Compensated multiple output attenuator

Help Support GroupDIY:

Dualflip

Electronic monk
Joined
Nov 15, 2008
Messages
1,487
Location
México City, MX
I read a book that mentions about the need to add capacitors in parallel with resistors in a resistive divider made up with many resistors in series and taps coming out of each resistor, in order to compensate for the PCB's parasitic capacitances, all the capacitors should have the same ratios as the resistors and the last capacitor should be variable and adjusted for best square wave symmetry, pretty much the same way as you would compensate a 10X probe on an oscilloscope.

How many of you have used this kind of arrangement?

If I remember correctly, the resistor values used in the example were relatively high, the highest resistor was in the Mega Ohm range, perhaps using much smaller resistors would negate the need for capacitors?
 

Newmarket

Well-known member
Joined
Oct 10, 2016
Messages
1,065
Location
Brighton Sussex UK
What sort of application was the book considering ?
High resistor values sounds like it might be high voltages as in kVs ?
The idea is that the capacitor values 'swamp' the parasitic capacitances such that the ratiometric differences in capacitance across the resistors are reduced.
 

PRR

Well-known member
Joined
Jan 30, 2010
Messages
11,144
Location
Maine USA
> perhaps using much smaller resistors

Say there's 100pFd anywhere you look.

With 100r the top-cut is 16MHz.

With 100k the top-cut is 16KHz.

With 100Meg the top-cut is 16Hz.
 

Newmarket

Well-known member
Joined
Oct 10, 2016
Messages
1,065
Location
Brighton Sussex UK
PRR said:
> perhaps using much smaller resistors

Say there's 100pFd anywhere you look.

With 100r the top-cut is 16MHz.

With 100k the top-cut is 16KHz.

With 100Meg the top-cut is 16Hz.

I get the maths (yes it's "maths" not "math"  :) ) but I was thinking that the 'stray' capacitances in question were across the resistive element rather than to 'Ground' / 0V reference ?
 

PRR

Well-known member
Joined
Jan 30, 2010
Messages
11,144
Location
Maine USA
> I was thinking that the 'stray' capacitances in question were across the resistive element rather than to 'Ground' / 0V reference ?

"Say there's 100pFd anywhere you look."

How would the electrons know "across" from "to ground"? They are not that smart.
 

moamps

Well-known member
Joined
Sep 24, 2005
Messages
2,165
Location
Croatia, HR
Dualflip said:
How many of you have used this kind of arrangement?

I used frequency compensation on the attenuator/divider network when I was very young and was building my first broadband AC voltmeter. Generally, this approach does not have to be used when the divider is low-resistance, it does not require very high precision and a large frequency range, which is mostly the case in audio gears.

https://wiki.analog.com/university/courses/alm1k/circuits1/alm-cir-voltage-divider



 

Attachments

  • tube voltmeter.png
    tube voltmeter.png
    163.7 KB · Views: 10

Latest posts

Top