Nissan Leaf Modules Li Conversion - Questions

Looking at a chart here: http://hybridautocenter.com/HAC4/index.php?option=com_content&view=article&id=72&Itemid=631

there isn't that much lost going down from 4.2V to 4.1V. From just under 4.1V down to around 3.65 it seems the curve is fairly constant sloping down at the same rate. Then below 3.6V there isn't that much energy left either.

That means, the bulk of the charge is between 148V and 130V, probably about 55Ah at the 1/3C (20A) discharge rate. The cells are 2P, so roughly equivalent to the same chart 40A, which is probably what the VX1 draws at low-ish speeds, I presume - have not researched how much it draws at various speeds. of course, as we should all know, 2x60Ah in parallel is *not* = 120Ah (hint: it's a bit more than that) :). I'm guessing the VX-1 probably draws around 40A during very easy riding on flat roads and relatively low speeds, so that chart is probably a good indication of the actual discharge during a fairly moderate riding style.

Each 1/2 cell according to this chart has about 220 Ah between 4.2V and 3.6V, so 36 times that is almost 8KWh of usable energy (with almost another 1KWh "reserve" where it is still safe to discharge down to just under 3V, but not worth it, in general, and probably shortens the life of the battery).

On my first charge I did 60km (35miles) going from 150V down to 142-144V in mixed driving. 144V puts me at the first 1/3 of the curve, down to around 4V and only 15Ah used. So there should be plenty more where these 35 miles came from - if the curve is accurate, it would be as many as another 60 miles of similar relatively easy riding to get me down to 130V where I would definitely want to recharge (though it would go a bit lower, I guess, based on your experience).

It's so nice not having to "pre-charge" every time I need to ride (the Li seems to keep charge better, the NiMh would self-discharge quickly if left alone fully charged). And also, to be able to take "the long way home" or make an unscheduled detour without fear I'd run out of range. I don't know what the actual range would be, but seems it will be at least 50 miles of good riding (more at lower speeds/more conservative riding). That range puts significantly more places within reach compared to the 25 miles I had before (with nearly 1/2 of them at reduced speeds to make sure I make it OK to my destination).

Again, thanks for answering my questions and sharing your photos and experience!

Good work Kocho and congrats on your convesrsion. Did you upgrade your main fuse, MC & Charger firmware.
I have my max at 148V and min voltage at 120V (3.3V/cell) which looks pretty much in the safe zone too. I am reckoning on about 140Km range.

Enjoy!

Fuse was already the new version (200A), MC is the latest (1017, I think), original ESD software works fine (can't charge in one go, but if I use half the battery or charge 2 times in sequence, it is OK, for now; I'm watching the charge voltage and stopping at around 148V usually - no need to pump the battery to 150V every time as I seem to be getting what I need from it between charges between just fine. I've not gone below 140V yet - just so happens where I find places and time to plug in, but probably won't be discharging below 130V if I can help it.

At 40Amps, in a flat road, you will be coasting at around 80Kph/50Mph.

That is from my experience with the FFM Software, that shows you the instant power drawn from the pack, among other things ;)

If this is correct, and the discharge plot is correct, and my math is correct too, it means that the VX-1 on Leaf cells should indeed be able to go for 125 miles at 50mph between about 148 and 130V. How do I get this? The chart is for 1/2 cell at 20A discharge current. We got two of these in parallel, so it should look about the same for 40A discharge rate. It shows about 2.5 hours cover the straight section of the discharge curve, between 148 and 130V, approximately. 2.5 hours x 50 miles per hour = 125 miles range on a flat road with no stops :), with up to another 10-20 moles possible if fully charging/discharging. Of course, real life range will differ ...

From 148V to 142V on my last charge I got so far about 30 miles of spirited city riding, half of it with a 130lb passenger for a load of 330lb (I'm around 200lb alone). With a linear discharge curve down to 130V, I can easily see another 30 miles will be possible. And that's with full or near full power when needed. If I ride like I rode with the NiMh pack (geeently), I'd get more than that :)

Again, I won't test the full range yet, not until I get my per-cell voltage monitors connected. Without them I don't want to risk damage to some potentially unbalanced cell, so I will stay in the middle of the charge-discharge curve.

Peter,

Are you referring to the 148km on the range display or actual 148km travelled? 148km is just under 90 miles. At low-ish speeds and calm driving I think that's realistic, based on what I'm seeing on mine.

So with 18 LEAF cells my 148Kms range when fully charged does seem realistic?

So, what "programming" specifications should we ask for when ordering one of these for a 18 cell conversion?

My ESD just failed last night and is not charging any more, so I might go with one of these if I can't find a cheap ESD or Runke replacement...

Hello Matt,
The best source for the elcon TC Charger is in Evassemble compagny, for 340$:
http://www.evassemble.com/index.php?main_page=product_info&products_id=17

And another quick FYI note on range. With 18 cells, I would stop riding when the voltage gets to 130V in use. Yesterday I rode for 60 miles city riding with lots of stop and go at traffic lights on streets with 25-45mph speed limit. Started from 147-148V (so not fully charged) and when it got down to 130V, it still had good power, but the voltage started to go down quickly, so I stopped. Based on the discharge curve, there isn't much lefty in the tank after that... I could see the voltage started to drop rather quickly, once I reached the 130V at about 40mph on flats. Just a mile or so later - I had another 2V drop.

So, the voltage decrease seems fairly linear from about 148 down to 130 or so, and dips quickly after that, just like the discharge curve for the battery module suggests it would. I'd say it would be good to start thinking about recharging by the time you start seeing 133-135V on the display while riding.

Archi, it is not the MC. The bike runs fine and I also have full instrumentation. I did use a light bulb for IRL, by the way. It is the charger that failed - see my other thread, I have a photo there.

I thought I'd post a quick update, in case anyone is still following this topic :)

I've put only a few hundred miles on the Leaf cells so far, mainly because my ESD charger died (nothing to do with the conversion, just happened - following the same failure patterns as many others here have shared). I replaced the ESD with a Runke, which had to be reprogrammed. Thanks to David from http://www.vectrixsupport.com , I now have a custom-programmed Runke for my 60Ah VX-1 Li conversion! It shows either the clock or the instantaneous current consumption while riding and at the same time shows the battery voltage. It's great to have both the current and voltage displayed at the same time! Makes you realize just how much power is being used in various situations and helps you learn to ride better and more efficiently. The fuel gauge is also working, though I have not fully tested it yet to tell if the bottom half of its range is as accurate as the top half has been. It is good that now folks with Runke chargers have an alternative to shipping their chargers and controllers to Fuel Free Motors in Europe for a firmware update - the software from vectrixsupport.com can be emailed to and installed by the bike's owner like any other firmware update.

But all this conversion work and charger trouble took time and so the bike has spent more time in pieces than on the road lately. Hope this will change soon, but the Summer is over, unfortunately...

The cells are holding-up well and are staying in very good balance so far, without a BMS.

Today I started to wire the male BMS harness extension (the female BMS harness is already wired to the battery). I terminate these wires to small 9-pin female harnesses like the ones for 8-cell R/C Li batteries. Each of these harnesses is about 1"x1/4" in size and very light. With these harnesses (5 of them by my plan) I can use a voltmeter to check individual cell voltages, or the total voltage of entire pack, or the voltage across a group of any number of cells in series.

I was planning to connect to these harnesses 5 small R/C voltmeters for Li cells. They are 1"x1"x1/2" in size approximately, and very lightweight too, so can carry one or all 5 in the bike. These can measure up to 8 cells each, so I was planning to use 5 of them simultaneously (to cover the 36 half-cells in my 18 cell pack). Unfortunately, at $2 a piece delivered from China, they turned out not as accurate as the specs say they should be. Out of the 6 that I ordered, one has 7 of its 8 channels dead on (with the 8th channel being 0.03V off), another has 6 of channels right, and the rest are not quite up to spec - instead of being accurate to 0.01V they are all over the map with some channels accurate and others giving +/-0.03V accuracy. Now, 0.05-0.06V is not bad at all if all we need to measure is the entire pack. But it is not good enough when one cares about 0.05V increments on each half cell - the total error could be as much as 1.8V across the pack, and slight cell disbalances will be hard to detect. In reality, these small errors turn out average out across the 36 channels, so the total pack voltage I get by summing the 5 voltmeters' readings is within 0.2V of what it actually is. I was hoping to use the 5 meters simultaneously, so I can at a glance each cell's voltage and quickly notice if a cell is unbalanced. But because these are not as accurate, instead I will use the best one and plug it in each of the 5 harnesses one at a time. This way I at least get both accurate and consistent readings across all cells. Perhaps I can rewire the harnesses to each connect to 6 or 7 cells (instead of 8), so I can rely on the one good voltmeter's 7 accurate channels.

Anyhow, the bike is now what it was supposed to be in the first place: good range AND good power. With 60-80 real-life miles of range, now almost all areas near where I live are within reach without the need to plug-in. This range covers all of my regular daily commuting needs. Plus I can ride more spiritedly or take a detour without too much worry of running out of battery before I get to my destination. Plus, the pack does not lose charge noticeably by just sitting for a few days - so no need to worry about charging immediately before riding, like I did with the NiMh pack.

Double-post (deleted)

An observation and question on the possible "self-balancing" property of the Leaf cells. Could they actually be top-balancing themselves during the normal charging process? And what might be the implications?

Yesterday, after I had ridden about 50 miles over a couple of days (many of these miles were highway, I measured my cell voltages immediately after I came off the bike. I measured as much as 0.1V difference between the lowest and the highest voltages (between 3.5 - 3.6V per each of the 36 half-cells). Letting the bike rest overnight, in the morning the differences were down to 0.05V between cells. Plugged-in to charge and when I got back 9 hours later, the pack was fully charged and the charger had shut town, probably a couple of hours earlier. Measured all cells and they were all within 0.01V! In fact, all were 4.07V and one was somewhere between 4.06 and 4.07V (my multimeter only has 2 digits on the display and it changed from 4.07 to 4.06V while I was looking at it).

So it appears, that when charging, their voltage got virtually identical even though they started from different voltages. I need to ride again and measure immediately after the ride, but with the batteries near 4V rather than 3.5V this time and see what voltage differences I measure.

I know it is not a very reliable way to measure the voltages at rest, but there seems to be an interval immediately after riding where I measured significant differences that later decreased but did not disappear - so some conclusions could probably be drawn...

FYI. My charger does CC at 10A to 146.7V actual (148V on the display of the bike). It may take 2-3 hours starting from the battery depleted to 130V or so at rest. Then holds the voltage steady at the 148V mark during CV for another 2-4 hours or so, decreasing the current gradually and ending the charge when the current falls to somewhere below 0.5A.

Thoughts?

Follow-up on the "self balancing". Came back from a ride today and measured all cells immediately after the ride (last full charge was a day earlier and I have not plugged after that to top-up). All cells were within 0.01V at 3.93V each, so very well balanced! So, the 0.1V imbalance I measured previously is now completely gone! Again, I am measuring at rest, so this may be masking things, but as I mentioned, also at rest I had 0.1V difference the day before and now that difference is gone...

Thoughts from the more experienced with these battery type?

Hi Kocho,

It is perfectly normal as the curve voltage / "state of charge" of this cell has a very little slope at end of charge while a big slope at beginning of charge.
It means that the same difference of state of charge makes more voltage difference around 3,5V than around 4,07V.
It is not a self-balancing effect.

Wouldn't it be the other way around? Edit: you probably meant that the right side of the curve has more slope than the area on the left. But the slope is still steep on the very left end (above 4.1V). The chart shows the curve is steep at near full and near empty SoC, so a small diff in SoC near and above 4V would result in a big voltage difference? Edit: but a good point nevertheless to look at the curve: 4.05V to 3.65V is a low slope, so they would appear better balanced there, where above 4.10 and below 3.60V any differences will be easier to see. Since I never charge above 4.1V I won't see the differences. But I might discharge below 3.6V and see then.

And one more edit: wouldn't a 0.05V difference at 3.5V be preserved as the cells are charged and remain there at around 4.1V? So if a 3.50V cell charges to 4.10, wouldn't a 3.55V cell charge to 4.15V when bulk-charged in series? I did not observe that.

no, because each cell has a different capacity

I see your point. More specifically, because the charge-discharge curves are not horizontal in the middle but have a significant slope, the curves of the individual cells will not overlap - they will intersect at that one point, at the voltage at which they are balanced.

So, it appears that my cells are fairly "top-balanced" then, since I don't get any measurable disbalance at around 4.1V (at least not measurable with my v-meter that displays 0.01V precision).
.

I would like to buy this modules, but where?

Hmmm, wasn't that mentioned up somewhere in the previous 100+ posts :)

I got mine (as I think did just about everyone in the US) from Hybrid Auto Center: http://hybridautocenter.com/HAC4

Availability varies, so be ready to buy when they have stock. The '13 model year is with the new chemistry (enhanced for hot and cold weather) and they also sell the older '11 and '12 versions. The '13 cells look physically a little different from the older ones, but are of the same dimensions.

EDIT: that said, if you are not in a hurry, used Leaf batteries do come from junk yards in the US from time to time. I missed one pack by a couple of days while I was still unsure if I want to upgrade. The price was $2,400 (plus shipping, or you pickup) from the junk yard for a fairly new '13 pack. There are 48 of these modules in one pack. You need 18 or 19 (per your preference) and you can easily resell the remaining for close to $100 a piece, so you will get your 18 for free, practically. I opted for "new" cells so went with Hybrid Auto and they delivered what they advertised - very flexible to work with too and I got all the parts I needed from them (the BMS wires and harnesses, bus bars, compression plates, rods).

FYI. A quick follow-up on one of my previous observations about disbalanced cells at low voltages that become balanced at mid/full voltages.

Today I had to push it a bit further than anticipated, and ended my ride with a total pack voltage of 115V. That is about 3.19V per each of the 36 pairs of cells in the 18 modules in my pack. Immediately after the ride, I measured the voltage of each of the 36: it ranged from 3.12V to 3.30V (only one or two cells were at these extremes, most were within a narrow 0.05V range of difference).

Recall the charge-discharge curve (larger view here: http://hybridautocenter.com/HAC4/index.php?option=com_content&view=article&id=72&Itemid=631).

3.1 - 3.3V is well within the steep right corner of the curve. That is about 5Ah difference in SoC and that translates to almost 10Ah difference in SoC between a module that is at 6.2V and one that is at 6.6V. So, while quite safe and with enough reserve for a few more miles, this is probably as far as one should take these modules on a regular basis (and higher would be better, I guess).

Another observation is that one can actually measure these differences after a ride: I spent several minutes measuring with a multimeter by hand each of the 36 pairs of cells (through my BMS leads). The change from when I started to when I stopped was only 0.01-0.02V or so. So, it takes more than several minutes for the cells to recover much compared to when I stopped riding. The total pack voltage was virtually unchanged over that period of time.

One final observation: the voltage does begin to go down dramatically quicker after it hits around 130V pack voltage. This I think is the practical Low Battery warning point, with a good reserve of perhaps 10 miles of worry free riding left at moderate speeds to get to the 115V mark. Power is still there and the pack voltage does not sag much when I drew 80-100A from it briefly at that point, so it is still safe to ride - no cells were empty or near reversing. Below 115V I think we should consider to be in "limp" mode. Ride it gently, and try to stop riding as soon as possible, to avoid damage to cells when the pack voltage goes below 100V (and some cells might be near empty at that time).

I'm charging now, so will update this post in a few hours to confirm that the cells are back in balance when fully charged. If I get a chance, I check voltage when the cells are a bit higher, say around 3.5V each, then again near full, to see at what point does the 0.2V disbalance disappear.