Sunday, May 17, 2009

Installing a New Potbox

As with many EV projects, much time is spent waiting for parts. The replacement Soneil chargers haven't shown up yet. On a more positive note, Synkromotive figured out some of the glitches in their new beta controller, so one should be available this coming Friday. This new system will require a different potbox, contactor and other wiring.

I'm excited to install the controller, so I installed a new potbox to get ready.


Here is a view on the driver side in the engine compartment just above the engine mount. I'm going to use the two holes in the engine mount bracket to mount the potbox.


Here's the PB-5 potbox that I purchased from Bob Bath with an aluminum bracket bolted to it. I was fortunate enough to keep this around in my pile of parts because it fits very well.


Did you notice the shiny silver thing attached to the arm of the potbox above? I had difficulty trying to figure out how to attach the acclerator cable to the arm of the potbox. The guy at the hardware store suggested that I use one of these tiny cable clamps. The "U" bolts fit perfectly in the holes on the arm and I can clamp the cable down tightly.


Here's the potbox mounted in place. Notice how it's tilted slightly to the left and not in parallel with the E-meter DC-DC converter in the black box to the right. This is because the accelerator cable interferes with the top of the shock tower if the potbox is mounted inline with the car. This leftward tilt moves the cable just inside the shock tower so it hangs better.


Not only did I have to attach the acclerator cable to the potbox arm, but I needed to mount the housing on th end of the accelerator cable in something so the cable could pull against something. To accomplish this, I took a spare piece of 3/4" angle iron and drilled a 5/16" hole in one end, near the edge. I then cut out the hole with a hacksaw to get this keyhole-like shape. This will allow the cable housing to slide in at its narrow point but not pull out when I slide it forward so that the wider threads sit in this 5/16" hole. I also drilled a 3/16 hole in the other angle iron face to mount this tiny piece with an 8-32 by 1/2" bolt.


Here's a piece of flat bar that will hold the cable housing. The two holes on the left are the same distance apart as the two posts on the potbox. The hole on the right end is where I'll mount the tiny piece above so the cable has something to push against.


After sawing off the tiny piece from the end of the 3/4" angle iron, I mounted it to the flat bar with an 8-32 x 1/2" bolt and nylock nut to keep it from falling off. This will rotate a bit and allow the cable some play as the potbox arm swings back and forth.


Here's the above bar mounted to the potbox with 8-32 x 1/2 button-head bolts with lockwashers.

Here's the same picture above, but with the cable mounted in it. The trick is to get the bracket on the right in the correct place so that you can use the two adjusting nuts to properly tension the cable so it gives maximum throw and stops at the zero point on the potbox.

Stay tuned for Friday when, hopefully, the controller arrives.

Saturday, April 18, 2009

Researching a Paint Job

One of the downsides of a home-EV conversion is that most of them don't have air conditioning. The AC tends to suck precious power from the battery and adds extra weight to the car while taking up space that could be used for batteries.

Thus, it's often best to get a white (or light colored) donor vehicle to repel the sun's rays during the summer. The conversion vehicle I purchased is dark metallic blue, which get's really hot inside, even on a mild day in the sun.

So I called a bunch of different car painting places to determine the cost of painting the car white. Much to my dismay, the places that got good local reviews wanted at least $3500 to paint the car. The local Maaco paint shop (which got horrible reviews) wanted $550 for their bottom-of-the-line paint option.

In short, I guess I'll be rolling down the windows and using a reflective shield under the glass for the near future. Note to self: purchase a white conversion car next time...

Planning to Replace the EV Components

In a month or so, I'll be getting a beta-test motor controller from Synkromotive which will have a variety of programmable features as well as more power. Since I'm replacing the controller, I looked into replacing the other components with off-the-shelf items that might work better than the current Belktronix system I have.

Here's the plan:
- replace the controller with the programmable Synkromotive unit
- replace the optical potbox with the original Civic TPS (throttle position sensor)
- replace the charging system with a bunch of individual chargers (Joule or Soneil)
- replace the integrated DC-DC converter with an IOTA DLS-45 unit


Here's a picture of the IOTA DLS-45 DC-DC converter on the right that I purchased at Northern Arizona Wind & Sun for about $150.

On the left is a jj12060 Joule individual battery charger from ElectricRider, supposedly modelled after the popular Soneil chargers. I purchased 13 of them, 12 for the batteries and one extra in case something dies or I need to charge an external 12V battery. They were out of Soneil chargers and said that the Joule was a drop-in replacement with the same specification (right!).

As is typical, things didn't go as planned. I saved this weekend for removing the old charging system and installing the Joule chargers. As a safety check, I clipped a Joule charger to one battery to observe its behavior. When I attached the charger, I heard a click (like a relay) from inside the charger. Hmm, if the battery is always powering an internal relay, then there's a constant current drain (that's bad).

I checked the current draw with an ammeter and found that the Joule charger pulls 35ma out of the battery when it's just sitting there! The Soneil chargers have the feature of not pulling ANY current out of the battery when attached but not plugged into the AC source.

In short, the Joule chargers are NOT a drop-in replacement for the Soneils and I'm going to return them and get Soneils instead.

Argh, more delays. At least the weather outside is beautiful and I can go for a hike.

Have a great weekend everyone.

Taking a Vacation

It's been over a month since I've blogged about the Civic and people have started to ask where I've been. I'm still commuting (even with a blown charge detector) and the car has 1500 miles on it. Several events have happened with the Oregon Electric Vehicle Association (of which I'm the Chair) and I also went on vacation.

At the end of march, I went on an amazing backcountry adventure, skiing the Wapta Traverse with a 45 pound pack on with Randonee' skis.


Here I am near the start of the trip. The scenery was mind-blowing.


Here's the group all roped together just behind an enormous crevasse, heading up to our high-point on the trip. The picture just doesn't do this justice.

Now that I've recovered from the trip, I'll be getting back into trying to fix the charging system.

Stay tuned.

Thursday, March 12, 2009

Charge Detector Blown Again

A few days ago, something funny happened to the vehicle. I drove home and my LVP light came on earlier than I expected. When I got home, the front battery was rather lower than the others in the pack. I attributed this to being just behind the front grille where a lot of cold air is blowing in. to remedy the problem a bit, I slide a large sheet of foam-core board between the batteries and the radiator grille which gives a bit of insulation and forces the air go around the side.

After measuring my batteries for several days after that, they seemed much better, but still didn't seem to get fully charged when sitting at work. Again, I attributed that to the low temperature around here (mid 30's).

Today, I drove home, plugged in the charger and just checked to see if the charge detector was working. Sad to say, the FET has probably blown open again since the OVP lights were not on and lots of current was dumping through the large external 3-ohm charging resistor, making it really hot. This would definitely explain the slow charging.

After trying to fix this device several times, I have a few options:

- fix it again (and wait for it to blow again)
- send the charging system back to Belktronix to get upgraded to a Gen2 system without the charge controller for about $400
- dump the Belktronix system altogether and go with something else, like individual Soneil chargers (Rob Connelly did this and is happy with the results)

I'm not sure what I'm going to do. The existing charger works, but it's just a bit slow. I'll ponder this over the weekend. I have an OEVA meeting to lead this evening.

Ugh,
Tim

Saturday, March 7, 2009

1000 Miles and Still Going...

There's not too much to report on the Civic. I just passed 1000 miles yesterday and it still seems to work fine. I've definitely noticed over the past several weeks of commuting that my range is highly dependent on my speed. Even driving at 55mph instead of 60mph on the freeway gives me a few extra percent of battery life on the Link-10 E-meter.

I've ended up driving with the LVP protection disconnected from the Belktronix controller because it tends to make my speed go way down even though I have plenty of battery left. The LVP circuit is still attached to the battery light on the dash, so I know when to start being very careful of my driving habits.

The charging system seems to be working fine now that I have the BatMon boards inside boxes, protected from the elements. I've driven in a hard-driving rain several times without any problems. The final voltage on the batteries differs slightly from battery to battery due to the analog nature of the Batmon circuits. The BatMon boards rely on the calibration and tolerances of analog components to activate the shunt resistors at 14.5 volts or so.

I'm starting an experiment with several prototype digital battery monitoring boards from Synkromotive. I'm going to see if I can determine which battery is causing the LVP fault earliest with this monitoring system. For you circuit geeks out there, the Synkromotive system (prototype only at this phase) has a 12-bit A2D converter which makes it more accurate than the PakTrakr system used in many EVs today. Those extra bits can make a big difference when your batteries are close to the end of charge and the balancing system needs to compensate for small differences like 50mV. The digital nature of the system (each battery gets an 8051 microprocessor) allows for easy adjustment of calibration values in case things drift over time.

I'll report back when I get some more results.

Cheers,
Tim

EV Driven Link

A few weeks ago, I added an "EV Driven" link to the right menu bar of this blog. EV Driven is a group of folks trying to spread the word about EVs and are behind the upcoming movie Revenge of the Electric Car. It contains several searching features for this blog as well as many other blogs related to electric vehicles. Please check it out when you get the chance.

Cheers,
Tim