I’ll be really happy when I get done with this heater, so I can put the dash back together and get on with the rest of the project. The phrase “dollar waiting on a dime” comes to mind. I did make some progress this weekend, and concluded most of the “fiddly” part of the process. The rest of the mounting procedure should be relatively straightforward from here.

One significant thing changed about the design, based on an idea that popped into my head several days ago. Instead of designing a replacement heater core of the same dimensions as the old one, I decided to use the entire available internal volume of the heat/vent module, creating a wall across it and making the heater elements into a “window” in that wall. This way, I can place them beside each other instead of having to stack them (which would cause the leeward element to run hotter and deliver less power). Also the large available area means I can fit all three elements, which gives me quite a bit of flexibility in how I wire them.

Since there are four rows of ceramic slabs per element, with the three elements I have twelve in all. And twelve is a very handy number, being evenly divisible by 1, 2, 3 and 4. It also means there are thirteen electrical connections around and separating each row of slabs; an odd number. So by alternating positive and negative and optionally skipping some connections, I can wire the entire set for twelve individual slabs in parallel, six parallel sets of two in series, or four parallel sets of three in series. (I could go on, but I don’t have enough battery voltage, e.g. to effectively power four in series). And I can do any of these arrangements while leaving all three elements physically attached to each other and electrically connected like one continuous element. The manufacturers of the space heaters these came from actually use a similar technique on the cheap models to provide multiple power settings without the parts cost of a continuous (smooth) control circuit.

So I decided to go with the plan to build the heater core as a flat board. One result of this was that I had not only the opportunity, but actually the necessity, of removing the flaps that previously occupied much of that space. There were two flaps, a hot/cold mixture flap controlled by a servo, and an insulating flap controlled by a vacuum actuator that opened whenever the mixture was at any setting but full-cold. Since the new heater core will be electric (and therefore can be switched off unlike the original core) these flaps become unnecessary since air can flow though the heater elements at all times. And building the new core where I wanted it meant that they had to go. So I removed the flaps, and then I cut out their supports so they won’t obstruct the flow of air. A hole was left by one of the supports, but that was unavoidable, and easy to close off later.

After these modifications were complete, I started modeling a mockup of the board that would form the “wall” in which the elements would be mounted. I created it out of foamcore, since it’s rigid, easy to work and similar in thickness to the material I’ll end up using. Armed with a caliper, contour gauge, a razor knife and an overcast afternoon, I basically hacked together a piece of the correct shape to fit each half of the module. Then these two pieces got traced together to make a single piece to fill the entire area. It tested out OK, but I noticed a potential problem.

Because I changed the angle of the wall from my original idea to make the contour simpler to design, I also changed its dimensions and ended up with one of the elements being very close to the foam-covered dash/floor vent selector flap with no extra room to move it. About a half-inch of gap, which I’m hoping will be adequate to prevent damaging the foam on the flap when the element is at maximum temperature. I’ll have to test this with a suitable heat source before I cut the design out for real.

I also started considering the control circuit. Perhaps there’s something I’m not thinking of, but I had a thought that there might be a good reason not to put everything back together when I’m done with the heater core construction. I do intend to design a control system for the heater, which will ensure that the element doesn’t get too much voltage or reach too high a temperature, and will also provide a continuously variable output from hot to cold (instead of simply switching the element between “off” and “hell”). While I’d planned on bringing out all the wires so I could put that project off until later, it occurred to me that a perfect place for such a circuit would be in the air path before the heater element. It’s an area that’s guaranteed to be inaccessible by humans or animals, and in the air stream it has a perfect means to cool itself. The result is a 100% efficient heater control, since all losses in the control circuit would be in the form of heat (mostly from the power semiconductors), effectively pre-heating the air before it reaches the heater elements. All power gets used. The only way to improve on that would be with a heat pump, maybe using peltier junctions, but that’s a little more elaborate than I’m willing to get.

Helping out for the day: Vivek Gani, Bryan Barnett, Erik Bigelow.