PNP Transistor state change

[drkidd22]

Hello,

I have this circuit which is partially working.
When I set the signal from the pic to LOW D32, D3 and D45 turn on and the output from U59 changes state.
The issue is when I try to turn it off, only D32 fully turns off. The gate of Q45 is therefore kept low thru the pull down.
But D3 and D45 remain on, slighlty, I tried adding a 10K pull up to the base of Q1, but that didn't help. Any input will be appreciated.

 

[MikeMl]

Where is the pull-up on pin 6 of U59?

 

[Nigel Goodwin]

Where is the pull-up on pin 6 of U59?

He could be using the weak pullup in the PIC, assuming he's using a suitable pin?.

 

[MikeMl]

Is there more to this circuit; like does the drain of Q45 drive something besides the base of Q1?

 

[drkidd22]

That's all there is to the circuit. The Q45 is not driving anything else. There is a pull up on the pic side for pin 6 of U59, but the issue I'm see is between Q45 and Q1.
It is a test circuit for now, but I will use Q45 to drive something else later. Want to understand what;s going on here first.

 

[crutschow]

Why do you need two optoisolators?
One should be sufficient for any isolation requirements.

 

[Mikebits]

The logic is goofy. You might get what you want by changing R154 to a 1kΩ, this should pull down your gate enough, but still enough gate voltage to turn it on when the opto detector is on. This was the simplest change I could think of that just required 1 simple change. If this is done on breadboard, you should rethink the whole logic scheme, such as what cruts is alluding to. See LTspice file, I did not have exact models, but mosfet is close. I kinda rushed this so look for mistakes guys

 

[spec]

Hi dr,
Welcome to ETO

The only thing I can see that may be causing your problem is if the MOSFET is oscillating. You could try putting a 22 Ohm resistor in series withe the gate of the MOSFET. The resistor needs to be mounted directly on the gate pin. Also, connect a 100nF or larger ceramic capacitor directly between the source and drain pins of the MOSFET.

Although the circuit should work, it has a couple of design issues and it is more complex than need be. Also a massive 16A MOSFET is not required, although I do appreciate that you may want to drive a heavy load later on.

spec

 

[Mikebits]

Although the circuit should work, it has a couple of design issues and it is more complex than need be. Also a massive 16A MOSFET is not required.

That is interesting notion (oscillation) which the 10k to 1k I suggested may fix. I agree with you and cruts on points mentioned above.

 

[kubeek]

Q1 has collector and emitter swapped.

 

[crutschow]

Q1 has collector and emitter swapped.

It's not connected wrong, it's just a PNP emitter follower circuit.
But Q1 seems to be unnecessary as I don't see what purpose it serves. You could remove it and the circuit would operate the same.

 

[kubeek]

It's not connected wrong, it's just a PNP emitter follower circuit.
But Q1 seems to be unnecessary as I don't see what purpose it serves. You could remove it and the circuit would operate the same.

My mistake, I overlooked the arrow direction.

 

[spec]

My mistake, I overlooked the arrow direction.

Glad it's not just me that makes mistakes.
spec

 

[AnalogKid]

The circuit impedances do not look high enough to need this, but you can try adding a large value resistor from U41 pin7 to pin 5. Some optocouplers don't turn off completely without a little help.

ak