200 MHZ high impedance JFET buffer

[alexm90]

Hello all,
I would design an voltage buffer with an input impedance of 1Mohm (and less than 10pf) with a bandwidth of 200MHz.
The dynamic of the input signal should be 10Vpp
Do you have some tips?
I'm thinking to use an JFET source follower to have an big input impedance, and polarize it with an current mirror. But the dynamics of the signal is very large

thanks
cheers

 

[schmitt trigger]

Please calculate what is the impedance of a 10pf cap at 200 Mhz, so you realize how preposterous the requirements are.

 

[alexm90]

yess, less than 80 ohm
i mean, an DC input impedance of 1 megaohm.. is obvius if there is an capacitance in prallel it is lower at higher frecuencies
i would use this buffer with an passibe oscilloscope probe

 

[FvM]

The bad thing is that JFETs have also a considerable real input conductance at higher frequencies. Need to check if it's sufficient for your application.

 

[dick_freebird]

Two things. One, a sane sized JFET for an IC design is
going to be sub-PF gate capacitance unless you're
talking an RF PA final stage, or something. 10pF is a
very high number, probably assigned without much
thought.

Two, your JFET will only present a high DC impedance
when the gate is reverse biased. A high slew rate signal
which the source cannot follow will let the gate conduct
forward on one of the transitions. 10V P-P certainly holds
that potential. A closed loop amplifier will be slow and
tend to allow transient forward bias on fast high amplitude
excursions. An open loop amplifier is liable to have poor
linearity over such a large range.

There are designed FET probes for 'scopes that might
be good study material, what they achieve and how.

Now I do not see any value in creating a 1Mohm /10pF
Zin FET buffer, to drive a 10Mohm / 10pF Zin passive
'scope probe. That's like wearing two pairs of shoes.

 

[alexm90]

i do no not have to do an ic design.
maybe i did not expressed well, i'm sorry
i will explain my project more in the detail.
i would to build an "active probe", i bought an 200mhz x10 passive probe and now i would design (using discrete components) something similar to the front end amplifier of an oscilloscope. for this reason i fixed the maximum input capacitance at 10pf, and the DC impedance at 1Mohm.