the Best way of switching a load at high speed?

[tbrew]

Hi, I'm working on a bit of a strange project, the aim is to verify the presence of a small capacitor which is in parallel with a small resistance in the range of the 10's of Ohms. The idea is to switch the load at a very high speed (10MHz +) and then use a high speed peak detection circuit and feed the output into a ADC... The peak detection circuit is designed and working well, I now need to drive the circuit from the board instead of the function generator. I have a voltage controlled oscillator to provide the control signal, Im just wondering which is the best way to switch the load. many FETS have turns off times up to three times larger than the turn on times and I would like to maintain a 50% duty cycle if possible.

Thanks for any suggestions.

[amspire]

Sounds like overkill.

How small is the capacitance?

If it is, say, 200pF or more, just supply a fixed amplitude 10MHz waveform (a square wave will probably do) from a 100ohm resistor and measure the voltage across the 10ohm resistor. Calibrate for no capacitor, so when the capacitor is there, the voltage will be lower.

If you are talking really small, like in the low pF, you have to take into account that both the resistor and the board layout have a signifigant capacitance of their own.

Also they type of resistor is important. Some have enough inductance to get some odd responses at 10MHz.

Richard

[tbrew]

Sounds like overkill.

How small is the capacitance?

If it is, say, 200pF or more, just supply a fixed amplitude 10MHz waveform (a square wave will probably do) from a 100ohm resistor and measure the voltage across the 10ohm resistor. Calibrate for no capacitor, so when the capacitor is there, the voltage will be lower.

If you are talking really small, like in the low pF, you have to take into account that both the resistor and the board layout have a signifigant capacitance of their own.

Also they type of resistor is important. Some have enough inductance to get some odd responses at 10MHz.

How is that any different to what I have stated above?

The peak detector and ADC is to measure the voltage across the 10 Ohm resistor. Variable frequency is to tune the device incase 10MHz produces some odd responses

[amspire]

I obviously misunderstood what you meant by "switching the load".

If you just want to be ably to switch on a a 10MHz signal, then it sounds easy, unless you have special requirements. There are lots of really easy ways to do it.

Also you don't need a voltage controlled oscillator or an ADC unless you want to accurately measure the capacitance rather then just verify the presence of the capacitance.

At this stage, I am not sure why you are talking about FET turn on and Turn off time. What is wrong with grabbing a 10MHz, 20Mhz or whatever from a micro if you have one, pass it through a nand gate with the second input to enable/disable the signal to a resistor that is connected to the 10 ohm?

Richard

[ejeffrey]

I agree, the best bet is to use your clock oscillator and drive a discrete logic gate.  If you need more current than a single 74HC series gate, you can use a hex-NOT IC and connect 2 or more gates in parallel.  Make sure you give the logic IC its own bypass capacitor.

If you need precision measurement you can do synchronous detection as in an LCR meter.  This would give you both R and C and probably work at lower frequency (like 100 kHz).

[tbrew]

Thanks for the advice I didn't even think about logic gates as I was think the current would be too high for a single gate.