Hey all - First, I am not an EE (though I really wish that were my field). i'm a software engineer fighting my way through making a trimmed down version of speeduino (open source engine management based on arduino) for a freestyle jet ski application. My board just does ignition control and a few GPIO. I need to make a latching start/stop circuit with two different button inputs to interface with typical lanyard start/stop switches. I tried using a bipolar mechanical latching relay, but it seems the vibration will make that unreliable. I've pieced together something that seems to work and I tested it out with an IRF540N mosfet and two 2N3904 transistors. it seems to work ok but I have no idea what I'm doing - I'll have to find suitable surface mount alternatives when I add this circuit to my board design.

Any advice, feedback, tips, etc is greatly appreciated.

here's a link to the simulation https://tinyurl.com/2hvjp6xx

 

oh also just to be clear, the LED in that circuit is taking the place of the speeduino - where the LED is would actually be a connection to a LM2940 5V regulator

 

I tested it out with an IRF540N

The IRF540 is a N channel mosfet. Your circuit shows a P channel. Need to check part number.

 

oh crap you are correct. ya I did test it with an IRF540N. I suppose the start/stop was just inverted. I changed the simulation to use an N channel (imagine the start/stop text is inverted).

 

The N channel will not operate to full saturation as a high side switch. You need to use a P channel mosfet you show in post 1.
Use a IRF9540.

 

The N channel will not operate to full saturation as a high side switch. You need to use a P channel mosfet you show in post 1.
Use a IRF9540.

Ok cool that’s no issue. I have some I can test with. Is everything else ok?

 

I believe so. How much current does that speeduino require?

 

Ok cool that’s no issue. I have some I can test with. Is everything else ok?

Not much. I think around .1 amp or so at the most. I do have two “high current” outputs that use mosfets to drive 12v out so shouldn’t have an impact on the draw on this circuit.

 

If it's only .1 Amp the IRF9540 would be major overkill.

 

If it's only .1 Amp the IRF9540 would be major overkill.

What would you suggest for a surface mount common mosfet? How bout https://datasheet.lcsc.com/lcsc/1808272021_Vishay-Intertech-SI2301CDS-T1-GE3_C10487.pdf this? Very cheap and JLC has a ton in stock.

I was originally hoping to just use two mosfets for the latch and only have two active components but I couldn't figure out how to make that work.

 

Gate voltage too low, need something at + or - 20 volts.

 

FDC658P mosfet. Available at Digi-Key for .70

 

ok maybe this then https://datasheet.lcsc.com/lcsc/1811081714_Infineon-Technologies-IRLML5203TRPBF_C2592.pdf

 

FDC658P mosfet. Available at Digi-Key for .70

cool thanks i'll check it out

 

ok maybe this then https://datasheet.lcsc.com/lcsc/1811081714_Infineon-Technologies-IRLML5203TRPBF_C2592.pdf

Sure, that will work too.

 

So I got my latest board revision and the darn latching circuit is not working. I tested it out before ordering with components on a breadboard and everything was fine. Granted, I switched to similar surface mount components but I'm really stumped. I can never power it off - the MOSFET just always has power running to the gate. One note, I did have to remove D4. For whatever reason, this diode heated up and drew a ton of power on first power up. Removed it and all seems fine with power regulation. Attached is the full diagram. I tested the transistors in the latching circut and all internal diodes seem fine. MOSFET seems fine as well. I have a second board that does the same thing. Really appreciate any advice here, I'm stumped.

 

I THINK I might have just figured it out. Drain and Source on the MOSFET might be swapped.... will take some rigging to test that....

 

One note, I did have to remove D4. For whatever reason, this diode heated up

Installed backwards or shorted. Don't see why it was needed though.

 

Installed backwards or shorted. Don't see why it was needed though.

Yup swapping drain and source fixed it! You mean the mosfet wasn’t needed at all? Ya I wasn’t sure about that honestly.

 

I was saying D4 was not needed.