FET not behaving as expected

[HappyOldtimer]

Hi all,

I've been following the build instructions, and I am coming across something unexpected:
The vacuum pump turns on as soon as the TinyG has power, regardless of laptop or software input. It seems to be fluctuating with something like a random port voltage.
When I tick the box in the software 'Pump on', the pump now makes more noise/runs faster, and stable. When I remove the checkmark in the software the pump resumes the fluctuating/random noise/running.

Just to make sure of the setup in general:
- everything is wired up according to the instructions
- I am using all the electronics from the kit
- TinyG seems to respond alright and communication with software is possible
- the pump still runs and fluctuates with the control wire to the tinyG disconnected, so the FET doesn't seem to be switching at all based on the control pin (!)

Thoughts/questions that I have currently:
- I have not connected the GND from the 3.3 side of the tinyG to the ground of the power supply. Might that be an issue?
- Could I have smoked one or both FETs in the process of soldering?

I'm hoping that someone with more experience with FETs might make something of this.
For myself, I'd probably be stuck changing the FETs to SSRs.

Thank you in advance

Regards,
Paul

[JuKu]

Sounds like current is flowing where it shouldn’t. I haven’t seen this before and couldn’t point to the problem right away. Emails are sent; we’ll put the solution here for others once we have found it.

[AnalysIR]

If you have the wires to the Drain/Source of the MOSFET swapped, there could be current leaking thru the protection diode of the MOSFET .

???

So verify you have the +Ve lead from 24V going to the drain for the MOSFET and the connect the source of the MOSFET to the =ve supply of your Pump.

See Datasheet linked below for pin-out,

LP default STD30NF06LT4 MOSFET

Of course, if as you say it is all wired up correctly, then the above could not apply.

Thoughts/questions that I have currently:
- I have not connected the GND from the 3.3 side of the tinyG to the ground of the power supply. Might that be an issue?
- Could I have smoked one or both FETs in the process of soldering?

The first point needs to be done as you suggested (common grounds) & the 2nd point(faulty/damaged MOSFET) is a candidate once everything else is verified as OK.

[HappyOldtimer]

Hi AnalysIR,

Thank you for your reply, any and all thoughts highly appreciated

Fair call, if I'd have the wires swapped. I checked, doublechecked and triplechecked. This is how the wires are connected:
- On the vacuum pump: I hooked the connector with the red dot to the 24VDC power supply + (I have checked the voltage on the power supply, it's 23.9/24 volts, the pump is now taken off the aluminum LP frame)
- I hooked the other pole of the vacuum pump to the tab of the FET, connection 2 in the schematic and the Drain if I'm correct
- I hooked the right side pole of the other side of the FET to GND, connection 3 in the schematic and the Source if I'm correct, which is now all connected to the power supply GND
- I hooked the left side pole of the other side of the FET to the TinyG control pin, connection 1 in the schematic and the Gate if I'm correct.

So as far as I can see, all connections should be ok.

I have followed the instructions by the way and reset the TinyG to conservative values in software according to the instructions Juha provided in the installation documents. So I guess the pump should be off by default, right?
However, I did measure the voltage on the 'spin' control pin after startup of the TinyG, and the voltage on the control pin of the TinyG remains around 1v when the pump should be off. That I cannot explain.
When I measure the voltage on the 'spin' control pin (for the vacuum pump) when I turn the pump on in the software, it rises to about 2 volts.

Odd thing that I only noticed now: the 'cool' pin for the solenoid remains at 2 volts no matter what I do, and the light on the TinyG stays on no matter what I do. Turning it off in software has no effect whatsoever...

I believe we can rule out a faulty TinyG: When i disconnect both the 'spin'/pump and 'cool'/solenoid control wires to the FETs, the light on the TinyG for 'cool'/solenoid stays off and both control pins show 0 volts, as they should.
Maybe something with the FETs after all, if the solenoid is also behaving oddly?

Regards,
Paul
PS I'll be traveling this week, so I'll probably only be able to read/respond in about a weeks time

[JuKu]

I'm not able to experiment myself today, as I'm not in my lab. However:
- Because of part shortage, I changed the FET type recently. They should be identical, but just in case: Which part are you using?
- To rule out bad TinyG, please disconnect both he solenoid and the pump, and measure the pin voltage with pump/vacuum on and off. You should see 0/close to 3.3v.
- Assuming a good TinyG, add a small (say, 470Ohm or 1k) between the the FET and TinyG pin to provide some extra isolation. I would actually be somewhat surprised if that changes anything, but as I don't know what is going on...)

I would find it odd if both FETs would be bad. The bad suspicion would be that the other FET is not actually good after all.; then I have a rather big problem.

[HappyOldtimer]

I've checked, in short:
Both FETs are the same, the housing says D30NF06L GF36S027 (S might be a 5, not sure)
With both solenoid and pump disconnected, voltage on both control pins is 0 volt when turned off in software (lights are off too), and 3 volts when turned on (lights come on then on TinyG)
Adding a resistor between de FET and the control pin on the TinyG did change things, and the result is a bit surprising to me:
- with a 560 Ohm resistor, the voltage on the pins changed from 1v when off on the pump to 0.5v. With the 560 Ohm resistor, the voltage on the solenoid pin changed to 0.3v.
- with a 1KOhm resistor, the voltage on the pump pin is now 0.03 in the off state, and for the solenoid 0v in the off state, both are 3v in the on state

Regardless of resistor value so far though, the pump keeps running. I can hear that when the TinyG boots up, it tries to shut it off, but then it just continues.

To my admittedly inexperienced eyes, it looks like the TinyG is trying to pull the pin low, but is somehow prevented by the FET? The TinyG also turned into error mode several times when I tried to measure the voltage, when I had just connected the ground side of my multimeter and not the other side yet... But not consistently. I find it all rather weird.

EDIT: addition
When I provide 24v to the pump, it runs just fine, so the pump isn't broken, and the connections there are ok.
When I just connect the solenoid, it shows the same reaction as the pump:no matter the settings in software, it won't shut off. So this seems like the behaviour I'm seeing from the pump.

[AnalysIR]

Have you fixed the grounds?

Otherwise, Connecting a 10k resistor from the gate to ground of the MOSFET could also help to deactivate it until actually turned on.(Although normally should NOT be required)

[JuKu]

That is the FET used for LitePlacer since the beginning. TinyG also works as expected (0/3v). As noted, I would think it is rather unlikely that both FETs would be bad and behave the same. As AnalysIR said, it sounds like a ground problem. It could also be a bad connection, and something that looks connected actually isn’t.

[AnalysIR]

With both solenoid and pump disconnected, voltage on both control pins is 0 volt when turned off in software (lights are off too), and 3 volts when turned on (lights come on then on TinyG)

You should measure this control pin voltage when referenced to the GND(not Mains Earth) of the power supply feeding the 24V to the MOSFET, which is what really matters.
(Red lead to Control pin output & Black lead to Power supply GND)

If you still get a steady 3V3(ON) & 0V(OFF), then the problem is likely elsewhere.

[infinitemach]

I had a similar problem when building my LitePlacer, when the vacuum was switch on in software it would run normally, when switched off it still ran, and as if it was getting lower voltage. I replaced the FET and everything worked normally. I never figured out how the FET became damaged.