MOSFET fall time

[carrotSnack]

I've been trying to switch this n-channel mosfet to switch a transformer powered by a 24v battery at 5kHz. As nothing was happening I removed the transformer and looked just at the mosfet with an oscilloscope and saw it was constantly on. I tried an npn instead and it worked fine (I want to use the mosfet though for it's faster switching and higher current rating).

I took the frequency down to 100Hz for a better look.
This is the 5v square wave I'm using:
**broken link removed**

This is switching the 24v battery with the npn:
**broken link removed**

And this is the same with the mosfet:
**broken link removed**

It takes about 5ms to fall from 24.6v to 0v so it's no surprise at gave me a constant 24.6v at 5kHz. The datasheet though says it has a typical fall time of 355ns.
I tried using a different mosfet I had and the fall time was only a little bit less, still not enough for 5kHz.
I'm still learning so have I somehow damaged the mosfet or am I not using it correctly?

 

[ronv]

A schematic may be needed.

 

[carrotSnack]

I'm just testing the mosfet, there's no other parts.
I have the square wave going to the gate of the mosfet and I'm looking at the voltage between the battery positive and the mosfet source. All the parts are grounded.

 

[ronsimpson]

I'm looking at the voltage between the battery positive and the mosfet source. All the parts are grounded.

So you have the MOSFET across the battery.

 

[carrotSnack]

Yes, I have the oscilloscope probes at the battery positive and the mosfet source (nothing else between them though, there's no current flowing), the wave source connected to the circuit ground and the mosfet gate, the mosfet drain connected to the circuit ground and the battery negative connected to the circuit ground. Nothing else.

 

[Roff]

Your description makes no sense. That is not the way an N-channel MOSFET is used.
Also, large MOSFETs have large capacitances. You can't realistically test them with no load.
You need to post a schematic of your test circuit.
Attached is how I interpreted your description.

 

[carrotSnack]

I put a load on it and it instantly worked as expected. I'm still not familiar with all the characteristics of the components so I just expected I could swap one type of transistor for another.
Roff, your interpretation was how it was set up. Thanks for explaining how to use it right.

 

[Roff]

I put a load on it and it instantly worked as expected. I'm still not familiar with all the characteristics of the components so I just expected I could swap one type of transistor for another.
Roff, your interpretation was how it was set up. Thanks for explaining how to use it right.

If it were really connected the way I drew it, the substrate diode would always be conducting, and you would only see about 0.6V change across the load, if you put the load in parallel with the probe that I drew. In your description, I think you have the source and drain swapped. It works if you switch them around.

 

[carrotSnack]

Whoops, obviously that is how I had it set up. For some reason when typing I was thinking that current went into the source and didn't bother actually looking at how I had the pins connected.