Switching big power and off in a friendly way

I’ve got a Baby-O using six high-side N-channel FET switch controllers to switch power on and off to up to six 12V resistive heaters. They can eat up to ~2 amps max on each channel depending on the particular heater connected.

These are resistive heating loads only, it doesn’t matter if they take a little time to fully power up. Should I dampen the onrush of current on these heater channels? With up to six loads switching on and off at 2A, is that a noisy electrical environment that would benefit from filtering? I’ve never ventured out of micropower before.


I know it’s not what you asked, but it might interest you to look at this item;


I use these for a few interesting things. Nice part about them is that they can be controlled over the network through web, API and can also have scripting onboard. Quite a package for the price, really.

Not sure what you are building, but the short description made me think of this item.



N-channel MOSFETs do not seem like a good choice for a high-side switch in your application. I don’t know if you’re using some fancy (and overkill) part or circuit to correctly switch the MOSFETs or if you’re using the MOSFETs incorrectly, but either way, 2A is not that much, so I think you should use P-channel MOSFETs. (By the way, easiest would be to use the N-channel MOSFETs as low-side switches, but you might have some good reasons for not going that route.)

It sounds like you are planning on turning the heaters on and off at very low frequencies (once in many seconds), in which case you won’t have the more typical noise you would have if you were switching these at some high frequency (e.g. using PWM at thousands of times per second to do variable heating). You can look at your power line with a scope to see if it drops a lot when you turn on all six; if so, you could add some capacitance to reduce the dip. If it’s an option, you should probably try to stagger changes to multiple channels. You might also mess with your MOSFET drive circuit a bit to limit how quickly it turns on (in the us to low ms timescale), but you should be careful not to turn it on so slowly that it gets destroyed.

In general, though, 2A is not that much, and the parts to do it don’t cost much, either, so I think you can just start building up the hardware and then look at what kind of fluctuations you see on your power lines.

- Jan

Great advice! Thanks.

I am using a Maxim high side N-channel mosfet controller which uses charge pumps I guess to generate the required voltages. I did work with N-channel low side at first but had trouble integrating a current sensing resistor. The high side switch lets me put a current sensor resistor right at the return ground and land it in the perfect range for my ADC.

Staggering the channel enable/disables should be easily doable. I’ll allot each of the six channels a time position within every 1 second when they are allowed to transistion on or off.

I was really happy with my Maxim N Channel high side controller, so hearing that it is fancy (and perhaps overkill) makes me wonder if I shouldn’t just ditch it and go with a P channel solution. The N channel high side switch let me put the current sensor at ground, and it let me use back to back FETs for total isolation when the circuit is off. Additionally, an off channel only has ground connected to the outside world instead of the full voltage. These were my reasons but I am not sure they are good enough reasons?


It sounds like your reasons are for a high-side switch, not for N-channel over P-channel. You’ll still have to do a little bit to drive a P-channel from a microcontroller, so if you have an N-channel solution that is working well for you, you can probably stick with it unless you are planning on making many units and you need to worry about the parts cost.

- Jan