Oscilloscope probe pulls signal low

Hi,

I have a pic connected to a MOSFET switch. One of the outputs of the PIC should enable/disable the MOSFET.

However, I have found that the MOSFET is switched on at all times - regardless of which state I program the PIC output to be.
If I touch the output trace with an oscilloscope probe to check its state, the signal gets pulled low and MOSFET switched off.

Can anyone shed any light on this??

Need to use a gate-source pull-down resistor to help pull the microcontroller output voltage down to 0V when the MOSFET is switched "OFF".

But then why would I not be able to measure the signal voltage level using the oscilloscope?

I can understand that the pull down will help pull the signal "low". But if I hold it high using the PIC, should this not be measurable? or even if I held it low?

The actual components I use are a sip32411 and a fdc6330L load switches - both enable pins connected to the ouput of the PIC. In the datasheet it states that the decide works with a voltage control logic interface without extra level shifting circuitry. I assumed this meant that no pull ups/downs will be needed either.

Audette said:

But then why would I not be able to measure the signal voltage level using the oscilloscope?

Click to expand...

What are you seeing at the oscilloscope?
I was wandering whether the PIC output is by chance configured as an input. If I recall correctly, you can write something in the output buffer of the PIC enven when the pin you are using is configured as an input. In this case, you have the very high impedance input pin of the PIC connected with the very high impedance gate of the MOSFET. You might have a situation for which you store enough charge that the MOSFET is in the ON state until you put your probe on the node. If your probe is 1:1, you have 1 Mohm towards the ground (If you have a 10:1 probe, it is the same but the resistance is 10 Mohm), which is enough to drain the charge and commute the MOSFET in the OFF state.
If there is a possibility that the pin controlling the MOSFET goes to high impedance, it is a good idea to put a pull down resistance to control the logic state. In some cases, a low value resistance in series between the gate and the output pin of the PIC can be useful with some MOSFET to prevent self oscillation (R_IN in the scheme provided by mister_rf).

Most likely in this situation the processor configured with output pins in high impedance mode, and leaving the outputs in an undefined state may causes this problems, so it is recommended to tie the MOSFET load switch input to GND. The external pull-down resistors, if provided, should block the load switch for the inactive state at the microcontroller output. The value of the pull-down resistor should be chosen such that not to increase the output current, usually a 10k up to 100k resistor can be used for.

You need to show your code as this sounds like a software problem.

Keith.

Thanks for all your input. After a bit of butchering of the PCB found the problem.

It was a SW problem after all... The pin was not properly initialised as an output - therefore the signal was left floating! I had a function call to initialise the pin as an output - but never called the function!!!!Yes kicking myself right now for not seeing this straight away!

From all that was mentioned here, I think I will put in a pull down resistor just in case as well.

Thank you again for your input.