500picosecond Pulse generator

[rineesh]

Hi All
Any one help me to do a circuit which can generate 500pSmonoshot pulse with input PRF 1mS to 1uS
I need from the generated pulse to drive an RF FET.So please consider the min Vgs too.

 

[mister_rf]

Read some ideas here:
https://www.edaboard.com/threads/102667/
**broken link removed**

Also see here:
http://wcsp.eng.usf.edu/UWB_Testbed_Docs/low_cost.pdf

---------- Post added at 09:52 ---------- Previous post was at 09:47 ----------

Also some idea here:
**broken link removed**

 

[rineesh]

@mister_rf
From http://wcsp.eng.usf.edu/UWB_Testbed_Docs/low_cost.pdf
This is avery good approach but LO shows that 10MHZ.But my LO is TTL with frq range1kHz01Mhz
can I keep the out put with 500ps with PRF with respect to input?

 

[mister_rf]

The transmitter employs SRD to form a step function necessary for the generation of the pulse using charging and discharging circuits based on a Schottky diodes, and the 10MHz generator are only to obtain pulse repetition rate of 10 MHz. This permits also to obtain pulses on a lower rate, so there’s no problem using a 1KHz – 1MHz generator.

 

[rineesh]

@mister_rf

Regarding with MMIC amplifier current one is SNA686.The output current <100mA.
I need 2A current.So Can I use Ant RF FET's (found one part PD20015-E 2Ghz opration ) as output driver?
Can I vary the output pulse width? how it is acheived?

 

[FvM]

I need from the generated pulse to drive an RF FET.So please consider the min Vgs too.

I think it's your job to give a clear specification. It should also include a load impedance.
Generally, 500 ps is pretty within the reach of fast non-saturated Si logic devices, e.g. ECL, so you don't necessarily need "exotic" parts like SRD, which have become rare and expensive these days. For high power output pulses, avalanche transistor circuits would be an option, too.

 

[rineesh]

@FvM

I need from the generated pulse to drive an RF FET.

This is the Impedence datafrom datasheet.

need "exotic" parts like SRD

I am in that path please read the first replay from mister_rf

 

[FvM]

You need to get an impedance specification applicable to wideband respectively time domain. The 2GHz input impedance already reveals a considerable input capacitance. It may be reasonable to start with the estimation of a capacitive input impedance respectively a RC series circuit.

Generally, these high MOSFET input capacitances can be easily driven in impedance matched resonant circuits, but are unhandy in wideband apllications. Peak currents are in an A order of magnitude for the intended pulse width.

 

[rineesh]

@mister_RF
Step recovery Diode and PIN diode

For impulse generation Can I use PIN diodes instead of SRD???????????????

 

[FvM]

Can I use PIN diodes instead of SRD???????????????

There's a recent thread discussing this topic. https://www.edaboard.com/threads/217841/

Although Agilent has mentioned a possible replacement in an application note (dedicated to frequency multipliers), it's not yet clear if a suitable performance can be achieved in pulse generator applications. Referring to your application, 500 ps is pretty slow for SRD. But the SRD on-state current must have at least the same magnitude as the intended output pulse current (respectively considerably higher, depending on the pulse former circuit ). This doesn't sound feasible to drive a power RF FET.