Charger for LFP cell testing

[karrak]

The Reaktor manual says that it supports an external resistor for faster discharging (since it can only dissipate 20W internally). The wording is very wonky but it sounds like it supports up to 300W discharge rate in that config. I could pick up a couple power resistors and set that up, to reduce how long my capacity measurements take?

Can't you use the regenerative discharge function of the Reaktor by using either a 12v-24V car/truck battery as the DC input rather than the inbuilt AC supply? Doing this gives you the full 300W discharge and means that you can use the energy transferred to the input battery by the regenerative discharge to recharge the LFP cell under test.
Another option is to do the testing using your current LFP battery, if you are interested I will describe how to do this.

Yeah, it's a lot of cells. If this weren't an RV, i would have stuck with FLA...But it wasn't doable with the space/weight capacity we have. Still, it grants us a LOT of flexibility and that's worth spending serious time baby-sitting the cells in order to measure and re-balance the pack before re-commissioning the rig.

You don't have to test the cells individually, you could test them as four groups of 5P or if you want to accurately keep track of the capacity loss of the battery just test one cell as a reference and only retest that cell as the battery ages . I am not even bothering to do that. I am just monitoring the capacity of the overall pack during normal operation and using this to see how much capacity is lost over time.
IMO the whole issue of balancing/rebalancing and battery management can be far simpler for LFP batteries than FLA batteries.


Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller

[zamboni]

Can't you use the regenerative discharge function

Yes, but I don't have anything to put all that charge into, except for the other cells in the pack (and the truck's batteries...i could unplug the alternator). If i were trying to keep the house battery on-line during all this, then i suppose i could pull 4 cells at a time out, regen them back into the pack to test, then recharge and swap for the next 4 cells. But that sounds like a lot of mucking around with interconnects and bus bars...i'll consider it. I guess i'm just simple; i feel like taking the pack off-line and testing cell-by-cell is the most straightforward approach (if not the most savvy), with the fewest opportunities to make a mistake.

And it'll be kinda fun to try and make up a suitable dump load from salvaged materials. I've read that wooden pencil leads are useful for such an endeavor.

IMO the whole issue of balancing/rebalancing and battery management can be far simpler for LFP batteries than FLA batteries.

For me it's not so much an issue of "complexity", just that the consequences for making a mistake are more extreme. FLA are just bomber. Bomber enough that for the money, "battery management" outside of adequate charge control, maintenance, and a basic LVD doesn't seem worth spending extra on. I'm talking about my application here, not mission-critical stuff. The first set of batteries in my rig were el cheapo golf cart batteries, which i did my best with, but they bore the brunt of most of my 'learning opportunities' -- for years. I'm really glad i didn't start out with LFP! For my money, FLA is still (at this point in history) a better choice if space or weight aren't important.

[karrak]

If i were trying to keep the house battery on-line during all this, then i suppose i could pull 4 cells at a time out, regen them back into the pack to test, then recharge and swap for the next 4 cells. But that sounds like a lot of mucking around with interconnects and bus bars...i'll consider it. I guess i'm just simple; i feel like taking the pack off-line and testing cell-by-cell is the most straightforward approach (if not the most savvy), with the fewest opportunities to make a mistake.

I agree that either using an external battery or a dump load gives you the fewest opportunities to make a mistake. If I was going to test all my cells individually I would use my pack while it is still in use. Here is the outline of the procedure I might use.

1. Take the whole battery down to around 50% SOC and set the charge controller bulk and float to 3.35V/cell which will keep the battery around 50% even with external loads.
2. Remove one cell from each of the parallel groups. Where each cell has been removed place a nut and bolt through the interconnect straps to restore the battery connections.
3. Discharge each removed cell to say 2.5V in turn, recharge it to say 3.6V, note how much charge was needed. The discharge/charge cycle can be done in one step using the Cycle function of the Reaktor. Make sure all the parameters are setup in the Setup menu, especially the max input battery voltage for regenerative discharge and minimum input battery voltage for charging.
4. Remove the second set of four cells and repeat steps 2 and 3 on these cells. Repeat this procedure until your input battery is down to single cells in series.
5. Remove the last cells and replace the rest of the battery.
6. Discharge the main input battery by 10% to take it down to ~ 90%SOC.
7. Carry out step 3 for each of the remaining cells.
8. Charge the main input battery up to 3.6V/cell and then reinstall the last cells back into the battery.

This would have to be done with care and I would always be monitoring the individual cell voltages of the main battery and the cells under test. Can you see any flaws in my logic?

And it'll be kinda fun to try and make up a suitable dump load from salvaged materials. I've read that wooden pencil leads are useful for such an endeavor.

Old car headlights or spotlights make a good load.

For me it's not so much an issue of "complexity", just that the consequences for making a mistake are more extreme. FLA are just bomber. Bomber enough that for the money, "battery management" outside of adequate charge control, maintenance, and a basic LVD doesn't seem worth spending extra on. I'm talking about my application here, not mission-critical stuff. The first set of batteries in my rig were el cheapo golf cart batteries, which i did my best with, but they bore the brunt of most of my 'learning opportunities' -- for years. I'm really glad i didn't start out with LFP! For my money, FLA is still (at this point in history) a better choice if space or weight aren't important.

The cost of a BMS need not be a very large percentage of the total battery cost and can make the battery immune to mistakes and faults and prolong its life. I would say the same goes for FLA batteries. If I had a large expensive FLA battery I would be monitoring the individual cell voltages and SOC to make sure that the battery was getting charged and equalised properly.
As for battery cost, if you look at the cost of each kWh cycled through the battery, the battery size needed, charging requirements, battery efficiency and maintenance I think you will find that LFP batteries are cheaper.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller

[zamboni]

I'm not sure i'm qualified to critique...but what you describe makes sense to me. I like it because physically removing the cells from the pack would help to avoid mistakes (i'm a spatial/visual thinker, which is a big part of why electricity is not my strong suit. It's all PFM as far as i'm concerned). I can't really do it that way though, because my cells would be a total witch to physically separate. The pack is spacered and strapped together, but this was done in-situ because the assembled pack could not fit through the compartment opening (the compartment itself is larger than the door). So the cells were loaded one at a time, then the strapping put on (allthread rods). Even after i cut in my new access hatch, it will still be too small to lift the pack out intact. That means in order to break out individual cells, i'd be spending a huge amount of time fussing with the bus bars and bypass/jumper connections; an accident waiting to happen. Instead, i will remove all bus bars and just test each cell serially. This will also ensure i don't accidentally skip any terminals when i'm going thru and cleaning/noaloxing/torquing the bus bars afterwards. I'm going to do so many things differently next time!

[zamboni]

So, i have an update but can't seem to post anything longer than a few words...?

[karrak]

Forum software chops the rest of the post off when it encounters certain character sequences or some UNICODE characters. The way to work out what is causing the problem is to make a copy on your computer before you post anything by typing ^A to select all the text and ^C to copy it to the clip board. Submit the post and if it has been truncated edit the posted truncated message and do a ^V to paste the full message you have saved. Look at where the message was truncated and delete the offending character and repost.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller

[zamboni]

Thanks for the explanation karrak, i'll try again. No big deal, its just apostrophes, quotes, hyphens, and ellipses -- what is this, 1991? I just spent 20 minutes re-submitting my post to get it to take. That's the last time im going to do that. A whole lot of work, just to come off looking like an illiterate goober who cant use punctuation. I dont need help looking stupid.

(Nearly) original post follows:

It took me longer to start my cell-balancing project than i had hoped, but im doing it now. Many thanks to this forum; im bringing my cells up to 3.6V in parallel groups of 5 (4 groups), resting, then individually charging each cell to 3.6V (then resting again and charging again) for a "top balance" per PNjunctions KISS post. After that, theyll be assembled back into a pack and will probably never see 3.6VPC again (but i can rely on my high-voltage disconnect and charge controller to cut off at the right time).

In case someone out there is actually looking for a 1S cell charger like i was, and has read this far in the thread, heres an update:

WHAT IVE TRIED
I started with the Turnigy 300W/20a AC/DC charger. It cost me $134 shipped. This charger has an alleged 20a output and can be plugged directly in to mains 120VAC power to run it. I bought it in advance of the project, and when i was finally ready to use it i found it was DOA!!! It threw a disconnect error (which isnt even mentioned in the manual) when i connected it to one of my LFP cells. I tried it on my trucks battery with the Pb profile (doesnt require a balance port connection), and it was still no good. HobbyKing took it back without a fuss but i was out the shipping and i sure as hell wasnt going to get another one.

Next i tried a HiTec PowerPeak E7, which was billed as their pro level charger, and priced at $165 on Amazon (!!!). Also able to be powered from 120VAC mains. The manual was available online, i read it completely before buying. It was much better than the joke that shipped with the Turnigy but the charger was still garbage. A toy, suitable for charging toys. Charge voltage was only settable in pre-determined intervals (NOT explained in manual), and couldnt even go very high or low. The charger registered voltage 0.15V lower than my Amprobe 560 - NOT COOL. It did come with a nice array of cables and accessories, though. I returned this as well, since it was useless.

At this point, i started looking at bench power supplies. Why the hell not? Well heres why: unless youre willing to spend upwards of $400 you wont get more than 10A output, and for a large-capacity cell (200Ah in my case) im not sure bench supplies are even up to the duty cycle of running for most of the day wide-open. I looked at some case-off reviews of these supplies and their heat sinks / fans were even more anemic then that of the Turnigy and PowerPeak. Plus, theres no coulomb-counting for total charge analysis, and no internal resistance measurement. Itd be cool to have a benchtop power supply but it would not solve my cell-charging requirements.

Finally i went back to Revolectrix. The PL8 is still overkill, but i got lucky and picked up a GT1200 on close-out pricing. Then i found a guy selling refurbished HP server PSUs as 12V sources for LiPo charging (beats the hell out of Revolectrixs $200+ PSU). The GT1200 is rated to 600W output with a 12V power supply. Since im only charging 3.xV cells, thats plenty. The PSU i bought is rated to 750W. THEN I bought a pigtail on eBay to adapt the bullet connectors on the PSU to some kind of blue RC plug on the GT1200. I paid $135 for the charger (shipped), $35 for the PSU (shipped), and $7 for the pigtail (shipped). I would have had to also buy parts to make up charging leads (banana plugs plus alligator clips), but i had the accessories left over from the DOA Turnigy.

The GT1200 is currently charging my cells at 40A and doing a great job. It is basically a huge heat sink with a pair of fans inside a plastic case, and some circuits with a display tacked on. Compare with the other chargers, which had pathetic heat sinks and tiny fans (yeah i opened them up). The PSU is much smaller than i had feared, and while it is certainly a pain in the ass to have all these wires everywhere, at least my cells are charging. I may build a simple case to hold the two devices together so theres less of a trip hazard.

MY CONCLUSIONS
1. Nobody makes a (retail, consumer) box that you plug in to mains power and that can charge large prismatic LFP cells. They just dont. Theres no technical reason against it, just need a somewhat larger box than usual, bigger heat sinks, more fans. I guess were too fringe to be a market right now. Dont be like me and get all bent out of shape that theres nothing off-the-shelf to do exactly what you want. Thats just where LFP is right now, be grateful its not harder!!! If you arent okay with this sort of thing, youre barking up the wrong tree. Come back in a couple years.
2. The best thing youre going to find is a Revolectrix product paired with a separate power supply. If you want to be mobile, compact, cheap, etc get over it. Its not even about quality or features, their stuff is just the only option thats capable.
3. Regrettably, the simpler Revolectrix products (like the GTs) are being phased out (see #1 about being too small a market). Get old-stock GT chargers while you can, or possibly the old-style PL8s, which are on clearance. If youre reading this thread 4 months from now, youll likely only have the new touch chargers available to you. Nice chargers, see if you can get a refurb to cushion the cost.
4. In most cases, it seems necessary to have the "balance port" connected, even though you're charging 1S so /balancing/ is a silly concept. I believe these chargers have their most-accurate voltage sensing circuits connected to this port though, so you have to use it. Which is great, because you generally also get 4-wire internal resistance measurements, too. But plan on making up the leads / jumpers / cables to make this happen.
5. This all said, products are constantly changing. If youre looking for a charger a year from now, heres the main sticking points i ran into that you should check for before buying:
* Make sure you know what the output capabilities of the charger are FOR THE PSU YOURE USING. 12V PSUs are more affordable but youll generally be limited to 40a output from the charger. Do the math and decide if you have that kind of time.
* If its an all-in-one unit, find out how much its de-rated when running on 120VAC mains. If the manual is not absolutely steel-clad clear on this, call up the company and be annoying until they let you talk to an engineer.
* Make sure the charger can be adjusted in the increments you want, and over the voltage/current range you want.
* Ensure the cell size wont mess things up. The HiTec charger freaked out when connected to a cell so much larger than what it was expecting. Revolectrix is unfazed.
* Dont forget about duty cycle. Charging 5 paralleled 200Ah cells takes half a day at 40A, so make sure the charger (and PSU) will run wide-open for that long without having a fit. Here again, call and ask. Dont trust manuals online (What is the world coming to?).

- Jerud
------------------------------------------------------------
1220W array (4x 305W Astronergy panels @ 48V)
1000Ah LFP house bank (5P4S HiPower 200Ah cells)
MPPT solar charge controller (MidniteSolar Classic 150)
2800W PSW Inverter (Magnum MS2812)
ME-RC, Trimetric, and JLD404

2001 Fleetwood Prowler 5th wheel 25 foot, self-rebuilt

[karrak]

I am sorry you have had trouble with the Reaktor 300W. I haven't had any problems with my unit charging and discharging at currents up to 18A.

If it had been me I would have tried to find out what was wrong with the unit you bought or gone for a replacement. There are lots of people on the internet who could have helped you with this.

Simon

Off grid 24V system, 6x190W Solar Panels, 32x90ah Winston LiFeYPO4 batteries installed April 2013
BMS - Homemade Battery logger github.com/simat/BatteryMonitor/wiki
Latronics 4kW Inverter, homemade MPPT controller

[zamboni]

I would certainly have preferred to fix the Reaktor; it would have saved time and money plus i hate "throwaway consumerism". When it comes to electronics through, i'm useless at repairs and don't have the setup nor tools to do the work. I checked that none of the internal fuses were blown and that was the limit of my abilities. Although i do want to learn how to do what is probably a simple repair like this, and could have likely gotten help on some forums, this was not the time. The insane winter storm did not help, either (living in an RV with the electrical system offline and temps in the teens is super not-fun). I have a window for this project and it's not big enough to pick up a new skillset. Also, i was just plain not impressed with the charger and didn't feel like the company should be rewarded for making it by getting my money. There were warning signs...i normally would have avoided it if i could find an alternative, but i could not, and the Revolectrix option (at that time) was a whole lot more expensive. I really wanted to avoid having 2 separate boxes. I was lucky that by the time i went through all this Revolectrix kicked off their sale on the discontinued GT models. In the end i got a good product and am satisfied that there aren't any readily-available alternatives. It just cost me a few bucks in lost shipping to learn this.

It sounds like the "Reaktor" i got and the Reaktor you have are similar only in name. Based on that long post on rcgroups (linked by sensij earlier) it believe Hobby King is using the good reputation of the earlier version to sell the newer, flimsy version. HK gave me no hassle with the defective unit, just asked to see a video then they accepted it as defective and of course didn't want it back. Just like every company that sells cheap crap and knows it. A few factors combined to make me uninterested in a replacement, however:

• The manual was poorly written, far too brief and incomplete, and seemed to be written for a different product entirely (likely the previous version), as it referenced menus which either didn't exist or had different names, and spent a lot of time discussing how to save log files using the USB port...which doesn't physically exist on the 120VAC model. I wasn't able to get anything more helpful out of calling HobbyKing, and that is the ONLY contact there is for this charger. No real manufacturer.
• HobbyKing didn't seem to care very much about the fact that the charger was dead; the staff all did their jobs but it was obviously call-center people working from scripts and flowcharts, registering zero concern. They asked me to video the malfunctioning charger but didn't say anything about how to actually get them the video. When i pressed them on that detail they were like, "i dunno maybe put it on dropbox?" Didn't think it through, did you? I prefer to avoid half-assed companies like this, which are just cranking out whatever products they have deemed profitable without really trying or caring. I don't shop at Walmart, either.
  • In contrast to this, when i called Revolectrix, they not only cared, but let me talk to a product engineer (without any hassle), who actually pulled out the charger i was considering and clicked through the menus on the phone with me, to verify that the settings and capabilities i wanted to use would actually work. That is the kind of company i want to give my money to.
• The discharging feature, which i was interested in (but no longer have enough time to make use of), was going to be pretty useless for me. The "manual" offered maybe 2 sentences on how to set up the external dump load for higher discharge rates. It wasn't clear what the limitations or requirements were, and at that point i wasn't even sure the charger would do it, because the menus didn't make sense. I had planned to make up my own resistor (using pencil leads), and was a little excited about that, but there was a strong chance i'd have to make and re-make the resistor along with doing a lot of other trial and error. By this point i had neither time nor interest.

I'm not angry or upset about the Turnigy...it's just how things are. It seemed that there just had to be a solution out there which was better than hobby RC chargers, for these large prismatics. The fault really lies with me: I simply didn't realize how new and un-developed this field is. I guess i assumed that it was just "new to me" and the main hurdles had already been figured out. I ought to be grateful there's anything available at all, or else i'd be doing all this manually with an expensive benchtop power supply.

The forum's truncating craziness seems to come from writing posts in a separate text editor and pasting them in. I can write like a human being so long as i type it all directly into the forum's "post" box. Ironically copy/pasting is exactly the kind of thing that a glichy forum encourages users to do! : )

[Sunking]

I would certainly have preferred to fix the Reaktor; it would have saved time and money plus i hate "throwaway consumerism"

You got what you asked for listening to Karrak. He sells Chi-Com junk and does not know squat about batteries or electrical.

[zamboni]

Karrak is the only person who actually replied with a suggestion that indicated he had read my entire initial question, so yes i did listen to him. That deserves respect in my book. It does not sound to me like the Reaktor he uses is comparable to the one i bought, which does indeed appear to be knock-off junk. That's why i didn't exchange it for another but got my refund instead. At some point the charger Karrak uses stopped being available at a good quality level. I could potentially have anticipated this from the RC forum posts...but i get so weary of spending my days second-guessing the whole internet. I tried for myself and now i understand a little better the games that these manufacturers play.

The GT1200 is working fantastically well and at half what a PL8 would have cost. But "use a PL8" is the gospel here. I tend to question that variety of repetition and this is exactly why. Revolectrix is undoubtedly a good company making quality products...but nobody talked about the one that was actually the best fit for what i'm doing. I would never have found the GT1200 if it weren't for all the mentions of the PL-8...but nobody here gets to say "told you so".

[zamboni]

Speaking of "told you so"

While the GT1200 is definitely doing a great job, and it's a good fit for me re: price point and frequency of use (rare), i've found a couple characteristics about it that demonstrate why getting a PL8 may be preferred if you work with the charger frequently.

First is that the "internal discharge" function won't let you drain anything below 3.0V. So long as you're working with 2S or more configurations, this poses no issue. But if you're discharging a cell 1S to 2.5V for a bottom balance (or capacity measurement), the process will error-out once you hit 3.0V. Note that the charger will let you SET a lower voltage without warning. It just will error and cover up the info screen so you can't read your stats up to the point of the error. The regenerative discharge does not have this restriction.

Second is that the info screen which shows the stats on your last charge/discharge cycle (duration, Ah, internal resistance) automatically clears itself about 15 minutes after the cycle ends. That information is not accessible once the screen is cleared. So you really don't want to leave this charger unattended if you desire that information.

Both of these issues would be moot if the charger was a "logging" type (as the PL8 and PL6 are, i believe). So this is a good example of the tradeoffs between the two price points.

When shopping for chargers, they all tend to look the same, and nobody really explains specifically why the PL8 costs so much more. These details end up being really important though and you can't get them from the manuals or any online reviews. So i guess i'll just keep talking to myself over here and hopefully someone will find it useful one day.


- Jerud
------------------------------------------------------------
1220W array / 1000Ah LFP house bank
MidniteSolar Classic, Magnum MS2812
ME-RC, Trimetric, and JLD404
2001 Fleetwood Prowler 5th wheel 25 foot, self-rebuilt

https://www.livesmallridefree.com

[coremax]

When you talking about the model No. I really don't have any idea about it. But make sure the charging voltage is constant 3.65V and output currently is 0.5C-1C charging is ok.

MOD NOTE: Please do not include any advertising links or websites concerning solar products without first getting permission from the Admin.