LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[exMaggieDrum]

I had trouble using my Mastek 2020EX base battery charger. I could get individual cells (and the whole string of 16 afterward) up to my target top balance of 3.65v but I could not get it to maintain that voltage. The charger could not be fine tuned between 1A and 0.0A.

At 1A it would continue to raise the voltage unless I turned the fine voltage control down with little bumps by my finger. I could get it to below 1A for just a second or two and then it would either go back up to 1A or more, or more usual, it would immediately go to 0A. Then the voltage would start to go down on the cells immediately. So I could never keep it at the set point I wanted at 0.1-0.2A that has been suggested.

So I just bounced back and forth for about 15 minutes manually bumping the voltage and then shut it down and went on. I have left the cells in parallel since then. They immediately started going down slowly. Three days later they were at 3.480v.

Mbartosch suggested that I should not leave them at this point, that I should draw them down to about 3.29v (how they came from Sinopoly). That sounds right but any comments are welcome.

I do intend to build the Edge Detector designed by Martin per his latest schematic with the added suppression diodes as suggested by OceanSeaSpray. I ordered the parts yesterday (Jameco.com and Mouser.com for the bidirectional suppressor diodes). I haven't put together any electronic circuits in years. It should be fun. I am doing the high side circuit and will test with my BMS, my power supply, and the Blue Sea relay. I will post what I get afterward.

Thanks, again, to Martin and OceanSeaSpray for their contributions.

I do want to figure out a way to make the end result boat ready with some kind of coating and suitable box/mounting. And I will build a spare too.

[mbartosch]

Originally Posted by exMaggieDrum

I had trouble using my Mastek 2020EX base battery charger. I could get individual cells (and the whole string of 16 afterward) up to my target top balance of 3.65v but I could not get it to maintain that voltage. The charger could not be fine tuned between 1A and 0.0A.

I don't know the Mastech device, but I assume it's a simple DC power supply, not a battery charger.

Generally what I would do with any DC supply for top balancing is this:
- power up DC supply without any load (only volt meter connected). Set output voltage to the desired value (e. g. 3.65 V).
- do NOT touch the voltage dial anymore afterwards!
- If the device has a current limiter, set it to a suitable max current, e. g. 20 A. If not, the device MUST be capable of withstanding a permanent short on the output terminals.
- connect batteries
- wait until current tapers off to < 1 A
- done

Reason: voltage with open outputs is set to the desired voltage, so if you attach a load (e. g. the cells) voltage will break down to the cell level. This is because the cells have a very low internal resistance BUT they also build up a "counter voltage level". The difference between the cell pack voltage and the target voltage divided by the total resistance on the terminals will determine charging current.
Cell voltage will rise with SoC, eventually the cell voltage will reach target voltage. Remember: don't mess with this! Changing output voltage when under load will not work and will not get the desired results.
If multiple cells are wired in parallel, during charing the ones "behind" will draw more current than the ones which are ahead. Once the current has tapered off to a very low value, it is safe to assume that all cells have the same SoC (because cell voltage rises with SoC at the high end of the "knee").

At that point it is OK to stop charging, no more work needed.

At 1A it would continue to raise the voltage unless I turned the fine voltage control down with little bumps by my finger. I could get it to below 1A for just a second or two and then it would either go back up to 1A or more, or more usual, it would immediately go to 0A. Then the voltage would start to go down on the cells immediately. So I could never keep it at the set point I wanted at 0.1-0.2A that has been suggested.

This sounds weird to me. You are trying to adjust things for current during chargin, whereas this should happen automatically (see above). If at all possible I'd recommend to give the above a try.

If in doubt, discharge the cell pack to about 95 % SoC and let them top balance again.

So I just bounced back and forth for about 15 minutes manually bumping the voltage and then shut it down and went on. I have left the cells in parallel since then. They immediately started going down slowly. Three days later they were at 3.480v.

It is totally normal and expected that they drop to a lower voltage level after the charging source has been removed. This also explains why they don't "self-balance" much when left connected in parallel. Balancing requires external excitation, i. e. raising the cell voltage to the desired target voltage.

I do intend to build the Edge Detector designed by Martin per his latest schematic with the added suppression diodes as suggested by OceanSeaSpray. I ordered the parts yesterday (Jameco.com and Mouser.com for the bidirectional suppressor diodes). I haven't put together any electronic circuits in years. It should be fun. I am doing the high side circuit and will test with my BMS, my power supply, and the Blue Sea relay. I will post what I get afterward.

I hope the high side version will work fine, I only tested the low side switching variant. Good luck, and I will be happy to hear from your experience.

[exMaggieDrum]

Thanks for your comments Martin. I did expect to see the resting voltage to go down after I got them to the target but since I am new to this type of battery I wanted to confirm this was normal. It always happens with other battery types.

Regarding the power supply (described as both a power supply but mostly as a battery charger by Mastek), it only displays voltage to tenths of a volt so I cannot set it to hundredths. I followed the instructions to the letter from the seller for setting to a constant voltage.

I set it to 3.6v to start. I hook it up to my cell and then the display drops to 3.5v. That doesn't concern me but it will often go up to 3.7v left to its own devices. Starting with a cell at 3.295v the amps will at first be in the 8-10A range but in a few minutes will go up to around 16-17A. I set the max amps to be 17A as the power supply is rated at 20A (per recommendation by MaineSail).

I check on the charger and cell every hour for the first 4-5 hours as the voltage rises. The amps will start coming down, as expected. I monitor every 30 min and then every 10-15min when the cell approaches my target. Even though I set the displayed voltage on the supply to 3.6v it will quit (0.0A at 3.5+v) every time unless I bump it up.

I would expect some voltage drop at the cell terminals from the power supply leads but with the low resolution in the setting I don't know how high to set it. And, it is not consistent from one cell to the next. I tried leaving it at the setting that I finally got to from the last cell (3.650v) and it would go over that with the next cell unless I lowered it. So I ended up changing it with the fine voltage adjustment with each cell.

I understand what you are saying but I have not found a way to set it and forget it. This may be something I am doing wrong with this particular model. I replaced the 12gauge leads that came with the power supply (clamp on) with 10 gauge wire and new soldered on clamps. The other wires got too warm at the upper end of the amperage.

[OceanSeaSpray]

It is a bad way of going about it anyway. If you are going to use a PSU at all, you only need a very small one, well regulated, to finish the job at the very end, and you set it just like Martin explained. You only need a few amps out of it.
The bulk of the charging can be done much more efficiently just using the alternator.

Then by leaving your fully charged cells lying around in parallel, not only you achieve strictly nothing, but it is in fact harmful to them and this is why I won't accept cells that are no longer factory packaged.
I only balance a pack if it can be loaded and discharged again immediately afterwards.

[mbartosch]

Originally Posted by OceanSeaSpray

It is a bad way of going about it anyway. If you are going to use a PSU at all, you only need a very small one, well regulated, to finish the job at the very end, and you set it just like Martin explained. You only need a few amps out of it.
The bulk of the charging can be done much more efficiently just using the alternator.

Totally agree. The balancing can be done with comparatively low current. I used my trusty old power supply which I have been using on my workbench for about 25 years now, and applied the procedure explained above.

Of course it comes with simple analog dials, but it is very accurate and stable, and I knew I could trust it not to exceed the set voltage. Any power supply will work, provided you can set the desired target voltage (and it keeps it) - and it does not only survive prolonged shorted outputs, but also does not drift from the set voltage.

My old power supply is limited to 2.5 A but despite the low output was easily possible to balance my 400 Ah bank with it.
Just as Eric explained: get the pack near full level with some bulk charging means and use the "good" power supply for the final balancing action.
You can see the device I used on the image in this blog post (but of course this does not mean you need to hunt one down on ebay )

[exMaggieDrum]

Thanks for the lastest feedback. I didn't dream this process up on my own. I tried to follow the protocol actively suggested by others on this forum, even to using the PSU that was recommended by some.

I don't think I have damaged the cells, at least based on what I have heard. I will discharge them today and take them out of parallel. I didn't mind taking the time to bring them up so much. I can see how an alternator would be faster for the initial charge up to near the top balance target. I don't see how I would get to a specific target voltage with any preciseness at all. I could get it close but the knee is so steep near the top balance the danger would be to overshoot the target easily.

Perhaps I didn't use the PSU correctly as to setting the voltage but it wasn't for lack of trying. I followed the distributors instructions to the letter. The unit is marketed specifically for charging batteries but obviously it is not a "fine" instrument.

This forum continues to have as many disagreements as agreements on many aspects of preparing and using LiFePO4. Who should someone like me follow?

OceanSeaSpray, did you disagree with the suggestions made by others before on top balancing? If so I missed it. The difficulty for someone not as expert is how to sift the chaff from the wheat on here. I have studied all the recommendations for months now and have made notes, underlined, and put together a summary of what seems to be the general agreements from the ones who seem to have the most credibility (including you). You suggest that I have permanently damaged my cells. Please share how you would get a proper top balance.

It seems it hasn't worked out so well from just the top balancing so I will try to be even more careful in the rest of it but not sure how I will do that. Lots of landmines in the DIY world right now it seems.

I am sure I must seem like an incompetent amateur right now to some of you. But I have more general electrical and other experience than some so I am not terribly out of the norm for the next generation of users of this technology. I'll just do my best and see how it works out.

Please note: I am not trying to be overly sarcastic and I still appreciate the contributions of all. But it seems that one of the best ways to find out the "right" way is to propose or actually do something and report on it and then find out what you did wrong after the fact. My mistakes may help some who follow though.

[exMaggieDrum]

Originally Posted by mbartosch

Totally agree. The balancing can be done with comparatively low current. I used my trusty old power supply which I have been using on my workbench for about 25 years now, and applied the procedure explained above.

Of course it comes with simple analog dials, but it is very accurate and stable, and I knew I could trust it not to exceed the set voltage. Any power supply will work, provided you can set the desired target voltage (and it keeps it) - and it does not only survive prolonged shorted outputs, but also does not drift from the set voltage.

My old power supply is limited to 2.5 A but despite the low output was easily possible to balance my 400 Ah bank with it.
Just as Eric explained: get the pack near full level with some bulk charging means and use the "good" power supply for the final balancing action.
You can see the device I used on the image in this blog post (but of course this does not mean you need to hunt one down on ebay )

Thanks Martin. If 16A for a couple of hours damaged my cells then there is something else grossly wrong here. The highest current I used near the top of my charging was 1.0A. Either I misused my PSU or it is not as fine.

One question is how important it is to supply very, very low current to 16 100Ah cells (0.1A) versus 1.0A to the same voltage. I certainly couldn't let them set at the low amperage for very long. Perhaps that is the issue.

[mbartosch]

Originally Posted by exMaggieDrum

I am sure I must seem like an incompetent amateur right now to some of you. But I have more general electrical and other experience than some so I am not terribly out of the norm for the next generation of users of this technology. I'll just do my best and see how it works out.

No, on the contrary. I think you are doing a very good job, but your latest posts seemed to indicate that you were uncertain with regard to the actual top balancing with your particular power supply unit.
Nobody wanted to tell you that you are doing it the wrong way; Eric and me were just trying to provide help by providing input on how we did it. You can see that we all had different approaches. This is fine, there is more than one way to do it, and there is no definite right way.
In the end all that counts is that the cell pack has seen 3.65 V. How you do this exactly is secondary.

Anyway, I don't think you have damaged your cells. We were just reminding you that it is not optimal for the cells to stay near 100 % SoC unnecessarily long, that's the reason for recommending discharge after finishing the balancing process.

After all, these cells are not "brittle", they will also take some beating, but we tend to babysit them so nothing happens to our precioussss. Maybe overdoing it a bit

[SV THIRD DAY]

The cells are much more tolerant to abuse than the panicked consensus would seem to indicate. So I wouldn't worry about what I view as a little hiccup...back away from the cliff..


What I do think this indicates again is the value in keeping things simple...and back away from the over engineering and messing with it. I have been involved now in setting up and sending out cruising about a dozen boats outfitted with LiFePO4 batteries and folks this isn't rocket science. It's really simple if you will just stop the overengineering compusion. For example, I chuckle at all this top balancing fuss, tie the cells together with 1ft 12ga wire and let them sit for a few days...how hard and risky is that? In setting up 13 boats its worked fine to date. The cells equalize and you move on to your install and go cruising. This crazy need to get back to 99.9997% full SOC is an engineering test question but not real life.

You have to ask yourself a question: are you guys building a science project or putting batteries on a cruising boat? Don't get lost in the process and remember the ultimate goal.

[evm1024]

I use one of those (VOLTEQ HY3030EX) lab supplies for various charging needs on my LiFePO4 bank.

They are OK but not research grade which means that the voltage/current settings stability are not as I might like. And the readout is lacking in precision.

That is not a problem.

When I want the pack full or when I want to top balance the pack I use my inverter/charger to do the bulk charging. I have the older model 3000w xantrex which does not have user settable voltages but does have a setting for maximum charge rate. This is a current limit in effect and runs from 100% to 10% of max current (AH setting /5).

In any case I need to monitor the pack and individual cell voltages manually.

Once the pack is near full I switch to the Hy3030EX for the remainder of the charge. If doing top balance I break the pack down and parallel the cells.

Once I've decided on a max voltage (oh say 3.65 or 13.9 or ???) I rough set the supply to that voltage and then put a load on the output. This load is either a 10 ohm power resister or a car light bulb. This resistance gives me about an amp or so (or 0.36 amp) and loads up the regulation circuity.

Then I use an external meter of known precision to set the voltage to what I want. Letting it run for a bit with the load allows for any thermal drift and mechanical oddities of the settings. Then the voltage is tweaked if needed.

With a stable voltage I then connect the pack without touching any of the controls and monitor the voltage with the external meter as the pack takes on the final charge.

Oh yes, I do have a shunt in series so that I can get a more precise reading of the current - but that is just me. The HY3030EX amp meter is fine for the precision we are looking at in current (10 amps vs 1 amp for example).

In any case I always monitor to be sure that nothing has gone astray. It is not like I've never made a mistake....

Just what I do.

[exMaggieDrum]

Originally Posted by evm1024

I use one of those (VOLTEQ HY3030EX) lab supplies for various charging needs on my LiFePO4 bank.

They are OK but not research grade which means that the voltage/current settings stability are not as I might like. And the readout is lacking in precision.

That is not a problem.

When I want the pack full or when I want to top balance the pack I use my inverter/charger to do the bulk charging. I have the older model 3000w xantrex which does not have user settable voltages but does have a setting for maximum charge rate. This is a current limit in effect and runs from 100% to 10% of max current (AH setting /5).

In any case I need to monitor the pack and individual cell voltages manually.

Once the pack is near full I switch to the Hy3030EX for the remainder of the charge. If doing top balance I break the pack down and parallel the cells.

Once I've decided on a max voltage (oh say 3.65 or 13.9 or ???) I rough set the supply to that voltage and then put a load on the output. This load is either a 10 ohm power resister or a car light bulb. This resistance gives me about an amp or so (or 0.36 amp) and loads up the regulation circuity.

Then I use an external meter of known precision to set the voltage to what I want. Letting it run for a bit with the load allows for any thermal drift and mechanical oddities of the settings. Then the voltage is tweaked if needed.

With a stable voltage I then connect the pack without touching any of the controls and monitor the voltage with the external meter as the pack takes on the final charge.

Oh yes, I do have a shunt in series so that I can get a more precise reading of the current - but that is just me. The HY3030EX amp meter is fine for the precision we are looking at in current (10 amps vs 1 amp for example).

In any case I always monitor to be sure that nothing has gone astray. It is not like I've never made a mistake....

Just what I do.

Thanks evm. A very helpful reply. Other than the need to discharge my cells to a lower SOC it looks like I might be there. Using

I was a little PO'd with the implication that I had damaged my cells so that someone might not want to buy them - not that I would have any intention to do so - with the other implication that this is what amateurs do. I wouldn't buy used cells myself. I was given the opportunity to do so when sourcing cells and I passed.

I realize that I'm being oversensitive so please pardon my outbursts. I have a large investment in both time and money in this already and will be putting more in to it. I do not want to make any irreversible mistakes with the largest $$ investment involved.

I did not want to remove my existing batteries on the boat and do the balancing there, and I don't have room to do it on there anyway. And I did not want to jury-rig a temporary set of charging cables to use either the inverter/charger or the alternator. I did get a very high quality DVM (Fluke 87) to monitor the voltage and used my Fluke 337 to monitor the amps going in. I decided the exact number of amps was not that important, but the magnitude. The Volteq HY2020EX (sold by Mastek) PSU I used was reasonably close to my measurements with my Fluke so I quit using it after a few days.

I also decided that there isn't any magical top balance voltage number. I was shooting for 3.650v as that was the max SOC recommended by the manufacturer. A top voltage of 3.635v looks to me to be serviceable. The only issue I had was whether a very low charge at this voltage for "hours" would be necessary.

Again, I do very much appreciate all the helpful posts on this forum. I would not even be trying this without them.

[exMaggieDrum]

Originally Posted by SV THIRD DAY

The cells are much more tolerant to abuse than the panicked consensus would seem to indicate. So I wouldn't worry about what I view as a little hiccup...back away from the cliff..


What I do think this indicates again is the value in keeping things simple...and back away from the over engineering and messing with it. I have been involved now in setting up and sending out cruising about a dozen boats outfitted with LiFePO4 batteries and folks this isn't rocket science. It's really simple if you will just stop the overengineering compusion. For example, I chuckle at all this top balancing fuss, tie the cells together with 1ft 12ga wire and let them sit for a few days...how hard and risky is that? In setting up 13 boats its worked fine to date. The cells equalize and you move on to your install and go cruising. This crazy need to get back to 99.9997% full SOC is an engineering test question but not real life.

You have to ask yourself a question: are you guys building a science project or putting batteries on a cruising boat? Don't get lost in the process and remember the ultimate goal.

I have never discounted your views on this subject. It is an example of how difficult it has been to decide which process and installation to go with.

I think my project is part science project and part practical. "Over designing" it has actually been fun in many ways. But I'm retired and have a lot of time, no family, etc. You don't have that luxury. It is helpful to hear about the robustness of the relatively simple systems you have installed.

[OceanSeaSpray]

Originally Posted by exMaggieDrum

OceanSeaSpray, did you disagree with the suggestions made by others before on top balancing? If so I missed it.

Charging and top balancing is simple and easy. I have disagreed and explained why in the past, but repeating the same over and over again gets tedious, and even more so when the same people keep throwing the same claims and misinformation into the thread with no end to it to try and become less isolated with their brilliant achievements.

Charging your cells right up and leaving them results in slow electrochemical degradation. You certainly won't have "lost" them, but you should rectify ASAP, because you are very, very slowly damaging them and losing capacity and this is irreversible.

Now, more specifically to your case, you keep talking about controlling the "current", but current just does what it wants to do when you are balancing. Forget about the current. The only thing that matters at the end when balancing is cell voltage and the tell-tale that you have reached full capacity is when that cell voltage holds up after disconnecting. Don't bother about trying to reach zero current, but you will - naturally - if you are able to control the maximum voltage, like with a small, good quality PSU.

[SV THIRD DAY]

Originally Posted by OceanSeaSpray

you keep talking about controlling the "current", but current just does what it wants to do when you are balancing. Forget about the current. The only thing that matters at the end when balancing is cell voltage .

I think you just nailed the logic for using the Charge Voltage of your various charging devices (Battery Charger, Solar, Alternator, Wind Generator) to charge without the need for a computer controlled top end monitoring and cut off. As long as you set your charge voltages below the level that will reach 100% SOC, then bingo. You can't over charge, unless you have a charge controller failure and in that case a simple high alarm or high voltage cut out keeps you safe.

"But Rich..you won't get back to 100% SOC by charging that way. Good God man, you are leaving capacity on the table that you paid for"

Yes and well SO WHAT that doesn't hurt LiFePO4 batteries and it's also an intrinsically safe way to charge without relying on a control system because we know electronics never fail on a cruising boat. Is that a "brilliant achievement" as said by some? Na...I believe back in the day it was called smart engineering: Getting the job done the easiest and most economical way possible, with the least possible pathways to a failure event.

"Witch...Heretic...burn him at the stake...We need to spend more money on controls"

...or do we....

[OceanSeaSpray]

Originally Posted by SV THIRD DAY

I think you just nailed the logic for using the Charge Voltage of your various charging devices (Battery Charger, Solar, Alternator, Wind Generator) to charge without the need for a computer controlled top end monitoring and cut off.

Precisely NOT. I am talking about charging once to 100% full for balancing and while terminating the charge manually.

Originally Posted by SV THIRD DAY

As long as you set your charge voltages below the level that will reach 100% SOC, then bingo. You can't over charge, unless you have a charge controller failure and in that case a simple high alarm or high voltage cut out keeps you safe.

And here we go with another instance of the ******** shovelled with no end in this thread, because there is no such thing as a magic voltage that will charge, but not overcharge.
Not only this has been explained over and over again, but I had posted the graph showing that just a few pages up, so when it comes to the period of decay of information here... sad state of affairs.