Source follower with bip. Trans and bootstrapping for impedance increase

Using a phemt and bootstrapping R4 might do even better.

Phil Hobbs likes to power the drain with an NPN emitter follower to do the bootstrap function.

A fast buffer, like BUF602 would be a neat drain bootstrap driver; gain would be close to 1. I suppose there's no reason why bootstrap gain can't be more than 1.

--

John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Perhaps bootstrap R4 to the output?

--
This email has been checked for viruses by Avast antivirus software. 
https://www.avast.com/antivirus

Why is R4 even there? If used, should be 100M, or better 1G, and bootstrapped. The bootstrap can be arranged so the DC Z-in appears to be much higher, 100G would be good.

High overall circuit slew rate capability is important. Replace J2 with a low-capacitance BJT current sink and run Q1 at much higher currents.

Fast Hi-Z HV probes are a useful area to explore. There's no good reason the probe can't work to at least 600 volts, but 1.5kV would be better. Then it could be used to probe critical points in my 1.2kV amplifier circuits. But only if it's also quite fast, needs to slew at least 2 kV/us.

--
 Thanks, 
    - Win

Am 20.04.2019 um 09:13 schrieb Robert Baer:

Bootstrap R4 to the output works only for low frequencies.

Am 20.04.2019 um 13:14 schrieb Winfield Hill:

One resistance (R4) is better than 2 (R5,R3) respective to the parasitic capacity.

The resistors have capacitance, which can be bootstrapped out.

It need a PCB guard too.

One question is, what use is a fF amp?

--

John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Am 20.04.2019 um 16:50 schrieb John Larkin:

I use it for the first stage of active antennas. The theory of active antennas ask for it due to the capacitive voltage divider of antenna capacity and FET input capacity.

Then you probably don't need a high DC input impedance, fast slewing, and kV capability.

--
 Thanks, 
    - Win

Am 20.04.2019 um 17:44 schrieb Winfield Hill:

I need a high impedance over a wide frequency band, maybe 100 MHz. JFETs and GaAs-FETs are the best component for the first stage of active antenna, but both, also MOSFETs, have a low input impedance at higher frequencies, not only because the input capacitance.

So I am trying to solve.

This fF-amplifier gives a few little advantages over normal JFETs. At

100 MHz the real part is more than 2 kOhm insted a few Ohms and the input capacity is very low. Problem is the high equivalent input noise voltage - I am working at it.

Regards

To reduce JFET voltage noise, run them at higher currents. You may also benefit from a higher- capacitance JFET, not necessarily characterized for RF use. Check the tables in AoE III. Your bootstrapping schemes can give you breathing room to use a higher-capacitance part, and run it at quite high currents, like 20 to 50mA.

--
 Thanks, 
    - Win

Am 20.04.2019 um 18:23 schrieb Winfield Hill:

thanks - yes - I have good equivalent input noise voltages with

2SK2394CP6 and BF862. Also both I use for the output stage, because of their high forward transfer admittance that allows to drive 50 Ohm.

AoE III. I have here, I will look into again ...

Regards

Is the antanna basically a small e-field probe?

Gate current noise is another thing to worry about. If you use a jfet, keep the drain voltage low, well below 5 volts.

--

John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Am 20.04.2019 um 18:32 schrieb John Larkin:

Yes, active antennas are small e-field probes. So much the better the equivalent input voltage noise and input capacitance and higher real part of input impedance so much smaller is the passive antenna geometry.

hmm, another problem at active antennas is the large signal persistancy, that asks for a drain voltage of 12 to 24 V.

thanks

Regards

Am 20.04.2019 um 18:32 schrieb John Larkin:

With a 2SK2394CP6 in the first stage of an active antenna, I get equivalent input noise voltages of 1.7 nV/sqrt(Hz), that is very good.

Regards

BF862 is good, high transfer admittance corresponds to low voltage noise. Also ON Semi's CPH3910, replacement for BF862.

Good!

--
 Thanks, 
    - Win

John is right about the Vds limit, necessary to avoid impact-ionization gate current. You can solve input- voltage compliance issues with proper bootstrapping.

--
 Thanks, 
    - Win

The pHEMT/SiGe:C NPN cascode gets down to about 0.4 nV/sqrt(Hz) in that band. See e.g. this plot from my old website: .

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC / Hobbs ElectroOptics 
Optics, Electro-optics, Photonics, Analog Electronics 
Briarcliff Manor NY 10510 

http://electrooptical.net 
http://hobbs-eo.com

Am 20.04.2019 um 21:16 schrieb Phil Hobbs:

thanks

That is very interesting for active antennas from 100 MHz to 2 or 3 GHz.

My calculator says 131 pF for your numbers. (131E-12)