What's up

It's been a little slow on this blog lately. Which doesn't mean I haven't been busy. Let's start with the converted Xsara. It's 12 V aux battery woes have continued. Also contributing to the issue is your's truly, forgetting the vehicle running and so on. So it seems I killed another lead acid battery. Maybe it's a hint I really should take and just not use them anymore at all anywhere again. To replace it I put in four of the worse 40 Ah SE40AHA lithium cells I had around. Worse because they don't really keep their voltage really well and are at least a little soft shorted inside, but not completely.

With those four cells in the can serving the purpose of aux battery I set off to the local inspection office to have the vehicle checked, as is required every year. This time I had no less than three officers looking under the hood and asking questions about the electric powertrain. The brakes have seen better days, but the car still passed. I should still take them apart and at least try to salvage the discs.

It seems there were no pictures taken of the battery swap or the inspection itself, but there are pictures of something I did a little earlier. The repaired 87.6 volt charger gave up the ghost again and this time I pretty much gave up on it. Again, I suppose, but not willing to pour any more money into it.

So thinking about how to solve the issue again, I decided I'd add some more batteries into the vehicle. It has 60Ah cells, but since I have those abused 40 Ah cells around from the motorcycle, I decided to augment the pack with a couple of them in parallel. Turns out I can just about get away with a string of 27 cells with the AXE 7245 controller in the car. It has a maximum voltage of 90 volts, but it will start if the voltage, as presented by Cycle Analyst, is less than 91 volts. With the 27 LiFePO4 cells in series and fully charged, I end up a little over 91 volts, but if I let the DC/DC converter run for a bit the AXE will relent and start up. This gives me about 400 Wh more to drive around with bringing the pack total a little over 5 kWh.

It's a bit of a mess, but there they are. I was sort of hoping I could use my 102.2 volt charger directly when increasing the cell count, but I had to put into a bit of monitoring in the end. It looks at 20 cells in the middle of the pack and makes sure they stay within 3.5 volts on average and then it tells the 102.2 volt TCCH/ElCon charger to stop charging until the voltage goes back down. Not optimal, again, but it works for now.

On the C-Zero I received a shifter cover panel from Ben Nelson, who has a flooded i-MiEV his parting out now. It took a little bit of fiddling to get it in the car, but I managed to do it and now it really looks the part. Thanks a lot for part, Ben!

Last, and probably least as well, but still quite important is the addition of two new USB charging ports I added to the C-Zero. I can turn them on or off with the switch. They're always connected to the car's aux 12 V battery, so I can provide power to some devices even though th car is not running. Very handy for CaniOn, which I've also configured to turn on when the phone which is permanently in the vehicle gets power, so I can always easily see what's really going on with the battery.

More "gears" to our C-Zero

Thanks to the great great folks at My i-MiEV forum I got a tip that PSA may actually have just mechanically blocked two of the shifter positions on the Citroën C-Zero and Peugeot iOn. It didn't take much convincing to make me crack open the cover on mine and behold, it was true. I quickly found out that even the screen would show the new B and C modes. On with the drilling and I had some new "gears" to play with.

I'll still need to modify or acquire a suitable cover plate with all the shifter positions.

A quick check using CaniOn and it looks like the difference from B to C is about 10 kW in regen power, with B having the strongest and C the least regen. Normal D right in the middle of those two. Letting the accelerator pedal go at 70 km/h gave about 10 kW regen at C, 15 kW at D and 20 kW at B. I finally have the low regen "C" option for cruising long distance. Good times!

Back to the original charger

The cheap charge controller I had installed gave up the ghost - or more like stopped reading the pack voltage correctly - and the next option, installing a PowerLog 6S to monitor the pack, also ended up with some smoke escaping and thus rendering that device inoperational as well. They run on the magic smoke, as you know. I now have some more electronic junk to be taken to recyling.

Cheap and readily available options exhausted, I figured what the hell, perhaps a local electronics repair shop could fix the original charger. The one that actually charges to a correct voltage right away. And what do you know, a couple of days in the care of Porin Radiohuolto and I once again had a working charger. Had to depart 85 euros for the job, but it's well worth it. These chargers do run at least around 500 € from China.

It had blown a fuse and a couple of caps had swollen, so they had replaced those. I had already suspected that I might have in fact aided in it's downtime by having it connected to the pack permanently, even while driving. It may have gotten on those caps a little and caused the fuse to blow.

So installing the charger back into the vehicle now, I added an 80 amp 12 VDC contactor, which only closes and connects the charger to the pack when the car is plugged into the mains. In the same power inlet as the main charger I just have a 12 VDC PSU originally from a Linksys router, which activates the contactor. I had a timer circuit as well, but decided to leave that out, just to keep it simple. In addition to the charger and the PSU there is also a aux battery charger, which keeps the small 12 VDC lead acid battery of the vehicle topped up.

Added insulation

I've been planning to add some insulation to the doors of my C-Zero. Insulation for the cold winter and also some sound dampening to reduce noise. Today I finally got it done. All four doors got some 1 cm thick foam with a sticky side and a side with aluminium foil.

Driver side front door before and after.

Passenger side rear door. Not as pretty, since I used leftover pieces.

Same door. Now you can see the foil in the unused speaker opening instead of cold steel.

On a quick test drive I noticed two things. The sound of the electric motor was more audible than before. I guess when you reduce other noises you start hearing others. I also noticed that the bass frequencies sounded louder on the car stereo. To be expected, since adding the insulation makes the doors perform better as speaker enclosures.

I'm also planning to add some anti-stone chip shield behind the front wheels. There doesn't seem to be much to speak of. It could also make the car quieter. At least concerning noises made by flying stones, if nothing else.

Update on charge controller

Noticed something nasty with the charge controller setup in my previous post. I really should have thought of this, but the 100V voltmeter is not isolated. It will connect your high voltage pack negative to your 12V negative, which is your vehicle ground. Not cool. In fact so not cool, that it seems to have killed at least one, probably two 12V lead batteries I used in the vehicle as the aux battery.

So, lesson learned, again, always check your pack isolation after you install anything that connects to it. Check voltage between both positive and negative ends of you high voltage and the vehicle chassis. If you see the pack voltage or something like it, you have a problem. If you get some fluctuating, small voltage, you probably don't have a problem. Note that with a brushed motor, you can end up having dust from the brushes build up and create a short, so it's not a bad idea to check even if you don't install anything new.

Charge controller

Went ahead and installed the simple 100V solar controller. It takes 12 volts and the pack voltage. When pack voltage reaches a set value, it will disengage a relay. It's set to disengage at 88 volts, which is a little over 3.5 volts per cell. The relay then controls the Elcon/TCCH charger's Enable pin, which will instruct the charger to stop charging.

Of course as the pack voltage drops, the controller will re-engage the relay and the charger will resume charging. This will make the pack voltage bounce back and forth pretty much forever, so it's not a perfect solution, but it should make sure the charger doesn't overcharge the pack right away. On the other hand there's no CV phase, so the bouncing is kind of necessary to get the pack full.

In any case, this solution will have to do for now.

It's dead, Jim

First things first. Godspeed, Mr. Leonard Nimoy and GNU Terry Pratchett. Also dead is my 2kW KP charger. Coincidence? Maybe, maybe not. In any case, I'm a little short on chargers (and great minds). Luckily, I have the Elcon/TCCH 2500W charger from my temporarily defunct electric motorcycle.

As it happens, the Elcon charger is set for 102.2 V, which is way too high for the 25 cell pack in the car, but it will still start charging although the initial voltage is lower than expected. For the 25 cells the 102.2V would be about 4.1 volts per cell. Not into dangerous territory, but not very good for the cells either. So you either need to watch the voltage carefully while charging or device something to end the charge earlier.

Knowing my tendency to forget things, an automatic shutoff would be much safer. Luckily I've had some experience with one of these voltmeters. It can measure up to 100V and control a relay based on that, among other things. I have one in my bottom balancer, but I'm reluctant to take that well functioning setup apart, which is why I've put another on order from ebay. Once it arrives I'll set it up to control the ENABLE wire on the Elcon charger, so that it will end charging at 87-88 volts. Actually it will bounce back and forth, but at least it will give me more time to pull the plug.

Or, as I just recalled, I also have a CellLog 6S, which I haven't come up with a use for yet. I could also set that one up so that it would monitor five series of five cells and control the charger as well. In that case the cutoff voltage would be 17.5 volts for five cells. It would even be more accurate, since it would monitor five cells at a time, instead of the whole 25. Certainly something to consider. On the con side, it does require more wiring, but maybe it would be worth it. We'll see.

Technically, the Cycle Analyst display could also do the job, but it would require opening the CA case, adding some wires, using a solid state relay and so on. A lot more work, all in all.

Update on cell testing

Just a quick update regarding my previous posts on the SE40AHA cells that seemed to be fine after all. So far, in testing the cells by draining them to about 2.6 volts I've found that out of the 33 cells I have, nine seem to have internal shorts. I can tell this by leaving the cells at 2.6 volts and monitoring their voltage. The bad cells have gone down in voltage, some already below 2 volts. They must therefore have internal shorts. The better cells have bounced back a little and then kept their voltage.

So it looks like I might have 24 usable cells. Unfortunately this doesn't really match anything I have in use. The car, Xsara or kWsara, has 25 cells in series, and the motorcycle, kWsaki, is set up for 29 cells with it's 102.2 volt charger. So it's either figure out something else to do with the cells, get a new charger perhaps, or try to find a couple of SE40AHA cells. Unfortunately GWL Power doesn't list them anymore. They are a couple of generations old now, so it's understandable that they'll be unavailable. Sinopoly has similar black 40 Ah cells, which might do just fine, though.

Regarding winter range

One of the questions I often get asked at the quick charge stations during the winter is how does the winter affect your range. It depends. That's the answer. Turns out, it doesn't just depend on one thing either. It depends on a lot of things.

The most obvious one, and the one people are most aware of, is the fact that current batteries perform better when they are in optimal temperature. That's usually around room temperature, or perhaps body temperature. Much above and you get into trouble. Freezing and below charging becomes slower and they're not quite as willling to part with their charge either. How much this matters on my C-Zero I don't really know. I've done most of my long trips in less than optimal conditions, so I don't really know well it will perform come summer. If I'd have to guess, I'd say it's in the 10-20 % range.

The second rather well known factor is heating. Electric cars tend to use the electricity in their batteries not just for driving but also for heating. In the C-Zero the heater can use around 5 kW. With a 16 kWh battery, with some usually left in reserve, it's not too hard to figure out that the heater can suck up a third of the range in an hour. Generally when I have 100 km range on the gauge, turning heat on will drop it to 70 km, which is probably quite accurate. So I'd say 30 %. In my car I've installed an ethanol heater, so I don't have to suffer this drop in range or the cold either. Some newer vehicles also have heat pumps, which should be able to reduce the consumption to around 1 kW in optimal conditions while still providing enough heat.

Some people are also aware of rolling resistance. Winter tires are designed to grip on snow and ice. And you can't have perfect grip with perfect rolling resistance. In other words, your summer tires will roll easier and winter tires will decrease your range. My guess, 5-10 %.

What I haven't seen anyone recognize is air resistance. If I've done my math right, going from room temperature to something like -20˚C can increase the air resistance or drag as much as 10 %. Simply because cold air is heavier than hot. That's quite a bit for something most drivers don't even think about.

I'll leave it to you to figure out how much all of these could dimish your range.

Lessons learned

A couple of things I learned from the events described in my previous post.

1. Cells at zero volts may not be dead after all
2. Cells at zero volts will not charge from a full cell connected in parallel
3. Bloated cells may actually shrink back to size
4. Bloating does not automatically mean internal soft shorts
5. CALB LiFePO4 cells can take hell of a beating

Point number two also casts a shadow on parallel bottom balancing. I think it probably still works if you start with cells in roughly equal voltage, definitely all cells above the bottom balancing voltage, but it will not cause lower voltage cells to rise to a higher voltage (or charge from it's buddies).

Magic recoveries

Remember the 29 cells in my motorcycle I destroyed last summer, when I accidentally left it turned on and the DC/DC converter has drained all the cells to empty and beyond? Something interesting has happened since. Actually, two interesting things.

I was actually going to take most of these cells in for recycling, but I just hadn't got around to it. Instead I had just left them in our basement. Before storing them I had connected them all in series and drained them down to 2.6 volts, figuring that they would bottom balance together and I'd see which cells had internal soft shorts.

Now the eight newest cells I had just bought to increase the cell count to 29 were at zero volts after my little mishap. I had also connected them in parallel with the others, figuring that they'd recharge from the other cells if they had any life in them. They hadn't and I had left them for dead at zero volts.

But before I get to those eight newer cells, let's get back the 22 older cells which had bloated. Actually 12 of them hadn't and I had connected them into three 12 V packs for whatever use I'd come up with later. I even tested them and each 12 V pack had more or less 500 Wh of capacity, so I thought these 12 cells can be saved for later use and the rest scrapped.

The remaining 10 old cells were pretty well bloated, but they seemed to retain voltage, at least for a while. To my no small amazement, they still had the same voltage after about six months of storage. All ten cells were around 2.55 volts. What that means is that they can't have any significant internal shorts. If they had, they'd keep going down until they're at zero volts. Alas they hadn't. What's even more amazing, they had lost their considerable swell and are now pretty much right size and shape! The had actually shrunk back to shape.

Back to the eight newer cells I'd left sitting for half a year at zero volts. I had even been so sure of their fate, that I had liberally sprayed some gold paint on them to mark them as the ones I'd definitely take to the recycling center.

After discovering that the older cells had gotten back to shape and were still holding their charge, I thought I'd give these cells one more chance too. What if they just didn't want to charge from the other cells while being in parallel? I made two 12 V packs of them, measured a voltage of about 0.5 V for both whole packs and connected a 14 V 5 A power supply.

What happened was that the voltage started to steadily rise as one might expect, so I just left them charging for four hours. When I got back, I disconnected the power supply and checked the cell voltages. They seemed normal, about 3.33 V, as they should for just charged, half full cells. Again I left the cells, figuring they'd probably start draining themselves. I was, again, quite amazed to find out they had not, but kept a voltage above 3.2 V instead. I repeated the same for the other pack with identical results.

These eight SE40AHA cells which I had left for dead and stored at zero volts for over six months were back as well! Now I don't expect the cells to have their full 40 Ah capacity left and I haven't tested them with a load, but I will bottom balance the rest of the cells now and connected them as a 29 cell pack. Charging them together will reveal if any cells shoot higher too early and what the remaining capacity of the pack is.

Right now it looks like bottom balancing saved the day after all.

Check your connections

I've been aware of the possibility of cell connection bolts loosening over time. There are ways to combat this, such as using braided straps and nord-lock washers. Having been reminded of these techniques by several recent YouTube videos, I thought I'd check my cell connections.

I've been cheap, as usual, so I've just used the plain copper interconnects supplied by GWL Power along with the bolts and washers which come with them. It's also been nearly a year since I put everything together, so I was interested to find out if they had indeed loosened. I had also noticed that the pack temperature does rise quite a bit while driving. No visible signs of extra heat were noticeable though.

Turns out not only where several bolts quite loose, I had even forgot to put in all the washers. So yeah, don't forget to check your battery bolts regularly, especially if you don't use nord-lock washers.