Vgs_max on low side MOSFET driver, kick-back diodes, decoupling capacitors

Using Arduino General Electronics
1

Hello everyone,

I'm revisiting and upgrading an old circuit. Part of the upgrade is swapping out relays for low side mosfet switches/drivers:

Watering_System_3.0.png
Watering_System_3.0.png6115×4115 207 KB

Low-side MOSFET driver:

The load I'll be switching are small 12v, 100mA solenoid valves and a 12v 6W brushless motor on a pump.

Earlier this year, when learning about high-side/low-side switches, @larryd posted this overview schematic:

DiscreteMotorControl.jpg
DiscreteMotorControl.jpg2128×1600 538 KB

In it, it says that when using a logic level low side driver, the V_gs of the MOSFET needs to be >12V. In Nick Gammon's explanation on the matter, that need is not mentioned.

I don't understand where the restriction comes from: The source is always connected to GND (0V), the gate is either pulled low (0V) or sees logic high (5V) by the µC pin. That means, it will never be able to exceed a 5V delta, correct? Why the need for a Vgs_max of >12V then?

Im planning to use this IRLML6244 logic level N channel mosfet for the job, which has a Vgs_max of +/-12V. Should I reconsider?

While I'm at it (sanity check), for the high side swich connecting the VCC to the analog sensors and the SD card module I plan to use this IRLM2244. It works fine in my prototype, but as you can tell from the example above, I'm not 100% sure I know what I'm doing here.

Diodes

I'm planning on using 1N4148 diodes for compensating the inductive kickback of the solenoids and a 1N4007 for the motor switch. That should work fine in my opinion, or would anybody strongly recommend using a schottky diode instead?

Capacitiors

Do the 100n caps parallel to the valves make any sense, given they are powered by their own 12V supply? Or are those wasted?

Thanks in advance for your input!

Watering_System_3.0.png
Watering_System_3.0.png6115×4115 207 KB

DiscreteMotorControl.jpg
DiscreteMotorControl.jpg2128×1600 538 KB

2

I see no reason for Vgs to be more than 12V, as you say it's never going to see more than 5V. Probably a specification above 8V to give a margin of safety.

The diodes for absorbing the kickback must have an If that exceeds the maximum current drawn by the inductive load they are connected to. Don't forget that for a motor that means the stall current. As a matter of convenience I would generally consider 1N400x before anything else.

No harm in the capacitors but I would not bother.

3

Thanks Perry!

PerryBebbington:
I see no reason for Vgs to be more than 12V, as you say it's never going to see more than 5V. Probably a specification above 8V to give a margin of safety.

Alright, glad I'm not the only one to see it that way. As the used MOSFETs have a Vgs_max of +/-12V, they should have plenty of breathing room here in terms of a safety margin.

PerryBebbington:
The diodes for absorbing the kickback must have an If that exceeds the maximum current drawn by the inductive load they are connected to. Don't forget that for a motor that means the stall current. As a matter of convenience I would generally consider 1N400x before anything else.

Unfortunately, I have no datasheet for valves or the motor (cheap eBay import stuff, the brushless motor is part of a mini pump), so I have no information on the stall current. But as it will not see any real load changes (pumping water with constant pressure), it will most likely operate mostly at the nominal rating of around 500mA. The stall current should never be reached.

PerryBebbington:
No harm in the capacitors but I would not bother.

Sounds fair, I'll kick them out in that case. Surprinsingly, they're not all that cheap after all...

4

Unfortunately, I have no datasheet for valves or the motor (cheap eBay import stuff, the brushless motor is part of a mini pump), so I have no information on the stall current. But as it will not see and real load changes (pumping water with constant pressure), it will most likely operate mostly at the nominal rating of around 500mA. The stall current should never be reached.

Any motor draws its stall current when you first connect power, it is, after all, stalled at this point. Measure the resistance of the motor while it is stationary. If you can, rotate the motor slowly while measuring the resistance and take the lowest reading. Divide the supply voltage by the resistance you measured to get the stall current. Add 50% for safety.

5

PerryBebbington:
Any motor draws its stall current when you first connect power, it is, after all, stalled at this point.

Hmm... I agree that that's the case for a shunt wound DC motor - I didn't think that it transfers one-to-one to a BLDC (or any synchronous motor), the same way it doesn't transfer to asynchronous motors.

For synchronous motors, I only know how to calculate the current based on load torque and load angle at synchronous speed/frequency. I have no idea how the current develops during the startup process or how to calculate it in gerneal when using an ESC/frequency converter.

I'll try your method, unfortuntely I can't access the motor shaft. Do you think it makes any sense to try measuring the startup current with a multimeter or will the transient be too short to be picked up?

6

I take your point and I think you are probably correct.

Do you think it makes any sense to try measuring the startup current with a multimeter or will the transient be too short to be picked up?

Yes, too fast for a multimeter, need an oscilloscope and a resistor in series with the motor to drop some voltage to measure.

What I would do it not worry about it too much, probably an 1N400x will be OK.

7

Your question got me thinking that I was fundamentally getting something wrong...

Of course a brushless DC motor is really an AC motor with some electronics to provide the AC from the DC supply. I have just tested a small 12VDC fan and checked with an oscilloscope:

There is no voltage spike to absorb.

There were some small current spikes sometimes, nothing like what you get with an inductive load.

8

PerryBebbington:
I take your point and I think you are probably correct.

Thank you - I may very well be wrong about the whole thing though, most of what's going on inside three phase machines is black magic to me. Especially anything non steady state in a synchronous one.

Either way, thanks a lot for measuring the fan motor! In that case I guess I'll put in the diode for good measure, should I for some reason swap out the motor for a small DC pump or another inductive load later on :slight_smile:

9

You have the "low side switch motor" drawn as HIGH side. In the "high side switch" drawing, what is the value of VCC?
What is S_HSS connected to?

10

JCA34F:
You have the "low side switch motor" drawn as HIGH side

Thanks for the heads up, I'll correct that!

JCA34F:
In the "high side switch" drawing, what is the value of VCC?
What is S_HSS connected to?

Vcc is +5V. S_HSS connects to the signal pin on the arduino, which triggers the HSS in order to power the analog sensors and the SD card.