Why the second MOSFETs in a two-MOSFET series shorts?

Thread Starter

fablau

Joined May 14, 2020
167
Hello everyone.

I am working on a project that involves the use of MOSFETs to turn a matrix of electromagnets on/off. The project is quite complex and involves several different circuits, but for facilitating this specific discussion, I have simplified the circuit to focus on this specific issue only.

The simplified circuit below works perfectly with just an N-Channel MOSFET which turns opens/closes the circuit (the MOSFET named N1 in the diagram), but as soon as I add a second MOSFET in series as shown in the diagram (MOSFET N2) and turn the second MOSFET on, it shorts for some reason between the source and the drain, burning it out. Nothing happens to the first MOSFET. I am trying to figure out why that's happening but I can't make a sense of it.

Is there a specific "protocol" to follow when chaining MOSFETs in series? For example: do I need to turn them on exactly at the same time? Or do I need to turn the N1 first and the N2 second? Or vice-versa? Or all that doesn't really matter and I am missing something else?

I am eager to know your thoughts on this. Thank you!

Schematic_Serial-Mosfet-Discussion_2022-03-13.jpg
 

crutschow

Joined Mar 14, 2008
29,755
You show no spike suppression for the L1 inductance to limit the transient spike when the inductive current is turned off.
Is there one in the circuit?
(If that's what D1 and D2 are supposed to do, they don't, unless they are Zeners.)
If not, you need one (such as a diode in parallel across the inductance, cathode to Vcc).
 

Thread Starter

fablau

Joined May 14, 2020
167
All legitimate questions guys, I was afraid the diagram was over-simplified and could cause confusion and misunderstanding of the whole system. Please, bear with me, I'll post a more self-explanatory diagram later today.

In a few words, the coil is actually bi-directional (can both attract and repel), even though I showed just one direction of the flux. So, there isn't an easy way to suppress transients around the coil (at least, per my own very limited knowledge, but feedback on that will be very welcome!)

What is in the box that says ”From microcontroller”?
What switches the MOSFET off?
Good question: The microcontroller drives an optocoupler that turns on/off a 12v line which triggers the MOSFET gates.

What is this second MOSFET supposed to do, given that N1 on its own switches power to the load?
Ah ah! I knew you were going to ask, my apologies I didn't explain it earlier. The purpose of the second MOSFET is to stop the flux coming from the shown coil to go to other coils connected to the matrix I mentioned after the first MOSFET. As I said, I over-simplified the diagram, and it is actually hard to understand. My mistake and my apologies again for that.

I'll post a new more extensive diagram as soon as possible that should address all your questions.

Thanks!
 

crutschow

Joined Mar 14, 2008
29,755
the coil is actually bi-directional (can both attract and repel), even though I showed just one direction of the flux. So, there isn't an easy way to suppress transients around the coil
Depends upon what you call "easy".
A bidirectional coil can have the transients suppressed using four diodes.
Two go from each end of the coil V+ (cathode), and two go from each end of the coil to ground (anode).
 

Thread Starter

fablau

Joined May 14, 2020
167
Depends upon what you call "easy".
A bidirectional coil can have the transients suppressed using four diodes.
Two go from each end of the coil V+ (cathode), and two go from each end of the coil to ground (anode).
Awesome! Let me post the new diagram later today (it'll take some time to make it, sorry), so we can better discuss all this. Thank you!
 

Ian0

Joined Aug 7, 2020
5,064
A bidirectional coil would normally be driven by an H-bridge, in which case the MOSFET reverse diodes would protect them.
 

AnalogKid

Joined Aug 1, 2013
10,050
there isn't an easy way to suppress transients around the coil (at least, per my own very limited knowledge, but feedback on that will be very welcome!)
Two zener diodes back-to-back (inverse-series) will clamp voltage transients of both polarities. The technique is so common that you can buy a single device with two diodes inside already connected. For discrete parts, try two 33 V or 36 V zeners.

Here is a TVS datasheet. The parts with "CA" are bidirectional.
https://www.littelfuse.com/~/media/...s/littelfuse_tvs_diode_p6ke_datasheet.pdf.pdf

ak
 
Last edited:

Thread Starter

fablau

Joined May 14, 2020
167
Here I am guys, thank you for waiting. It took a while, but I think this will clarify most of your questions.

Here is the circuit as it is right now, and it works great (it has been working without issues for the past few months). I just focused on a single H-Bridge output, but in reality, I have 9 of those from 5 L298N. But the rest is just a duplicate of the one below, so, no need to replicate it.

Sheet_1.png

Now, the goal of the added N-channel MOSFET in series after the N1 above is to just "stop" the propagation to a specific coil if I want to. In fact, as you can see from the following modified circuit, I have added a new N-channel MOSFET (N2) in series after N1, right before the right pole of the coil:

Sheet_2.png

I haven't added flyback diodes on these circuits, but I guess I should (I had no issues without it with the first circuit above). But even by adding flyback diodes as I drew in my first post diagram, the second MOSFET (N2) keeps shorting between the drain and the source.

Eager to know your thoughts and, possibly, any further questions. I hope the diagrams are clear enough.

Thank you!
 
Top