Climate control mod to turn off heater

[coulomb]

One of the niggles with the early Leafs is there is no separate button to turn off the heater. So it's a hassle to just get the fan circulating air, without using main battery energy to produce heat (and it's a resistive heater, not a heat pump, in the early models).

Weber kindly pointed me to this MyNissanLeaf post on how to modify the climate control electronics to turn off the heater whenever the air conditioning switch is off:

https://www.mynissanleaf.com/viewtopic.php?f=37&t=11412

I chose method 3. I've finally gotten it working, so I thought I'd post a few notes and pictures. Here is the schematic from the above post:



The way it works is simple. When the air conditioning switch is off, the LED pulls up the input, which turns on the MOSFET, which puts a 330 Ω resistor across the temperature Negative Temperature Coefficient sensor. So the control system thinks that the car is plenty hot, and claws back the heater power to zero in about 10 seconds.

With the headlights on, all the dash lights dim, including the air conditioning LED which is used to trigger the MOSFET. It dims the LED by pulsing it at about 200 Hz, 10% duty cycle (so the input to the mod is low 10% of the time, and high for 90%). The idea is to keep the capacitor voltage low with the diode when the input is low, and it charges far more slowly (with a 100 ms time constant) via the 1M resistor R2. 100 ms is so much longer than the 5 ms of the pulses that hopefully the MOSFET will stay off, even without R3 and R4.

To cut a long story short, I finally decided that these resistors are needed after all.

The original post suggests using a Schottky diode. I worried that the extra leakage (compared to a silicon diode) might cause problems with this circuit, since the diode leakage has to look negligible compared to the current through a 1 M resistor (of the order of 10 μA), so for the final iteration of the circuit, I used a silicon diode. When I measured the leakage of the Schottky diode with a multimeter, it did not register on my Fluke meter (so at least 400 MΩ).

Increasing C1 to 1 μF didn't help either (before R3 and R4 were added), although I was still using a Schottky diode at that point. There is a small chance that the leakage, despite my measurements to the contrary, was significant and was affecting the circuit.

I don't recommend the suggested value of 220 kΩ for R4, since the MOSFET will only be getting 0.22/2.22 = 10% of the auxiliary battery voltage, less perhaps two volts from the LED. So perhaps (13.8 - 2) * 10% = 1.2 V. I measured my MOSFET's threshold voltage at about 1.95 V. So it would never turn on properly. So I used another 1 MΩ resistor for R4.

Do view the video on the MyNissanLeaf post before taking apart the climate control unit. In Australia, we don't have an air bag warning light at the bottom (we have just a blanking plate), so that's one less connector to remove. That might mean that it's safe to turn the car on with the climate control unit unplugged, but I would still not risk it.

The hardest connect to get get undone is the first one, the hazard warning lights switch. Initially, it might seem that you can't remove the unit very far, making it very fiddly to get to the connector behind the hazard switch. But with a moderately gentle pull, you should be able to get it out this far:

Pull out for hazard sw.jpg (132.31 KiB) Viewed 529 times

This is the area you are aiming to press with a screwdriver (circled). You might do some good with the slot, but I didn't. This is the view from the windscreen; it's a pity that cameras are smaller than human heads I found it helpful to have one hand wiggling the connector from below, with another attempting to press with the screwdriver and wiggle from above at the same time:

Hazard switch connector.jpg (117.27 KiB) Viewed 529 times

Here is the hazard switch plug removed:

Hazard switch plug out.jpg (107.35 KiB) Viewed 529 times

I had a lot of trouble removing the two cable tie Christmas tree things:

Christmas tree cable tie.jpg (24.7 KiB) Viewed 529 times

The two tabs that hold it in seem to like to bend into crazy places, sometimes breaking off. I ended up siliconing mine into place after the final re-assembly, since they became so loose.

There is no need to remove the four large screws holding the climate control switches to the frame. Just remove the six smaller screws holding the black cover in place. With that removed, the printed circuit board just lifts out.

When making the mod, it's important to keep in mind that the printed circuit board rests on a number of pieces of plastic, so that there are a couple of "keep out" zones where you don't want any parts or wires at all, since the printed circuit board rests on the plastic in these zones:

Keep out zones.jpg (160.96 KiB) Viewed 529 times

Note how the original poster will have trouble when he goes to replace the board. I found it best to keep most of the parts above the large connector. Here is my final placement. It's rather messy since it was modified several times, and the silicone from the previous attempts was difficult to remove, and the final silicone was not yet in place:

Final placement.jpg (197.32 KiB) Viewed 529 times

Note how in my case, LED6 is populated. It's still quite easy to solder a wire to it, however. Some Australian Leafs may have the controls mirror imaged, in which case LED5 is populated, and LED6 is not.

[ Edit: replaced link to the schematic that was mysteriously lost. ]

[jonescg]

Awesome. I love an old school patch on new technology.

[weber]

Top marks for actually posting about it, Coulomb. But man! That's ugly. And I don't understand why you think you need the extra two resistors.

I've been sitting on the photo below for the 6 weeks since you watched me do it (for my wife's Leaf), and I gave you the parts to do yours. It works just fine with headlights on or off.

That's a 100 nF capacitor, a ZVN3306A MOSFET and a BAT85 Schottky diode. I remember I'd run out of BAT85's, so the Schottky I gave you was a 1N5711, which has slightly higher forward voltage, but much lower reverse leakage (0.41 V and 30 nA vs the BAT85's 0.32 V and 400 nA). With your MOSFET gate threshold of almost 2.0 V, it should have worked just fine. I note that the OP only had a 1.0 V gate threshold. The LED driver should be saturating at less than 0.5 V. The 100 nF capacitor voltage shouldn't rise more than 0.6 V in 5 ms, and a 1 µF would have reduced that to 0.06 V. So it should have worked without the extra two resistors, even with a 1N914/1N4148 diode (0.65 V and 20 nA).

https://assets.nexperia.com/documents/d ... /BAT85.pdf
http://www.learnabout-electronics.org/D ... 1N5711.pdf
https://www.vishay.com/docs/85622/1n914.pdf

LeafClimateMod.jpg (218.32 KiB) Viewed 506 times

I note that the MOSFET source, and the lower capacitor lead, are soldered to a patch where I ground the solder mask off the fat ground track. I believe that was your (excellent) suggestion, Coulomb.

[antiscab]

out of curoisity, does the car throw an error code if you physically disconnect the heater element?

[coulomb]

out of curoisity, does the car throw an error code if you physically disconnect the heater element?

It would seem so. From the Method 1 post:

• "Caution: You should avoid leaving the connection open or shorted out while the car is turned on. This will cause NO damage but it will log an error (DTC codes B277C or B277D). "

[ Edit: OOPS! That likely refers to the sensor connection, not the heater element connection. Sigh. ]

That's why the cheapest mod costs 5¢ instead of being totally free; you need a resistor in place of the temperature sensor.

Edit: I'd guess that the heater element might be hard to get to, or the author just didn't want to deal with high voltage cables or connectors and the associated safety issues.