that the supply starting SOC is inappropriate.

I've done a little bit of arithmetic to try to determine the appropriateness of the supply battery. If a few reasonable assumptions are made, the the battery appears to be capable of absorbing more than 4 time the energy released by 8 approx. 20Ah cells from full to empty. Starting point might be an issue.

A deep cycle RV 12V battery normally has a Ah rating of 95-105Ah when fully charged. That translates to 1200 wh. (later edit: I was thinking 6V RV batteries. Though not stated the battery I'm using is probably in the range of 60-70 Ah, so only about 600 wh and things get quite a bit more dicey. Maybe I get another and hook in series to have a 24V supply. OR maybe I just use one of the "off" LTO packs to test another LTO pack???? That actually seems simpler and more appropriate.)

8 LTO cells have a nominal voltage of 18.4V and a rating of 20Ah. That translates to 368wh.

So it would appear that the supply battery is well within capability to absorb the discharge energy of the 8 LTO cells, particularly if inefficiencies are included. i.e. If the cell charge/discharge cycle starts with the supply battery full, then it should certainly be able to absorb the cell discharge after charging. (One potential problem might be the starting SOC of the 8 LTO cells.)

The "fuel" table question: From the manual, p55, para 3, "If you use Auto mode (Auto 1C, 2C, or 3C) for simplified charging, and your battery is anything other than a generic 15C to 20C LiPo, you probably need a custom fuel table."

And yet, the Fuel Table Wizzard, is nonsensical at line 7 for LTO batteries. I can't tell you exactly what the options are because my PL8 is dead, but it is nonsensical. As I recall, the only choices available were appropriate to stand Lithium cells.

Anyway, I'm currently dead in the water with the PL8 until I can get it fixed - if I can. Called tech support for help. They walked me through so reset attempts but concluded that the unit is broken.

One more mystery, I don't understand at all why the unit would need to time out, which I did see at least once. Maybe that is some sort of built in protection for small capacity cells, but it is going to be very difficult to work around the unit timing out, regardless of the designer's technical motivations.
 
#36 ·
Thanks Ian for the additional info. I guess one has to try some stuff to gain insight. I had not seen either of the tabs in your attached pictures so that is helpful.

I think the charge window on the RV battery is so small that it is going to be very difficult to watch. i guess RC folks like to do that because it it cheap, and probably works well with relatively smaller packs.

It'll be a while before I get back to experimenting. The PL8 is broken and has to be serviced.:(
 
#37 ·
Running Again

It is up and running again. Some successes, still some frustrations or lack of understandings.

Power supply capacity: Gave up on the 12V RV deep cycle. With the mandatory 50% capacity stop, obvious it isn't adequate and I'm not buying another. Instead, turned to using an "off" LTO pack. Tried 1/2 pack first, couldn't make that work, so wired the two halves in parallel for a 12S2P configuration. Given the "safety" considerations, and the supply overvoltage cut-off, still can't get the 8 test cells charged without doing multiple charge cycles and recharging the "supply" battery.

Though the cells display works pretty well, the balancing current display never reaches stability before the test cells reach 100% "fuel" or charge, and the test cell do reach balance, though the voltage display shows balance. (Double checked by Fluke.) According to Fluke, imbalance remains at around .008mV, highest to lowest so +/- 4 mV against average.

I'm beginning to think I misunderstood the balance specification:

Quoting:

Measurement accuracy:
Voltage resolution: 78uV (16 bit)
Voltage tolerance: +/- 6mV
Charge current: +/- 1%
Capacity added to pack: +/- 1%
Percent capacity (“Fuel”): +/- 5% (with accurate fuel table defined for battery being charged)

Maybe I'm just expecting too much and will have to manually balance any pack immediately before installation????

I'm getting cell level Internal Resistance(IR) and maybe if I use internal discharge (avoiding stops) at 10A I can get a measure of discharge capacity per 8 cell segments. Doesn't look like I can hope for much more.

I don't think these hobby chargers were ever designed for testing huge batteries. Seems like if the supply battery is huge compare to the test battery things might work smoothly, but given same order of magnitude of the two, problems.

I suppose that the constructed supply batter might have a weak cell, but I'm being overcome by uncertainties. May buy the PL power supply just to get beyond the uncertainties.
 
#38 ·
Though the cells display works pretty well, the balancing current display never reaches stability before the test cells reach 100% "fuel" or charge, and the test cell do reach balance, though the voltage display shows balance.
You can get better balancing by adjusting the profile to favor balancing.

For example .. I have used my 3 PL8s to balance the 48s LiFePO4 in my car 8 cells at a time .. 3 PL8s .. each doing 8cells at a time .. 1st one half (3x8= 24s) , then the 2nd half (2x8=24s) .. end result displaying within +/-1mV .. for example .. 2.800 and 2.799 .. this +/-1mV difference also represents a +/- 0.036% variation .. or a ~99.964% accuracy .. I don't actually believe the device is actually that accurate , that's just what is displayed.

---

'good enough' is subjective .. there is just a finite limited faith / confidence in any measuring device.

Always remember .. The number of digits on the display is arbitrary , and has almost nothing to do with the devices actual accuracy .. thinking that the number of digits displayed does in some way directly correlate to a devices accuracy is a psychological trick .. commonly exploited in every day life by gas station pump displays .. many U.S. state department of weights and measures allow for up to a +/-2% pump tolerance .. that sticker on the pump .. only tells you on that date the tester certified that pump was no more than +/-2% .. even if we for the moment ignore any error in that tester's own testing .. on 10 gallons that could be only giving me 9.8 actual gallons .. the number of digits displayed on the pump is arbitrary .. 10.000 displayed , is not any more of an accurate pump than 10.0 displayed .. it is a psychological trick to think the 10.000 display actually is anywhere near that accurate of a pump.

2.800 shown from a +/- 0.1% device .. doesn't mean it is 2.800 .. it means a display of 2.800 could be as low as ~2.797 and as high as ~2.803 .. an actual 2.803 could show as a 2.800 just as a 2.797 could show as a 2.800.

---

To increase / improve the PL8's final balance .. you just have to keep in mind how the PL8 does cell balancing when you make adjustments to the profile.

There is a main +/- current across the whole series string of batteries.
At whatever rate you've programmed .. up to the ~40A limit.

The balancing works by bypassing a certain % amount of that current around a given cell .. up to the maximum Amp balance rate you've set the preset to do.

The amount % of bypass (100% = the max Amps of balance you programmed) displays on the PC charge control software .. and is also (by default) logged in the data file.

Soo .. on a charge or discharge of 40A and a 1A max balance .. that shows at a given moment a 50% bypass of cell #2 .. 50% of that 1A set is 0.5A .. that means cell #2 only sees 39.5A of charge or discharge.

Soo .. because the PL8 gives so many options you can adjust .. this can be adjusted / improved numerous different ways .. or some combination.

Option#1> Increase the max Amp rate of balancing.
Given the OEM balance wire size (that came with mine) I don't myself go above 1Amp .. but the PL8 itself can go up to 3A .. but personally I would recommend larger than the 22ga OEM balance wire to go up to 3A for any prolonged period of time.

Option #2> Increase the % of operation spent balancing.
If balancing on charging .. do you start to balance with 10% SoC left to go .. or 20% or 30% .. etc .. the further out you start to balance the more time that gives the balancing to work.

Option #3> Lower the C-Rate for ending the top CV phase of charging.
It will reach C/3 sooner than it will reach C/10 .. this also effectively gives the balancing more time to work.

Subtle detail note about balancing in the CV phase :
If in the CV phase your total for the 8 cells is presently at 2A of pack charging .. you can not get 3A of balancing at that same time .. you can set it for that 3A, but you won't get that 3A .. even a 99% bypass would max out at 1.98A bypassed .. which means that cell at the top charges slower , but it doesn't truly go to 0% charging.

Option #4> Lower the total bulk charge rate.
40Amps of charging gives a 1A balancing rate less time to work .. than a 20A charge rate with the same 1A balancing rate .. 10A to 2A .. etc.

Option #5> Multiple Passes /cycle on reasonably matched cells.
If .. for example .. out of a 2Ah balance difference .. you corrected for 1Ah on the 1st pass .. you could get 1Ah more on a 2nd pass .. but only if the cells are reasonably matched .. if extremely far apart from each other the multi-pass might not be the best option .. but with all the balance voltage sense wires in place, it won't do any damage.

I don't think these hobby chargers were ever designed for testing huge batteries. Seems like if the supply battery is huge compare to the test battery things might work smoothly, but given same order of magnitude of the two, problems.
Agree-ish.

Especially on supply batteries that swing in voltage under a given power of charge or discharge.

For example .. my normal 'goto' supply battery is 4p4s A123 (~900Wh or so) .. if I were to do a significant number of cycles .. because of less than 100% round trip efficiency .. I would have a bench power supply setup with CC-CV settings to do refilling of that supply 4p4s battery as needed .. that is easily 16x the size of a test 1s1p A123 cell.

--
An idea I've sometimes toyed with .. would be to rig up bidirectional supply side power .. tell the PL8 it's using a supply battery .. but in reality .. it gets supply power from a AC/DC power supply , and it gives regenerative discharge energy back to a DC/AC grid tie inverter .. I have a 340W AC/DC power supply and I have a 500W DC/AC grid tie inverter .. the trick would be to coordinating the two so they take turns correctly .. oh well .. another project to stay on the back burner for a while yet.

Although I myself like the regenerative discharge feature .. if one wanted to use just a power supply and the PL8's internal discharge .. Keep in mind the internal discharge is resistance based .. as the test cell voltage gets lower and lower that reduces and reduces how much your max rate is for internal discharge .. This especially pops up as you get down to trying to test voltages of 1 cell at a time .. the regenerative discharge feature uses a DC-DC buck/boost (not resistance based) .. so .. in addition to the efficiency benefits of regenerative discharge there is also testing benefits for low voltage test cells.
 
#39 ·
Thanks Ian. You have given me a lot to think about so my responses won't be immediate.

I don't have the flexibility you have in constructing a supply battery. About all I can really do, because of physical configurations, is a 2P12S LTO which is my current attempt. Since that doesn't remain strong enough to get an uninterrupted charge, and probably an uninterrupted regenerative discharge(vastly more important) I'm thinking of buying the Cell Pro power supply just to have a reliable supply. That will cut back on my discharge rate, but I may be able to get an uninterrupted discharge measurement by using internal discharge of 10A. That would suit me for a capacity measurement. The bugaboo I worry about is the "time out" which I have seen at least once. The manual is unclear on how this occurs. Can you tell me how the rig makes that judgement and when it might "bite" me??
 
#40 ·
The bugaboo I worry about is the "time out" which I have seen at least once. The manual is unclear on how this occurs. Can you tell me how the rig makes that judgement and when it might "bite" me??
'time out' for either charging or discharging is configurable for the presets.
I attached some screen shots of what tab that is under in the above post #35.

---

If you want charging to time out at 2 hours .. set it to 2 hours .. if you want 4 hours set it for 4 hours .. or 16 hours .. etc.

Same thing on Discharge .. if you want 2 hours set for 2 hours .. if you want 12 hours set for 12 hours.

---

The time out .. only comes up .. if your primary end criteria was not reached by that point .. like the top voltage for charging .. for the bottom voltage for discharging.

It is up to you .. if you want .. you can just set them to values soo high you won't reasonably get there .. at 10A 12 hours would only possibly come up on batteries with CV phase and also over 100Ah .. or no CV and over 120Ah.

---

Each of the time based setting you can adjust separately from the others.

the charge time out

the discharge time out

the rest period of time between each (charge or discharge).

You can also configure the PL8 to limit the CV phase as well .. not just to the C/3 or C/10 etc (ratio of amp you set to the Ah you set) .. you can also set a maximum amount of time for it to do the CV phase .. 15minutes .. 30 minutes .. etc.

---

You got options .. you can have it your way ;)
Whatever way that is.
 
#43 · (Edited)
Got the Cell Labs PL8 working pretty good, with a lot of help from Ian, but these hobby chargers don't really like to work with huge Lithium batteries. Eventually had to use internal resistive load, rather than regenerative discharge, because I couldn't come up with a large enough supply battery.

The PL8 will balance charge 8 cells at once. That means I can divide each pack into 3 sections and test each section with charge/discharge.

I just got things working pretty well today and did a 2.7 to 1.8V/cell discharge of one section of pack #4, which I named 4/9-16(pack ID/cell #s). The PL8 forced me to do the discharge in 2 sections because of some weird timeout problem, not yet diagnosed, but I totaled the two back to back segments and got 18.280 Ahr for that one segment, call it 19 Ahr if I had gone to 1.6.

Just now getting decent data, so results on a lot of segments on these used packs remain to be seen. Anyway, glad to have a number.;)
 
#44 ·
Some Battery Testing Results

Finished testing of my first full pack. Because of the PL8 ability to balance charge up to 8 cells, the pack is divided into 3 sections. Each section is balance carged to 2.6V and then unbalance discharged to 1.8V. There is very little capacity outside this range, and the end margins provide a bit of safety. In theory, if testing techniques are consistent then results can be compared.

For this 24 cell pack the cell level internal resistance varied from .6 to 1.1 milliohms. The three discharge capacities, between the assigned limits, were 18.280, 18.130, and 18,839 Ah.

Now, the remaining packs will be tested and compared. The 3 packs installed in the car remain to be tested, but are performing well and tracking well. In normal driving the BCM fooler in original configuration the pack stabilizes at 170-171V.
 
#45 ·
I went ahead and ordered a 24s 1.1kwh Fit LTO sub-pack.

From Here

I wanted to do more testing eventually anyway .. and if they test out good .. They might end up being a candidate for a future home battery supply upgrade.

My current ~0.9kwh 4p4s A123-LiFePO4 .. or .. my ~1.1kwh 10s Old Rav4 BEV NiMH packs both still work just fine .. for 'little' stuff like PL8 testing , exercise generator (pedal or hand crank) , etc .. But , I'm considering doing a battery system able to run my whole house (for a little while) .. so expansion would be in order .. and I do like the long cycle/use life of LTO .. soo .. more testing .. which I usually enjoy anyway :D
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top