Saturday, February 21, 2009

Despite the recent snowfall and subfreezing temperatures, I have managed to stay relatively productive. The most recent developments (as of last night) are a little more progress on my motor controller design. So far, I've completed the isolation circuit for the IGBT driver. I'm using an optocoupler between my PIC's output and then input of my driver IC, and a 5v to 15v isolated DC/DC converter to power it. I'm using a resistor to return the driver to a logical low after the PWM pulse from the optocoupler. All things considered, I'm pretty proud of the circuit. It allows for full isolation (a few kv, need to review the datasheets to be sure) between the high voltage and low voltage systems, while still requiring only one low voltage input and without tapping the high voltage to power the driver circuitry. This keeps the parts count minimal (6 ICs including voltage regulators and stuff) and the price down. The most expensive part on the driver board cost me $8.

Right now I'm trying to come up with an internal precharge circuit. I want to avoid a high voltage tap across the contactor if at all possible. Right now what I have in mind is a circuit that disconnects the caps from the high voltage while the contactor closes, and then reconnects them through a MOSFET or IGBT once it's juiced. That way, the solid-state part of the circuit would take the brunt of the charge current and minimize arcing. Still working out the details.

After that's all taken care of, I'll be working on designing the PCB. I'm hoping to have only one board for the whole controller. This should be possible if I use the same board (with adequate gaps for isolation, of course) for the high voltage cap circuit as well as the low voltage driver circuit. If I do this right, I'm pretty sure I'll only need one internal connection outside of the PCB- the one that goes to the gate of the IGBT. This means less opportunity for failure, more vibration resistance, etc.

Anyway, I'll be doing my best to get this thing ready to test while I'm home over spring break in about 3 weeks, so I'll be staying busy. Stay tuned!

Tuesday, February 10, 2009

So, above we have a picture of my home built controller project in its current state, along with a couple other staples of college life that are continuously inhabiting my desk. Might not look like much right now, but it's actually quite a step forward. I've got it up to the point where it's able to power my 600A IGBT. Granted, it's currently running a motor I pulled out of a VCR that was in the trash behind one of the frats, but it should, in theory, be able to take quite a bit of juice. All that's really left are a couple little programming tweaks, some, uhh, minor wiring upgrades, etc. and it'll be ready to drop into a car. I've got a monster heat sink taking up the other side of the desk, and a few 2200 uF caps are on the way as I type. I've also got a few isolation issues I have to take care of. The logic input to the gate driver will be isolated with an optocoupler, and I'm currently searching for an isolated DC/DC converter to provide power to the "dark side."

Oddly enough, this is actually really exciting for me. I've been working on and off on a home made controller for a good long time now, probably a little over a year, and I finally have enough knowledge that I can probably make one work reliably. Since I've used a surplus IGBT and mostly inexpensive electronic components, the cost for this thing should come in well under $200- quite a bit better than the $1800 or so you might find yourself spending on a Curtis. Of course, that doesn't include the two programmers I've gone through in the process of learning to program, or the mini laptop I ended up buying because it was the cheapest thing I could find with Windows XP, or the several dozen chips I've fried for various inane reasons, but it's still worth it.

Anyway- stay tuned, more developments soon!

Monday, January 19, 2009

Howdy folks! Those of you that have been paying attention despite my constant attempts at boring you to death may have noticed that I've been 17 for roughly a year. This train of thought would eventually lead to the conclusion that I'd have to turn 18 sometime soon. Rather than keep you all wondering, I might as well just admit it: I'm finally 18!

In addition to me just wanting to brag about having survived another trip around the sun, I might as well bring everybody up to date. College has, oddly enough, been time-consuming! Nevertheless, I've managed to get a little work done on EV-related projects every now and then. I've been doing quite a bit of electronics work, and have succeeded in getting a few gadgets up and running with varying degrees of success. I've built a few motor controllers of various types on the small scale, digital voltage gauges, and a few other non-ev-related items that I am equally unrightfully proud of- digital clocks, stepper motor controllers, that kind of thing.

Anyway, right now my main focus is on the motor controller side of things. Mainly, I'm working on scaling up a motor controller that can handle 6A to something that can handle 600A. I'm probably going to be building either a single- or double- IGBT controller, to save space and boost efficiency. The main goals of the project are to keep size, weight, and of course, cost to a minimum while maintaining capabilities suitable for road-going vehicles.

One of the coolest things about a project like this is that it allows you to do something about everything you didn't like about the production stuff you used before. For instance, the throttle sensors used by most modern controllers really suck. Even my Raptor, which is supposed to have auto-sensing capabilities for throttle positions, requires a lot of tinkering to get it working to its full potential. My controller will have a programming mode that requires some setup beforehand, but can be calibrated to just about any throttle setup.

Well, that's about it for now. As always, I'm going to try to keep everybody a little more up to date, but as we've seen in the past that doesn't always work out as planned. Nevertheless, some pretty big developments might be coming out of the GreenflightEV lab (aka "the dorm room") in the next couple months, so stay tuned!