I've just found the following circuit on a board I'm repairing. There are 3 N channel mosfets and 1 P channel mosfets driving the relay coil.
The signal is provided by two 5V micro pins (labelled IO7 and IO8 on the schematic).



I'm kind of struggling to understand what makes it tick.

 

I've done a bit more working through the circuit and think a pulse from IO8 would latch the circuit on, keeping the relay conducting without the micro having to constantly output.
And then a pulse from IO7 would turn the relay off ? I can't figure this out - time to look at how P channel mosfets work.

 

There are 3 N channel mosfets and 1 P channel mosfets

hi fb,
With the +12V supply at that position, all the MOSFET internal diodes are forward biassed.
Are you sure that your drawing is correct.?
E

 

LOL, I've just noticed that. For some reason my schematic program has inverted them. I will amend and repost ... soon.
Thanks

 

So here is the corrected schematic ...

 

hi,
OK,
Check up on your N and P type MOSFET's .....
P have a Neg voltage on the Drain wrt the Source and N types have a Pos voltage on the Drain.

E


Update:
Are you sure the relay code is correct, ie: Non latching.???

 

Thanks for the help.
I will check the relay code tomorrow. In the meantime I have taken another look at the devices and had transposed my N and P channel devices, so now the circuit schematic looks like this

 

hi fast,
If the relay was a latching type, the change in voltage polarity on the coil would make sense.???
E

 

Hmmm, the relay is marked Finder 40.61 on the side, but on the top is then marked 12V BS. My guess is BS=bistable -> latching relay. So the correct replacement part will be FINDER_40.61.6.012.0000.

So one IO is the On pulse and the other is the off pulse.
It's getting late here, I'll go to bed and sleep on it and see if it all makes more sense in the morning.

 

I still struggle to understand how the circuit works, so wonder why they didn't just run with something like this

The BJT could be mosfet too. R6 and R7 would need to be changed to work with the relay's coil current and the 12V supply.
circuit from https://forum.mysensors.org/post/60174

 

hi fast,
I will create and run a LTSpice simulation of that first circuit, let you know when done.
E

 

hi fast,
Here is the function of the circuit, I would say for sure the Relay is Latching.
Note the current reversal.
@fastbike
E

 

So my reintepretation of the circuit in post #10 above would be something like this


I just need to size resistors R7/8 to match the current requirements of the relay coil.

 

hi fast,
Here is the function of the circuit, I would for sure the Relay is Latching.
Note the current reversal.
@fastbike
E
View attachment 232924

Thanks for the simulation- it looks like our posts crossed. I'll take another look at your simulation.
My question is why are 4 mosfets required in the board I'm repairing when other circuits I have found only show 2. What advantage does the extra component count bring ?

 

hi fast,
The original circuit is exclusive, either A or B but NOT A&B or /A&/B

E

 

hi fast,
The original circuit is exclusive, either A or B but NOT A&B or /A&/B

E

Nice, I did not catch that feature, so hats off to the original designer.

 

hi fast,
Your circuit looks OK to me.
E

@fastbike

Corrected the image to show the correct A, B. sequence

 

So part FINDER 40.61.6.012.0000 seems to hard to find in stock.
Would Omron G5RL-U1-E-DC12‎ work as a substitute?
It has the same footprint, capacity, coil voltage etc.

 

hi fast,
The spec's look similar, it should be possible to use this alternative.
E