Car Heat · 31 January 06

I’ve pretty much wrapped up Eve’s heater work, at least the installation into the air duct part. Thankfully it wasn’t anything at all like the duct-work in the movie Brazil.

There’s still a little bit of electrical wiring to do under the hood: wire up a couple of high voltage relays, fuses, and find an enclosure to keep it all clean and dry.

Kind of has a bug-eyed look to it, eh?

Here’s what the installation looks like:

This is the view from the car’s fan section. The opening is about five inches vertically, as is the widest point horizontally. As luck would have it this is about the same size as the ceramic heater core when it is mounted on its insulated frame.

You can’t see it here, but there are two ceramic heaters, mounted back to back (well, their frames are, so about 3” apart in center). I drilled the frame mounting holes out a bit to fit some long bolts that were already laying around the shop. Almost a mistake, since the framing material is some sort of heat resistant plastic that is quite brittle.

This is the view from the other side, the side which is connected to the conjurer’s box of magic flaps and levers for redirecting airflow to the rest of the car.

I’ve change the orientation of the photo just so you don’t get overly comfortable in your spatial orientation. ”:^)

The nozzle looking thing is a drain port. Since this is the old air conditioning box it had to have a way for the moisture that gets frozen and re-thawed on the cold AC coils to escape. This would be hooked to a rubber hose that goes through the firewall towards the bottom of the car. No longer needed.

So that’s about it. But before I close out this segment on heating I wanted to share results from a little weekend experimentation.

Another of the popular “you have an EV?” follow-up questions is, “What about heat in the Winter?”

If I’m feeling bold I’ll explain all about ceramic cores, inverse resistance, on-demand instant heat, slight reduction in range, blah, blah, blah… Most of the time I simply reply that it uses an electric heater that does a pretty good job.

Which is a fine response for most folks. When faced with a curious tinkerer/engineer these kind of generalities never seem to satisfy and I’m left flailing for specifics.

So I borrowed a digital temperature probe from the office, a decent multi-meter (mine’s kind of cheap), and set out to answer the question I’m sure has been keeping all of you awake at night:

Hot or Not?

No, wait, that’s not the question, and definitely the wrong website. The real question is how well does an electric heater work?

At first I just set up the fan and heater box with the temperature probe hanging out the opening, but that’s a big opening and probably not at all representative of what we’ll read at, say, the defroster output. So I added the flap section, with flaps set on defrost mode and the temp probe hanging in one of the two output ports.

Still not perfect, but re-assembling the whole dash on the garage floor was too much, even for science.

The fan was hooked to a twelve volt battery jumpered to one of the four possible speed settings: low, medium, high, BLAST. The chart above shows the results of using BLAST mode with one heater core and then two heater cores.

Actually the first couple of times I tried measuring the 2 core setup I got the same results as a single core. Turns out the extension cord was bad. Phew! Still, I was expecting blazing heat, what’s up with that?

On the opposite end of the spectrum here’s what you get with the fan set on LOW. Low is pretty low speed, but it gives us a range to work with.

A ceramic heater heats up when electricity is applied. In addition to the heat rising the heater’s resistance also goes up. This makes it somewhat self-limiting, i.e. it won’t go into melt-down mode. If I had a variable speed DC fan I could have probably figured out what the maximum heat is, and even the best fan speed to optimize heat and air circulation.

BTW, I have only one data point in regard to heat rising capability. It was 46°F/7.7°C outside and that’s where all of the tests started from. How hot would they go if the temperature was twenty? Zero? If we get some colder weather I may set up the test again to find out.

As a final test I decided to face off the electric heaters against our VW Golf’s heater.

First a little back-story on traditional car heat. When a gas engine combusts fuel about 20-25% of the fuel is turned into energy to move the car (and how efficient it is at using that depends on tire pressure, aerodynamics, weight, etc..). The other 75-80% of the energy is lost as heat: out the exhaust, the surrounding air, and into the engine core.

Think about that for a second. When we pay $2.00 a gallon for gasoline what we are really doing is paying $1.50 a gallon to heat the world. The other fifty cents worth of fuel goes towards getting us to work.

The good news is that we get “free” heat to warm our car in the winter…and in the Summer!

I should be clear that there’s no “free” heat with an EV either. On my last EV the heater pulled from 8 to 12amps (from the 144vdc pack) depending on fan speed. Compare that to the 75amps it took to drive the EV at 40mph on a flat road. The good news is that, unlike a gas engine, you can turn it off and no more energy is wasted making heat.

I only had one shot at testing this since the engine core would stay warm for a long time afterwards. I taped the temperature probe inside the car’s defrost duct, opened all of the windows (no circulation or re-heating), turned the fan up to High (not Blast) since that’s what I normally use, and turned on the car and waited.

It was pretty boring.

I’ll admit that I probably swayed the tests a little more favorably (if that is possible) towards the car. Instead of just letting it sit idling I revved up the engine a bit. Still, it was slow, with over three minutes before any kind of useable heat.

Even at the end, when it was reading 140°F/60°C, the heat gauge on the dash was showing a bit more than 180°F/82°C. I don’t know if there’s temperature limiting for the heating core or if this just showcased how inaccurate dashboard gauges can be.

All in all it was a fun experiment and I’d like to thank Faith for being my data collection assistant while I tended wires. It was a poor-man’s (woman’s?) data collector, I wrote a quickie PERL program that ran on my Powerbook. She’d read a number off the meter , typed it in, and the program time-stamped and sent it out as a comma separated value (CSV). The CSV data imports into spreadsheet programs nicely.

For those curious few, here’s the program:

    #!/usr/bin/perl
    while(<>) {
        chomp;
        my ($sec,$min,$hour,$dummy) = localtime time;
        print "$hour:$min:$sec,$_\n";
    }

I put it in a file called grab.pl and invoked it from an OS X (or Linux/Solaris) terminal window with:

perl grab.pl > output.csv

In retrospect I should have just saved the elapsed time from the first data point to all subsequent data entries, since that’s what I ended up re-calculating to make the chart.