Author Topic: A strange compensation network on pre-amp input  (Read 284 times)

0 Members and 1 Guest are viewing this topic.

Offline igor78Topic starter

  • Contributor
  • Posts: 24
  • Country: us
A strange compensation network on pre-amp input
« on: January 15, 2025, 12:16:20 am »
I'm working on freq counter pre-amp. Freq range is HF and maybe VHF.
I want it to be like an active probe with JFET input. The idea is straightforward - the signal goes to JFET as a source follower and then immediately into an amplifier.

Clearly this has been implemented many times over, so I'm studying prior art. I found a few similar schematics where the signal is fed through some sort of compensation network? I've highlighted it on the images.



I'm guessing this is the frequency compensation to improve linearity? Does not seem to be relevant here.
Or is it meant to form capacitive divider and matching resistive divider with JFET gate capacitance to source? Attenuation? Why do you need attenuation here? Component values do not match in the top picture then. 10pF/gate-capacitance is not like 39k/1M Ω

DC blocking capacitor on the input 470nF in one case and 100nF in the second case would have SRF somewhere within 0-50MHz band for sure, are they compensating for the SRF of the input cap?

I often see caps thrown at input. Are they operating them as low value inductors on the other side of SRF? Getting 100nF cap that work to VHF will be challenging.

Two level limiting diodes would add to input capacitance, why it's not a concern?
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 15231
  • Country: de
Re: A strange compensation network on pre-amp input
« Reply #1 on: January 15, 2025, 09:07:25 am »
The resistor R1 is there for protection. The parallel capacitor helps with the gain at high frequencies, kind of compensating the 39K/1M divider at least somewhat. With a frequency counter the response is far less critical than with a scope, but it still helps to have an about flat response and one would definitely not want an large extra drop to high frequencies. The 10 pF may be on the low side, but one may not want too much to still have ESD protection.

The input capacitor SRF should not be that bad. At the high frequency the capacitor starts with a very low impedance and the Q is limited with X7R or similar capacitors. So even in the resonance the impedance of the capacitor can still be low compared to the amplifier input. Having a lossy capacitor here helps.
 

Offline PGPG

  • Frequent Contributor
  • **
  • Posts: 471
  • Country: pl
Re: A strange compensation network on pre-amp input
« Reply #2 on: January 15, 2025, 11:52:00 am »
The parallel capacitor helps with the gain at high frequencies, kind of compensating the 39K/1M divider at least somewhat.

Rather compensating 39k/(2x 1N4148 capacitance) divider.
 

Offline PGPG

  • Frequent Contributor
  • **
  • Posts: 471
  • Country: pl
Re: A strange compensation network on pre-amp input
« Reply #3 on: January 15, 2025, 11:57:38 am »
I often see caps thrown at input. Are they operating them as low value inductors on the other side of SRF?

If you need DC blocking you need capacitor there.
It is not important if it works at its C slope ch-c part or L slope part. Important is that it isolates DC and has enough smaller impedance than following circuits.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf