The Td(on) and Td(off) parameters are simply the turn on delay time and turn off delay time, respectively. When they say 'delay' how are they defining when the channel is considered to be not delayed? In terms of Vgs, they say the channel initially opens when Vgs = Vth, which is the voltage at which 100uA passes through the S-D.

Since there is a Tr (rise time) also thrown in the mix, what exactly is going on here?

 

The conditions for the delay and rise time measurements are typically given in the data sheet for the MOSFET.
Below is an example for a BUZ11 device:
As you can see, td(on) is the time from the input pulse rise to the start of the MOSFET turn-on, and tf and tr are the 90% to 10% waveform fall and rise times.

 

In terms of Vgs, they say the channel initially opens when Vgs = Vth, which is the voltage at which 100uA passes through the S-D.

The Vth value has only a small effect on the switching times for a 10V input signal with a fast rise/fall time.

 

The Vth value has only a small effect on the switching times for a 10V input signal with a fast rise/fall time.

Since the rise/fall times don't necessarily coincide to the 50% point of the pulse, how would I go about ensuring my pulse width is what I would want it to be? Dividing the td(on) and td(off) parameters by 2 was my first thought but the pulse doesn't directly match up with those.

 

how would I go about ensuring my pulse width is what I would want it to be?

If you are talking about the 50% point pulse width, you would subtract the rise-time from the fall-time and add that signed value to the input pulse width.
How close do you need the output pulse width to be, compared to the input pulse width?

 

Since the rise/fall times don't necessarily coincide to the 50% point the pulse, how would I go about ensuring my pulse width is what I would want it to be? Dividing the td(on) and td(off) parameters by 2 was my first thought but the pulse doesn't directly match up with those.

Do you have an oscilloscope? Could you start with settings that get you close, then just check with oscilloscope and fine tune from there?

How long is your desired pulse, relative to rise and fall times? How precise do you need the width to be?

 

Do you have an oscilloscope? Could you start with settings that get you close, then just check with oscilloscope and fine tune from there?

How long is your desired pulse, relative to rise and fall times? How precise do you need the width to be?

I do have an oscilloscope in our university lab, and will end up doing a lot of testing after I get everything put together. The problem is that I need to order the parts and have a limited amount of time to do that left. It will be important to make sure that I consider all factors before ordering a power mosfet of some type to get what I need.

Essentially the pulse will be driving a 5MHz transducer, and these typically are excited by a 1/2 cycle pulse of high voltage. So I need a 100ns pulse, preferably less than 100ns if it was going to be +/- some margin of error.

I do not know how accurate it needs to be, but I'm hoping to get within 50ns to 125ns range.

 

If you are talking about the 50% point pulse width, you would subtract the rise-time from the fall-time and add that signed value to the input pulse width.
How close do you need the output pulse width to be, compared to the input pulse width?

Since I'll be using a power mosfet as the switch, I wasn't really thinking of the 'input' as a pulse. Ideally I'm trying to get a 100ns pulse to excite a 5MHz transducer.

Some of the mosfets I looked at had really high rise/fall times relative to 100ns. (tr = 16ns, tf = 45ns, td(on) =24ns, td(0ff) = 62ns) Which in total would be literally double my goal width. I'm going to be looking for different ones, but if I had no choice would the high/rise fall times cause irreconcilable problems in making a 100ns pulse?

 

What current and voltage does the transducer require?

 

What current and voltage does the transducer require?

Unfortunately, the manufacturer didn' have any sort of datasheet on the probe. PT-12 by Yushi Instruments. I was able to find out that it operates by at short pulse of 220V by emailing them. In terms of current, I know atleast that the input pulse will be less than or equal to 4A and quickly decrease as the capacitor discharges over the 100ns.

 

Do you know the value of the capacitor?

 

Do you know the value of the capacitor?

This is the schematic I'll be using.