N-channel MOSFET as a switch. Dying somehow?

Thread Starter

SkyForCe

Joined Jan 26, 2019
6
Hey guys,

currently working on my diploma for school. Trying to build a guitar playing robot. So far so good, but when designing the circuit which controls the pushing magnet (don't know how it's correctly called in english, datasheet attached, couldn't find an english one) I am running into problems.

I got a PCB with 10 Inputs which I get from an myRIO, which outputs 3,3V to control each MOSFET. When the MOSFETS gets the 3.3V it should let the current flow, and if not it should stop it. At the end i connected my 2 pushing magnets which should presss the strings on the guitar.
When I tested each lane everything worked fine except 2 lanes. I began to measure and test if there was a shortvut womewhere, but I couldn't find one. So I replaced a MOSFET and tried again. And look, it worked now.

But then I realised that another lane had stopped working. Maybe I destroyed it in the measuring process? I don't quite know.
Hope you can more or less follow me here. I attached my schematic.

The first time I tested I had 120 Ohm Resistor on the Gate.
Later I also tried 620 Ohm and 3k3 Ohm Resistors at the Gate. It worked all the time. The resistor shouldn't be that important anyway due to me just driving the thing with an myRIO which can be compares to something like an Arduino and I don't have huge Voltages on the Gate.

The pushing magnet is powered by 24V and usually takes about 300mA. (for more view the datasheet)

Is there a problem with my circuit? Is the Resistor wrong? Any recommendation?

Thank you very much in advance,
Hope you can help me. That's the first time I am working with MOSFETs
 

Attachments

djsfantasi

Joined Apr 11, 2010
6,965
There’s a slight language problem. For example, when you say “pushing magnet”, do you mean solenoid? That is an electromagnet, a coil of fine wire that produces a magnetic field when electricity is applied.

Can you show the schematic of everything used to activate your pushing magnet?

Do you know what a snubber or flyback diode is?
 

Thread Starter

SkyForCe

Joined Jan 26, 2019
6
There’s a slight language problem. For example, when you say “pushing magnet”, do you mean solenoid? That is an electromagnet, a coil of fine wire that produces a magnetic field when electricity is applied.

Can you show the schematic of everything used to activate your pushing magnet?

Do you know what a snubber or flyback diode is?

Thank you for helping.
Just googled solenoid and thats exactly what I meant.

Well, isn't that my antiparallel Diode I put in to prevent the current/voltage spikes from damaging the MOSFET when the solenoid is getting turned off?

The attachment is pretty much all I've got on my PCB. With the only difference that it exists 10 times.

I attached the full schematic but be aware that the diode in the schematic is wrong and actually got built in the other way around.
I also uploaded an english datasheet of the used MOSFET
 

Attachments

wayneh

Joined Sep 9, 2010
16,400
3.3V on the gate is marginal, even for a logic-level MOSFET. Is your MOSFET getting hot?

If the diode is not reverse biased, that’s your problem!
 

Thread Starter

SkyForCe

Joined Jan 26, 2019
6
No, the MOSFET isn't getting hot at all.

What do you mean by Diode is not reverse biased? You mean not antiparallel?
If you mean that, then you might not have written that the diode in the latest schematic is drawn wring and hasd been put onto the PCB the other way around.

My english sadly isn't perfect.
 

Norfindel

Joined Mar 6, 2008
326
In the last schematic you uploaded, you're just shorting 24v to GND with a mosfet and a diode. Well, 10 mosfets and 10 diodes, really. I'm surprised it didn't go instantly thermonuclear.

That's not what you have to do. The first schematic you uploaded "Hubmagnet-Transistorschaltung.jpg" was correct for one channel.
 

wayneh

Joined Sep 9, 2010
16,400
The first schematic you uploaded "Hubmagnet-Transistorschaltung.jpg" was correct for one channel.
Agreed. I still wonder if the gate voltage is adequate to avoid MOSFET overheating. The datasheet shows it should be fine, as long as the voltage is not sagging somehow.
 

wayneh

Joined Sep 9, 2010
16,400
I got a PCB with 10 Inputs which I get from an myRIO, which outputs 3,3V to control each MOSFET.
When the output goes low, does it truly go low, or "open"? Perhaps you need pulldown resistors on the gates. Those are a vey good idea anyway, as they ensure that the MOSFET is turned off when the control signal is indeterminate or missing.
 

Thread Starter

SkyForCe

Joined Jan 26, 2019
6
Agreed. I still wonder if the gate voltage is adequate to avoid MOSFET overheating. The datasheet shows it should be fine, as long as the voltage is not sagging somehow.
That shouldn't be a problem. It didn't get hot when I was testing and the MOSFET will be turned on for a maximum of like 10 seconds with breaks inbetween.
Thanks for the suggestion. I'll try to do a test build with a pulldown on the gate. What dimension would be best for the pulldown?
 

Thread Starter

SkyForCe

Joined Jan 26, 2019
6
In the last schematic you uploaded, you're just shorting 24v to GND with a mosfet and a diode. Well, 10 mosfets and 10 diodes, really. I'm surprised it didn't go instantly thermonuclear.

That's not what you have to do. The first schematic you uploaded "Hubmagnet-Transistorschaltung.jpg" was correct for one channel.
In the second schematic there is a connector for the solenoid, which also exists for the other ones. I now also changed the direction of the diodes.

It is my first design and there are definitely a lot of improvements to make.
 

Attachments

jayanthd

Joined Jul 4, 2015
904
That shouldn't be a problem. It didn't get hot when I was testing and the MOSFET will be turned on for a maximum of like 10 seconds with breaks inbetween.
Thanks for the suggestion. I'll try to do a test build with a pulldown on the gate. What dimension would be best for the pulldown?
10k pulldown for all mosfet gates. Mosfet gates to ground through 10k.
 

LesJones

Joined Jan 8, 2017
2,726
As the driver board is connected to the mosfet board with a plug and socket (As shown in post #3) and the mosfet board provides no protection to prevent static damage to the mosfets you will need to be careful not to touch the input connections. You schematic does not show any ground / common connection on the connector. Do you have a proper ground path between the driver board and mosfet board ? When connecting the boards connecting the ground path before the signal paths would be a wise precaution. You should also be using an antistatic mat and grounding strap while working with mosfets. The suggested pull down resistor on the gates will make the mosfets less susceptible to static damage.

Les.
 

Thread Starter

SkyForCe

Joined Jan 26, 2019
6
Okay guys, thanks for being patient with me.

I learned a lot and I am going to be more careful and also add pull down resistors.


Have a good one :)
 
Top