iVac Switch Box teardown and modification

[morrone]

I have an iVac Switch Box, and I would like to increase the current threshold that it uses to turn
on the second device (usually a shop vac).

Attached are photos and my hand-drawn reverse engineered schematic.

It looks like in "auto" mode, a current transformer senses the current going to the tool.  The circuit
uses half-wave rectification, and then clamps the maximum voltage to 10V with a zener (1N4740A).  There
is a capcitor for smoothing and, I think, for a small delay after current is sensed and then a bit
longer delay when current goes away.  The voltage feeds the gate of a power mosfet, an STP3NK60ZFP
with typical Vgs(th) of 3.75V, and a range of 3V-4.5V.

Are there any suggestions for easy ways to adjust the device's turn on current threshold higher?   Just
changing resistor values to scale down the voltage at the mosfet gate would seem to also significantly
throw off the turn-on and turn-off times, which is not desireable.

[morrone]

Actually, what if I just replace R3 with a 1Mohm potentiometer, and connect the gate of the mosfet to the wiper rather than its current node in the schematic?

[thm_w]

As you say there is a short delay for turn on as 10uF is charged through 43k, then a longer delay for turn off as 10uF is discharged through 1M.

I think it depends on the ratio of the current transformer if your method would work or not. If its already hitting the 10V zener cap now with a small current, tapping off the 1M would just extend the turn on and turn off times.
If the voltage is lower, lets say 5V DC out of T1 (after diode) with your small current, then your 1M divider would work. So it might be worth measuring that voltage with whatever loads you are optimizing it for.

Another option would be loading down T1 output with a suitably sized resistor, but a pot may not be able to handle the power dissipation.

[opinionated_hack]

As you said, R2 & C1 set the turn-on time while R3 & C1 set the turn-off time.  However, the best way to change the threshold would be to adjust R1 which controls how fast the zener will hit 10V for a given tool current.  A larger value for R1 will allow a smaller current to charge the gate.

[morrone]

I measured the MOSFET gate voltage with various loads up to about 90W.  See the attached chart.

This particular MOSFET seems to turn on an trigger the relay at around 4.2V, and a tool current of 290mA (which pretty much matches the iVac Switch Box advertised threshold of 0.3A).

I have abandoned the idea of an adjustable threshold, since there seems to be no place in the circuit that it is actually isolated from mains.  I'll settle for a fixed higher threshold.

I tried splitting the R3 into two resistors in series, and connected the mosfet's gate in the middle.  That worked, but of course I failed to measure the tool's actual current.  I set the threshold to 50W, since the tool is 35W...but the tool's power factor is bad, so VA is around 55.  Sigh.

This approach probably reduces the turn-off delay unfortunately.

@thm_w, how would you suggest loading down the transformer.  Just a resistor across the secondary terminals?

@opinionated_hack, the turn-on voltage hits well below the 10V zener kicks in.  I think that R1 is just there to limit the current through the zener when it allows current through.  Otherwise increasing the value of R1 has little difference on the gate voltage than increasing R2.  Both would result in a much slower turn-on time.  Or am I missing something?

[thm_w]

I tried splitting the R3 into two resistors in series, and connected the mosfet's gate in the middle.  That worked, but of course I failed to measure the tool's actual current.  I set the threshold to 50W, since the tool is 35W...but the tool's power factor is bad, so VA is around 55.  Sigh.

This approach probably reduces the turn-off delay unfortunately.

@thm_w, how would you suggest loading down the transformer.  Just a resistor across the secondary terminals?

Nice chart.
Yes placing a resistor directly across the secondary.

But based on your measurements, splitting R3 should work well. Just a bit painful to set the threshold and also manage turn off (by increasing total resistance of R3 >1M). You can still use a small trimpot inside the unit to make it a bit easier. Maybe two pots in series for adjusting turn off time as well (if its important)?