Probably MOSFETs on the other side of the board.
I took some measurements by powering my Inovelli modules through a Kill-A-Watt. Could use some help figuring out the relationship between the cap value and the motor power reduction. I naively tried to calculate the motor’s impedance using Ohm’s Law and see if the cap’s impedance scaled the power linearly, but the numbers didn’t make sense and I’ll probably need to learn more about AC circuits to know where it went wrong.
My line voltage was measured as 122.3V.
My first test involved connecting a Smart Fan wall switch (VZM35SN) to the Panasonic AC bathroom fan mentioned above. The measurements were as follows:
H: 70 VA, 62 W, .572 A, PF .88
M: 34 VA, 16 W, .278 A, PF .47
L: 22 VA, 7 W, .179 A, PF .32
My second test was to use a Smart Fan canopy switch (VZM36) to control a Casablanca ceiling fan:
H: 89 VA, 89 W, .727 A, PF 1.0
M: 43 VA, 21 W, .341 A, PF .49
L: 26 VA, 7 W, .212 A, PF .29
My final test involved setting the Smart Fan canopy switch to the high speed setting, and then using the pull chain on the Casablanca fan to lower the speed:
M: 55 VA, 32 W, .449 A, PF .59
L: 27 VA, 7 W, .220 A, PF .28
I did not take apart the Casablanca fan, so I was not able to confirm the capacitor values. I see information online that suggests that these boards have 4.5uF, 6uF, and 5uF capacitors. One of these is to control the start winding rather than the fan speed.
I also did not take apart the VZM36 (it’s in a junction box up in my attic), so it may or may not be using the same 4.7uF and 8.7uF caps as VZM35SN. Can @EricM_Inovelli confirm?
Based on these results it looks like the Low speed is very similar on VZM36 and the Casablanca. I think that means that both are using ~4.7uF caps for Low.
Medium speed is significantly higher on the Casablanca pull chain than on the VZM36. Wondering if this means the pull chain is using 10-11uF rather than 8.7uF? Or maybe more
And of course M/L speeds on the Panasonic bathroom fan are just abysmal. This motor is obviously quite different from the ceiling fan motors. I don’t know how large of a cap I’ll need to get reasonable performance out of this unit.
I bought 10uF, 15uF, and 20uF MKP capacitors from ebay with the intention of testing whether they improve the anemic medium and low fan speeds described above. These turned out to be way too large for the enclosure:
@EricM_Inovelli is this why your design maxed out at 8.7uF? Was that the highest capacitance that would fit in the box?
The 10uF won’t even fit? Its hard to tell, it looks way too big then in another picture it looks like it might.
Is the medium speed OK? If so, a firmware change that turned on both capacitors could give another slightly higher speed.
10uF is too wide (with the side leads) to fit in the enclosure and much thicker than the 8.7uF. Also, I was really hoping to use 15uF-20uF for Medium speed, and 8.7uF-10uF for Low. Medium needs a huge boost in my setup because it’s currently only 25% of full speed. Low is useless; the airflow is imperceptible.
One other difference, I think, is that the originals were rated for 125V and the replacements are 250V. Possibly, 125V replacements would be physically smaller.
(Is is safe to use a 125V cap here? AIUI, one of the contributors to the notorious Insteon reliability problem was using capacitors with a voltage rating that was too low to handle occasional spikes/surges: Repairing a failed Insteon PowerLinc Power Line Modem (PLM) 2413U — 1 Project A Week )