The Struggle Continues: Successes and Failures in Winter's Final Stretch

We're having a time in the short days of winter. Here at the end of January, there is a bit of light at the end of the tunnel as the winter at this point WILL end at some point. But we struggle.

This week, we look at the final touches on the Electric Swallow. This is a neat little car with a neat little body and we had fun doing it. I was a little annoyed over the throttle issues but we more or less solved it with a 12v to 5 v converter even though it wasn't as it should be by the book.

We did do something we are going to try to incorporate in all future builds, including the Elescalade. That is interlock our controllers using the little relay on David Kerzel's J1772 board that we use. This little board is $37 and it makes your car respond to proximity switch and copilot signals from J1772 EVSE. But it features a little relay. On this build, we routed a 12v interlock signal through the normally closed contacts of this relay. In this way, if you plug in the charging plug, the car is disabled and you can't drive it away.

We actually routed the same signal through the Xantrex so it could also disable the controller when the SOC got down to 10%.

The car drove well and Lee took off for Denton with no real issues. If you stepped on the throttle all the way the controller kicked out. I showed him where to put a little resistor on the 5v DC-DC to fix that. And he seemed happy when he left Saturday afternoon. The car drives very nicely and as predicted, by using the 100Ah cells and the AC-50, it is very spritely.

We were hoping for additional reports when he got home on actual range and acceleration as we had no time for them here before he needed to be back to work.

Unfortunately, he apparently towed the car all the way to Denton with the VW tranny in gear. On arrival, he found transmission fluid all over the passenger compartment and nothing working. We're hopeful he can get all that turned around but it's a setback after all this work.

Meanwhile, I blew experimental A123 module La Troiseme number two WIDE OPEN.

I've already heard from a number of viewers DEMANDING to know what happened. Well, I'm not sure and I'm not sure we'll ever know. We don't have a "script" like House. Whatever happens happens and I don't always know what caused it.

Before casting, the cells measured 77.75 Ah which I thought was quite good. We cast it in resin and let it cure a few days and discharged it - but we really only got 65 Ah out of it and that was down to 24 volts or 2 volts per cell. We probably over discharged a couple of cells at that point. Then we put the charger on it. After 80 Ah it never did reach above 41 volts but it DID split the resin wide open and the most positive set of cells were hugely swollen and leaking electrolyte. Later, another set split it in another place. So at least two cell segments failed.

No fire. No smoke. But some heat. And the module did split wide open.

Module one failed in an interesting way as a result of something I just did wrong. I've been playing with these new light object voltmeter controllers. They have a voltmeter that sports two relays with two voltage set points each. You can use these for all sorts of things.

I was using it to switch the first prototype Troiseme to a load. Of course the load is very low resistance. The cables to the load are also very low resistance. But we were sampling the voltage at the wrong place. So when the controller closed the contactor, the voltage dropped to under 2 volts because of the drop across the cables. The ratio of cable resistance to load resistance is very low. When it goes to 2 volts, the controller opens the contactor. With the load removed, the voltage shoots up above 3 volts. This causes the controller to close the contactor.

And so what I set up was a series of perhaps 50 cycles from zero to several hundred amperes all occurring in 30 seconds until I could figure out what was going on and get it stopped. This caused a lightning display inside the battery module. It now shows full voltage, but any attempt to draw the least amount of current results in a total collapse to zero volts - maximum internal resistance possible.

With Swallow finished, we are working more on the Elescalade and have good progress there really. Rod fabricated a couple of 3/8 inch polycarbonate covers for the massive 76 kWh battery pack and box in the back of the Cadillac. We've mounted an Manzanita PFC-75 to this cover and a Vicor Megapack power supply we will use with six 5v 40A modules and one 15v 10 Amp module for a 15 volt output at 90 amps for our 12volt system.

We also moved the Aurora Inverter down to the shop. This allows us to hook up the 192v pack to this inverter and drain about 36 amps from the pack to produce 21 or so amps of 240vac right into our panel. In this way, we can use the truck to power the shop. Because of anti-islanding circuitry in the Aurora, this will do us NO good as far as backup power goes because it won't make any if we lose grid power. Why they have done this mindless thing makes no sense to me. It does not achieve the safety issue they sought.

But in any event, it DOES let us drain down a pack and instead of wasting the energy as heat, we can use it to run the bandsaw and the lights. That's pretty cool.

After taking the pack down to 170 volts, we'll trim from there manually using a little setup from the batt lab.

I basically mounted a 12v supply, a 0.1 ohm 250 watt resistor, a contactor, and this voltmeter controller all on a little piece of the aluminum aircraft decking we've been using. We have one set of cables to carry the 30 amps or so of drain current and a SECOND set of voltage sense wires with alligator clips we will connect DIRECTLY to the batt terminals to prevent the oscillation that caused so much trouble with the A123 module. In this way, we sense the actual battery voltage instead of the voltage after the current drop. Again, this is caused by the very low 0.1 ohm load. It's actually about the same amount of resistance as we see in the connections.

We'll set the lower limit at 2.60v and the upper limit at 2.77v. In this way, the cell will discharge into the resistor until it reaches 2.60v. At that point, the controller will disengage the contactor. The cell voltage will creep back up slowly. When it reaches 2.77v, the controller will again engage the contactor and take the cell down to 2.6v again. This cycle will be repeated until the cell just can't make it to 2.77v. That should be about 2.75volts near enough.

In this way, I can connect this device to each of the 57 cells in turn and basically walk away. An hour later, that cell should be at 2.75v. At some point, they all will be. Then we'll charge the pack and set our Manzanita for the top voltage.

The J1772 receptacle has some pretty stiff wiring but the 75 Amp Manzanita will certainly heat that up. Fortunately, it is just a foot or so long. We've terminated that in some terminals under the truck right where the J1772 comes in at the gas cap. We'll run 4 AWG cable from there to the Manzanita to handle the 75 amps this charge really can draw if you put it close enough to a panel. And of course, we have 4 AWG cables from the Manzanit to the cell terminals, although with the charger mounted on TOP of the battery box, those cables are quite short.

Again, we're going to use David Kerzel's little AC31 board to do the magic and make use of his relay to do a couple of things. Interlock the controller again of course. We'll also probably use it to turn on the heater pump through a seasonal switch. In this way, in winter with the seasonal switch set, the charger will cause the pump to come on circulating our glycol in the system. We're going to put two little 250 watt rubber heater pads on our system that run off of 240vac. When the J1772 connector is in, this will use wall AC to gently heat the system - all night long in most cases. So the pump will run and the very low level of heating should keep both cabin and batteries reasonably above freezing and I suspect, running all night, really quite warm which is why we are using such weenie heaters.

During the day while driving, our much larger 14 kw heater running off pack voltage can be used as desired. But I don't want THAT much heater running all night. A thermostat failure and we could really heat things up and cause a fire.

So we'll have an alternate system powered by AC when charging.

Today's video is of a bit more reasonable length. Our 3 hour mini series are just too much editing for me, and two much video for you.

Warmest Regards;

jack Rickard