[dtbaker]

I have been doing an initial top-balance charge of 38 x 100ah TS cells wired in parallel at 3.7v for....6 days now, and still going at 5amps or so. total energy added per kill-a-watt meter has been 7 kWhr.

Seeing as how this is a little over 50% nominal capacity pumped in, with no signs of slowing yet, I am wondering what other peoples experience s with the initial balance charge of a new pack.

I don't really know how long these cells were sitting around, but probably not that long, so I am surprised that I have put in more than 6kWhr. What have YOUr experience been for the initial charge?

 

[GizmoEV]

I think I put something like 14kWh into my set of 40 when I got them. Remember that you are only putting 0.13A average into each 100Ah cell. That is a very slow charge. Mine didn't take 6 days, however. Probably because I was using a 10A charger.

What is the terminal voltage right now? There is very little energy above 3.4V. I did a test where I recorded current and voltage by hand of a 100Ah cell starting at 3.40V and ending at 4.00V and got at most 0.65Ah into it. Because of the way I calculated Ah this number would error on the high side.

 

[dtbaker]

I think I put something like 14kWh into my set of 40 when I got them.

hhmmm, must be a lot of loss at the plug, and surface charge that is not really sinking IN to the cell....

Remember that you are only putting 0.13A average into each 100Ah cell. That is a very slow charge. Mine didn't take 6 days, however. Probably because I was using a 10A charger.

my power supply is listed at 10a max. started at about 8amps, and is hovering around 5a all day today.

What is the terminal voltage right now?

... well, the power supply is set to 3.7v, and still reads 5amps. If I measure across cells, I get values between 3.36 and 3.38 .... before charging they were all around 3.30

 

[dtbaker]

Lemme come at this from a different angle to get a feel for how long I can expect this charge to take. Lets assume my 100ah cells were only 50% charged from factory, which I think is about where they say they are? If each cell is going to need around 50ah, then if I were charging them SINGLEY with the power supply it would take at least 10 hours and maybe more as the rate slows toward end of charge. Am I on track with this?

Then in parallel, it could take 38* 10 hours (minimum)... like TWO WEEKS! ?!

...and the total kWhr that my meter is showing is not really all getting into the cells, so I might see a total of 12 or 15kWhr at the plug to get the 6 or 7 kWhr into the cells. Correct? or, not?

I am thinking it may be worth my time to temp wire in parallel and pump some juice in with my on-board charger until the pack is 'close' to full and seeing a little voltage rise say up to 3.6v per cell, then go back to the parallel to finish and balance to 3.7 or 3.8 for the final top balance? ....

comments?

 

[GizmoEV]

That could work. When I watch my cells voltage after they reach 3.45V they start rising rather fast. This is with a relatively low current like 6A. (Obviously if you were to pump 100A in them the terminal voltage would be a little higher even though SOC wasn't higher.) While are around to watch the cells you could raise your power supply voltage to a higher value to get the current up so it would charge faster.

As for charge efficiency, these LFP cells are quite efficient compared to lead acid. I would think that as the charge rate diminishes that a higher percentage of the energy coming out of the wall is going to be spent on charger cooling and other "waste" energy than what might be lost while charging at a higher rate. As a comparison, with my lead acid pack I could go 4 miles/kWh out of the wall where as with the LFP cells I go 6 miles/kWh.

If you have a BMS which will tell you when a cell reaches a high voltage point you could just hook up all your cells in series and use your bulk charger to get them to that point and then reconnect them in parallel again and finish it off. Just be careful of that top end voltage. It climbs really fast. Remember, after 3.4vpc will only take about 4 minutes to fill them up to 4vpc if you are charging at a 10A rate so if one cell reaches full ahead of the rest you could end up way over charging it. With my Zivan NG1 it only takes about 10 minutes to finish filling my 200Ah pack once the charger switches into constant voltage mode.

 

[Overlander23]

Hehe... now you're discovering why I went series with my 5kw charger for a bit, and then back to parallel.

At 3.7v and 10amps you're only talking 37 watts, and as you mention not all of it is getting into the cells. Even though my charger is capable of 5kw, I only had access to 110v so I was only pumping only 2000w into the series pack. That still took several hours... and then two-three more days of paralleling at 3.8v with my 20 amp Mastech. Otherwise, I was looking a two to three weeks of straight paralleling.

If you go the series route, just be sure to check the cells regularly to prevent any one going above 4.0v. Then, either bypass it, or stop the series charge and wire in parallel again. When I did it, there were a handful of cells that hit first. After bypassing the first three, I rewired in parallel and just let it all sit there on the power supply at 3.8v. But 3.7v works, too.... It's really wherever you want them to all line up, as long as it's at or above 3.6v.

 

[dtbaker]

ok, I re-wired cells in series and hooked up the regular charger (elcon 1500), and it ramped up to 11.5 amps and chugged away. I let it go for about 10 minutes while rounding up my multimeter and then started checking cells individual voltages.... which were all over the map.

The terminals the charger was wired to showed 138v (as expected w/ curve 611), and the charger draw from the wall started dropping from 11.5 as expected. BUT a couple cells were already at 4.0, 4.1, while most of the others were down at 3.55-3.60. So, I unplugged charger before the high cells cooked.

It is now apparent that in addition to my mastech power supply I was planning to use to bring up individual cells on the low side, I will also need a decent, easy, cheap way to drain a little excess off the individual cells that are too high for this initial top-balance. Since there were only a couple high ones, and a lot of low ones, I am thinking it will be way faster to drain a little off the high ones and re-start the pack charger.

Anybody have some great ideas/examples of cheap resistor setup to drain a little juice? Can I just wire up a 100 watt incandescent bulb across an individual cell for a couple minutes to make enough difference?

 

[GizmoEV]

With them all at or above 3.55V it looks like they are all close to being full. I'd hook them up in parallel again and finish them that way. The ones at 4.1V will drop in voltage quicker than the ones at 3.55V. Also, if there is any way you can throttle back your charger you might find that the 4.1V is a little higher than rest voltage. You could hook a lightbulb in series with your charger to limit charge current if you can't do it any other way.

 

[dtbaker]

I found a local electronic supply place that had some big resistors ( 1 ohm x 25 watts for $1.50 ea), so I got two and put them in parallel to use as a 'load' to drain some juice from the high cells.

This is kinda time consuming, I spent a hour taking a first pass cell by cell... to check each cell, drain for a minute on the high ones, go back and see how things recovered.... I should be able to take another pass a series charge to see how things are at the top end. When all cells finish within .01, I'll re-wire in parallel and let the cells auto-balance that last little bit while I work on revising battery racking.

 

[dtbaker]

...getting frustrated w/ lack of success balancing... I went thru the pack several times with my resistor, and discharged the several that were higher than most until they were all 3.34-3.36 . Then, I applied my series charger until the first ones hit 4.1, which was only about 10 minutes. Overnight most have relaxed to 3.34-3.35, but the same ones that were high before are high again at 3.40, 3.42ish

so... is it time to go back to the parallel power supply and trickle it up to 3.70?


d

 

[EVfun]

I'd knock an amp hour out of the high cells, that is about 18 minutes with a 1 ohm resistor, then check the resting voltage the next day and then put them on a series charge for a few minutes and see where they go. 138 volts for a 38 cell pack is only 3.63 volt per cell. At the end there will be some scatter but you should be able to leave them with on charger for an hour after hitting the voltage wall while the current tapers down (naturally, I don't mean walk away during the initial charging.)

 

[GizmoEV]

...getting frustrated w/ lack of success balancing... I went thru the pack several times with my resistor, and discharged the several that were higher than most until they were all 3.34-3.36 . Then, I applied my series charger until the first ones hit 4.1, which was only about 10 minutes. Overnight most have relaxed to 3.34-3.35, but the same ones that were high before are high again at 3.40, 3.42ish

There is so little energy above that point I'd go parallel and top them up to 3.6-3.7V. It will take a fraction of a second to bring the high ones down to the low ones and so little energy that I doubt that you will even be able to detect any spark on connection. That is less than 0.1V difference between your extreme high and extreme low cell. Unless you have shunting BMS boards and can put them on and throttle back the charging current to below the shunting current in which case I'd do that.

If you stick around you can turn up your bench-top power supply above your cutoff voltage to get it to pump in at its highest current until the terminal voltage is where you want it. Then just taper the voltage back to your target voltage and let the PS finish its job.

 

[dtbaker]

I'd knock an amp hour out of the high cells, that is about 18 minutes with a 1 ohm resistor,

hhhmm, maybe I wasn't taking the high ones down enough... they seemed to keep bouncing back almost to where I started. I have two 1 ohm 25 watt resistors in parallel. I guess I will try one more round of balance, another series charge, and then go back to parallel to finish up.

 

[dtbaker]

I was struggling with repeat loading and charging... The high ones I'd let drain for 15-20 minutes at 50 watts. Then when I fired back up the series charge I was seeing a couple hit 3.7 and rise rapidly to 4.0v while the rest were still sucking up juice at 11 amps and sitting at 3.50 v or so. It was getting very time consuming and not seeming much better balanced, so I rewired to parallel and am letting it trickle along unattended with CV=3.70

Unfortunately, it really is just a trickle with bench supply. Only about 3.8 amps with CV=3.70. With 35 watts going to the pack, this could take a LONG time.

Would I blow things up if I hooked up the series charger, that would kick out 11amps, and watch it like a hawk? Or, would I accomplish the same thing by cranking up the power supply to its max CA (10amps) and watch the cells manually while they are in parallel, and back down to CV when I start seeing cells up around 3.55-3.60?

 

[rwaudio]

I was struggling with repeat loading and charging... The high ones I'd let drain for 15-20 minutes at 50 watts. Then when I fired back up the series charge I was seeing a couple hit 3.7 and rise rapidly to 4.0v while the rest were still sucking up juice at 11 amps and sitting at 3.50 v or so. It was getting very time consuming and not seeming much better balanced, so I rewired to parallel and am letting it trickle along unattended with CV=3.70

Unfortunately, it really is just a trickle with bench supply. Only about 3.8 amps with CV=3.70. With 35 watts going to the pack, this could take a LONG time.

Would I blow things up if I hooked up the series charger, that would kick out 11amps, and watch it like a hawk? Or, would I accomplish the same thing by cranking up the power supply to its max CA (10amps) and watch the cells manually while they are in parallel, and back down to CV when I start seeing cells up around 3.55-3.60?

When you are draining the cells, I assume you are using the pair of parallel 1 ohm resistors. I'm taking a stab that the voltage under load is 3.3 to 3.4V, if that's the case you are only drawing 6.6 to 6.8Amps which works out to ~22-23watts not 50watts. (that's good for the resistors though)

As for attaching the series charger, I think your asking to blow something up then, best case the charger thinks it's a short circuit condition and turns off. It's just not going to work down to that voltage, I'm sure there is an output range on it somewhere.

I've been using some of these http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=380195053685&ssPageName=STRK:MEWAX:IT
in parallel adjusted upto 3.65V to charge some headways I'm testing. I put 4 dc/dc converters in parallel to charge 6-12 parallel 10AH headways and they charge pretty quick. It would be easy to parallel more for even higher currents. I use a 48v 60A power supply to power the dc/dc converters but any good size 36-75V supply would work.

 

[dtbaker]

ok, I will continue to be patient in parallel.... sigh.

 

[GizmoEV]

Or, would I accomplish the same thing by cranking up the power supply to its max CA (10amps) and watch the cells manually while they are in parallel, and back down to CV when I start seeing cells up around 3.55-3.60?

That is what I suggested the other day. It is also what I did when I was testing out a single cell so doing that on your paralleled pack it would be even easier because the voltage won't move that fast.

 

[dtbaker]

FINALLY, amps are dropping and pack is looking almost full in parallel. maybe just a couple more days.....

 

[dtbaker]

whoo hoo! amps dropped to Zero.... so parallel pack charge is done. I will unhook power supply, and let them sit in parallel for probably a day while I work on other stuff.....

I am tempted to re-wire in series on the bench, apply a load to drain a little, and then use my series charger to make SURE they finish re-charge in balance before I have them all wired up in the car. Am I being too fearful? Is it worth the hour to rewire on the bench and verify charge balance, or am I being overly chicken?

 

[dtbaker]

pack is relaxing a little, and still within .01v across all cells. yay.

I had set the power supply to 3.8v, which 'finished' at 3.83-3.84 according to my multimeter. A little above where I hope my series charger will take things (3.65) if all goes well. I decided to go a little above to purposely get to the steeper part of the curve to nail down the balance as closer than where the curve is flatter.

I left things in parallel for about 24 hours now, and cells have relaxed to 3.73-3.74. I won't be working on anything today, so plan to leave in parallel for a while longer.

still debating on whether it is worth some piece of mind to take the time to wire in series, load the system to bleed a little, and use the series charger to verify that they remain in balance and hit the finish at the same time....