Batteries


So after the last deep discharge of the nicad pack a month ago, the pack has been very soft lately. The voltage drops quickly long before I get to work. It is winter, but tonight I found out something new. I pulled the lid off of the pack and was wondering if the 24 redtops were way over charged compared to the rest of the pack, due to their higher capacity. This would allow the pack voltage to rise quickly and not fully charge the rest of the pack. So much for that theory. I took my discharge equipment and applied a 55 amp load for 12 minutes. That discharge took out exactly 11ah out of those 24 redtops. When they were almost finished, I checked them with a DVM only to find that many of them were nearly at zero volts! No good. So I put the constant current power supply on that group of redtops at 10 amps for 3 hours. So they took on 30ah worth of charge. Some cells were venting liquid which means they were over filled. I had refilled them a while ago, not knowing they were actually low on charge, which makes them low on fluid as well. Hmmm. So I’ll look forward to a stiff pack tomorrow morning on the way to work.

This past Tuesday the truck got another deep discharge on the nicad pack. The range was low, but it had also been cold lately. The pack went from -4kwh to +5.3kwh or -17ah to +29ah. This is encouraging. On the 5.7 mile drive to work, the pack voltage never went below 310 volts. This is also a good sign.

I’ve also been working hard on an analog Lithium BMS for 160ah cells. Later, it will be adapted to 40-90ah cells. The board has been iterated a couple of times. Each time better ideas were added for the layout. This last version needs to be built yet as it had some serious changes. The first version is built on a breadboard and works fine.

This deep discharge should be done at least one more time. The nicad manual states that sometimes up to 3 deep cycles are required to completely erase any memory issues. I will check the entire pack for water levels this weekend. A spot check showed that most levels are ok, but were checked while the pack was fully discharged. It has to be checked when fully charged.

Last night I took a drive on the highway. It was obvious the nicad pack was doing much better. So I drove at about 72mph which is the max the truck will go. Same thing driving back. About 11.7 miles round trip and the pack was above 300v 98% of the time. I burned over 4kwr and there was probably 1-2 kwh left. I’ll do a range test to see how much the capacity improvement really is.

On the way to work this morning the pack was over 315v the whole drive.

I will do the discharge maintenance again just to ring out any extra capacity that is held back by the memory effect. The manual says sometimes up to 3 complete discharges are required to restore the pack from deep memory effect.

The pack was down -5.2kwh. It charged to +6.3kwh. The truck took on 11.5kwh as of this morning. So it recharged with more than 100% overcharge. The pack voltage on the way to work this morning was significantly higher. It stayed in the 315-320v range while driving. This is about 15 volts higher than before the discharge maintenance. I won’t recharge at work to see how well the pack voltage holds up on the way home. At some point I’ll have to do a load test to see how many kwh’s the pack is good for now. It is possible that the deep cycling of the pack may need to happen once or twice more.

All of this work is important as it keeps the pack’s soc higher since the total pack capacity is kept higher. The cold weather brings the capacity down as well.

Starting yesterday afternoon I had to short circuit the pack after driving it until it would barely roll. The pack seemed to develop the classic memory issue again. The pack was only good for 4.5kwh during a drive that day. Not good. So this time I loaded each half of the pack to 1.5 amps with several light bulbs. After a few hours the pack was low enough to load it will all the light bulbs across the entire pack. This brought the load to 2.5 amps. 22 hours later I connected the new 240vac Variac from Ebay to it’s bridge rectifier and started charging the pack.

I had to keep turning the Variac’s voltage up to keep the current at a good charge level of between 6 and 10 amps. That 240vac Variac really sped things up. I didn’t have to do half the pack at a time as with a 120vac Variac. We’ll see if this gets the range back up.

The weather has been in the 40’s to 50’s and I suspect it will lower the range.

When using Dolcom for looking at ground fault info, the standard nominal reading is about +20 counts above the pack voltage.

In my experience when the ground fault reading taken with Dolcom is higher than nominal then the ground fault is between the negative most battery in the string and mid pack. When the ground fault reading is lower than nominal then the ground fault is between the positive most battery in the pack and mid pack.

So far with mild rain exposure the #1 truck had only had a ground fault reading swing about +30 to about -20 from nominal. It looks like the new gum rubber weather stripping is working.

Here is Lanny’s work on putting Hi Power Lithium modules into Ford Ranger EV’s .

He’s really moved the Rangers ahead and out of the hands
of a terrible fate. I’ve exchanged a lot of email with him as I’ve
worked on Rangers. I’m hoping he has a good BMS picked out or on the
design board.

I’ve had the Thundersky equivalents of those Hi Power cells that Lanny
is using. The Hi Power’s that he is using I have ordered and will
arrive next week for testing.

The issue for us is how will lithium function in a real EV. Bench
testing has major limitations compared to real world. Real world
driving beats on batteries like no bench testing ever could.

I’m also in a group that has a working Lithium BMS that is being
majorly tweaked at the moment. It has no street time yet, but will
soon. I was suppose to have some street time already but this second
truck has taken all of my energy.

The good news for us is that a small or medium pack will fit our
trucks fairly easily. My nicad pack is about 1400 lbs lighter than the
bone stock Hawker pack. The Lithiums would be 1800 lbs lighter than a
bone stock Hawker pack. But each of the 100 cells has to have a BMS
module. All of those 100 cell modules have to talk to a master to keep
the driver informed.

The good news about using around 100 cells is that our chargers can be
tweaked to handle lithium. I’ve done it. The max voltage output of our
chargers is 400vdc. 100 cells could only charge to 4.0v each. 3.6v to
3.8v has been suggested. The 400v limit will help keep from
overcharging the cells. But they still have a mandatory requirement
for a BMS.

108 40ah cells would fit easily. 100 60ah cells will fit nicely. About
100 90ah cells with a second batt box will fit. The range due to the
light weight would be great. The lifespan in an EV has only been
tested to 25k miles. Then they were accidentally discharged overnight
and ruined in the blink of an eye.

40ah pack = $6.9k. 60ah pack = $9.6k. 90ah pack = $14.4k. All of these
prices are without a BMS. Right now a BMS looks to be $1.5k+ if you
assemble it your self.

There may be cold weather issues as was reported by a lithium user
some time ago. So testing is everything because toasting these babies
is vastly more expensive than buying the truck in the first place.

You’d think this format of Lithium would fit in the Prizm’s nicely.
But I have never heard of anyone trying.

A fellow EV’er commented about using 40 foot long greenhouse soil heaters
to warm his battery packs. Liking his idea, here is what I wrote back.

The ACP cars use battery heaters. It only draws a few hundred
watts when it’s running. It takes a day or two to heat up a cold pack,
but holds it there nicely after it does. The lead packs I’ve worked on
in ACP cars put off heat for a whole day once they are in the 30-40C
range. When we pull a pack out to replace a weak battery you can feel
the heat in a big way.

The pack heaters are also one of two reasons the ACP vehicles get 30k
miles from a lead pack. Heating up a battery to 30-40C also increases
it’s capacity, making it less likely to reverse a cell. The second
reason for such great longevity is having a BMS.

The ACP battery heaters are powered from the pack itself I just
remembered. That’s why they don’t have to worry about what voltage
they are charging from. They do have a simple IGBT circuit that turns
on, off, or pulses the heaters to keep them at the same temp with a
thermal sensor.

A little math shows that the ACP heaters total about 450 watts. So on a
336v system that’s about 1.34 amps. So our packs would supply a little
less as our nominal voltage is 312v. So it’s not a huge drain when
it’s running.

Yesterday the truck failed an important test. It rained and it cause a ground fault. Even with all new seals on the lid and gasketed washers, it still ground faulted. So I got into the pack and checked for ground faults with the volt meter. The ground fault was floating around as it had on me earlier this year. So I tried the meter in low current mode set to milliamps. It showed the same location for a ground fault. Removing the cells showed some KOH streaking. Cleaning it up did not change where the ground fault was shown to be. So I went to the new process that I had figured out last Feb. I removed a cell interconnect at the lowest voltage reading relative to ground. I kept pulling one copper bar at a time until the voltage suddenly shot up. It ended up showing me that the ground fault was in the rear drivers corner of the pack. Removing the 8 cells in that area showed that there was water between the corner of the cell and the wall of the battery box. So I cleaned off the walls and cells with diluted vinegar and reassembled the pack. That took care of that ground fault!! This time I moved the mylar insulator into the corner to help isolate that corner cell from the box if it gets wet again. There is another smaller ground fault on the passenger side of the pack. I just vacuumed in that area and called it good for now.

The rear edge seal was a different material since I had run out. So I tore it off and installed the correct material. Since the two ground faults I found were right under that seal, this could help.

Also I adjusted the ground fault setting in the software to give it more tolerance since having 252 nicads or 504 terminals is almost 5 times the terminals that a regular lead acid pack would have. Between that and having flooded nicads, the pack is more prone to having ground faults. So I will see during the next rain how it all functions.

The video chip did finally arrive thanks to Peter getting involved. Here are 3 of the BMS modules running on top  of 3 Thundersky 40ah cells. Everything works fine. I just need to make a couple lengths of wire to connect the communication lines from one board to another, then install them and the master board into the truck.

The video chip did finally arrive thanks to Peter getting involved. Here are 3 of the BMS modules running on top of 3 Thundersky 40ah cells. Everything works fine. I just need to make a couple lengths of wire to connect the communication lines from one board to another, then install them and the master board into the truck.

50ah HiPower lithium cells from China are now on their way. I want to bench test them and install them into the truck just like with the Thundersky’s.

« Previous PageNext Page »