DIY seat heaters

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Actually, I'm more concerned with finding 2N2222 transistors and a +5V three-terminal regulator. 7476s (or at least 74LS76s) seem to still be available, but I haven't looked for the transistors yet, and I obviously need something to interface the JK outputs to the power relays. And I also need to verify that a 7476 is guaranteed to power up in a "reset" state, or failing that, find a simple way to guarantee it myself.

And what I REALLY don't relish is the idea of dealing with CMOS and its legendary sensitivity to the slightest static charge.
 
Looking at the Dorman 628-040 kit on Amazon. With the weather starting to cool down, and my back giving me some grief this week, I'm thinking very seriously of going for it. Seriously enough that on my lunch break today, I took the trouble to figure out how the seat back zippers work.

I'm definitely going to have to reverse-engineer the power switch, because (1) the switch that comes with the heaters is too damn big to cut into a 68492-5SA0A switch panel blank (like my "flap ajar" indicator), and (2) I want, at least in the final installation, to have the seat heater power unlatch itself when I power down the vehicle.

Looking at the instructions (on the Dorman site), I note that it says, in so many words, "Do not place seat heater pad over any air bags or occupant sensors." Does anybody know, offhand, where the sensors are located on a Nissan Leaf seat? (Just the driver's seat.)

And while I can certainly find the fuses for heated seats, and use a voltmeter to determine whether they're on the "battery," "accessory," or "run" bus, there is an awful lot of plastic between me and the back of the fuse box. Does anybody know of an easy way to find the wire coming off of any given fuse?
 
Well, after over two hours of struggling with the upholstery (the valleys are held down with plastic rods sewn to the outer skin, linked to steel rods embedded in the foam by wire rings that aren't quite hog rings, and aren't quite split-rings, but are an unhappy combination of their worst features), and five minutes of jury-rigging the kit harness to a lighter plug, I now have a heated driver's seat.

Before doing anything else, I disconnected the 12V battery, and let it sit, while I checked my email, did some record-keeping, and had lunch. About 40 minutes later, I started unbolting the seat. I broke the bolts loose with a torque wrench, so that I'd have a number to shoot for when reinstalling: 30 pound-feet.

I really don't think they could have made that seat any harder to work on if they'd tried. There are three valleys in the bottom, each with four of those miserable wire rings holding the plastic rod down, and barely -- BARELY -- enough space for one of the heating elements to go in and lie reasonably flat. I had to open up ten of those rings (four in the front, and the front three on either side), in order to peel back enough of the fabric to get the heating element in, and then hook them back over the rods and close them back up after it was in place. And there wasn't any convenient place to route the cable: I had to poke a hole in the foam, Gorilla-tape the plug onto the female end of a socket extension, and pull hard to get it through to the underside.

The back only had two valleys, and it was only necessary to open up one of them, the inboard side. Which is really good, because on the outboard side, the side-curtain airbag is attached to the outer skin. (I will note that there's also a tag attached near the breakaway stitching: do not re-sew this cover.)

On the bright side, the valleys should keep the heating elements from doing any migrating.

Once I had everything back together, I hooked up the wiring and bolted the seat back in, torquing the bolts back to 30, and reconnected the 12V battery. After nearly 3 hours of no 12V, all I had to do was reset my radio presets (all two stations: KNX for all six AM presets, and KUSC for all twelve FM presets) turn the charge timer back on, and reset the default dashboard display to the big state-of-charge indicator.

Then, with the lighter plug that I know I have lying around somewhere completely buried, I went over to O'Reilly for another one. It took me all of five minutes to jury-rig it to the kit harness, and verify that the heaters work as well in the car as they did when I bench-tested them. This will do, until I've built my own control mechanism (and yes, I now have a small stock of 74LS76 dual JK flip-flops and 2N2222 transistors, as well as some relays to handle the roughly 5A the heating elements draw.

And I did it without the side-curtain airbag blowing up in my face.
 
Well, by last weekend, I learned that my control mechanism that works perfectly in an online simulator fails miserably (even without connecting relays!) with a real 74LS76 on a solderless breadboard, and by this weekend, I knew a few reasons why:
First, I was ridiculously overloading the outputs, trying to drive LEDs in an active-high configuration directly out of the flip-flop; I needed to add transistors. Fortunately, I'd bought a whole bunch of 2N2222s.

And second, I learned that trying to get clean clock pulses directly out of a physical switch is a fool's errand. I found a schematic for a debouncer, and I have some Schmitt triggers on the way.

Today, with nothing further to do on the control mechanism, I opened up the dashboard, and (once and for all) doped out how to get into the back of the fusebox. I tapped the actual front seat heater fuse (that's an awfully fine wire to be on a 10A fuse!) for my power (verifying that I had the right wire as many different ways as I could think of, before clapping a "suitcase" tap on it, and then verifying it again once I'd driven the blade home on the suitcase), then put the fusebox back where it belongs.

I then tried (and failed) to dope out how to drive an LED to follow the panel light dimmer, but at least I tapped an unused switch plug (maybe for a heated steering wheel?) for a 12V line that will go on and off with the headlights, so as long as I set a comfortable lighting level, it should be fine.

And I already have a wire on the accessory bus, from when I tapped the lighter socket for my "flap ajar" indicator and a dedicated socket for my GPS navigator; with a non-latching control mechanism, I don't really need to go to the additional trouble of finding a 12V source that only powers up when the traction motor is online.

So now, I have everything I need (including knowing how much room I have behind the button-panel) to hook up my control mechanism, once it's built and tested.

And still, nothing has blown up in my face.
 
Sunday, I finally had time to assemble a debounce circuit for the control button, and now my control circuit cleanly and reliably cycles through "low power," "high power," and "off" states. Unfortunately, I don't yet have the pre-clear working reliably, with the result that it powers up in "low power." Fixing that could be as simple as adjusting a resistor value, or it might involve a couple of the leftover Schmitt Triggers (they come 6 to a chip) from the switch debouncer.

*****

Tuesday evening, I changed the pull-up resistors on both the preset and preclear inputs from 1k to 100k, and added a 390k resistor in parallel with the pre-clear capacitor, and now I have something that works reliably, at least spread across two solderless breadboards.
 
When I moved everything onto a single solderless breadboard, with the spacing tight enough that I'll be able to get everything onto a matching PC board, I got a new malfunction.

This morning, I tried tweaking resistor values, with no effect, then finally swapped in a spare 74LS76. Not sure if I fried it, or if one or more pins just weren't making good contact, but I set it aside for further study. Because I'm doing this in TTL, I bought 3 of each IC. If I'd had the guts to do it in CMOS, I'd have probably bought 10 of each, because they're so much easier to fry.
 
Even as my Leaf sits in a body shop in Gardena, I've decided to abandon the TTL J-K flip-flop, one-button approach (the parts-count was just inflating beyond reason), and go for something lower-tech: SCR latching. Because SCRs and other thyristors always power-up in an "off" state, and don't require a regulated 5V supply, or control button debouncing. The only downside is that instead of one button cycling between low, high, and off, this approach requires a pair of very small (but not cheap) SPDT momentary rocker switches, in order to fit the controls and the indicator light in a hollowed-out blank plug (68492-5SA0A) for the button panel.

And yes, I'm proceeding despite the possibility that somebody will try to total my Leaf, instead of fixing it. The operative word is "try."
 
Back
Top