B
bo
All,
I am designing a PC card type oscilloscope. On a design inherited, the
attenuation network of 3 RC parallel combos in series with a 500K/43pF
parallel circuit:
A --500---- 453K---45.3K---4.53K---gnd
43pf 47pf---470pf 4700pF-gnd
B C D E
with each of the above sections paralleled.
Here's my question/problem:
When a 10x or 1x probe is connected at A, the resultant waveform
captured by tapping at any of points C,D or E, is UGLY. However, with
a coax input tied to point A, the waveform is pretty. The idea of the
circuit is divide by 2, 20 or 200. A 2nd problem, is that my freq
response is pitiful ie <10MHz.
Any ideas on how I can make this circuit immune /tolerate the probe
capacitance? and increase the 3dB freq response?
Regards,
Paul
I am designing a PC card type oscilloscope. On a design inherited, the
attenuation network of 3 RC parallel combos in series with a 500K/43pF
parallel circuit:
A --500---- 453K---45.3K---4.53K---gnd
43pf 47pf---470pf 4700pF-gnd
B C D E
with each of the above sections paralleled.
Here's my question/problem:
When a 10x or 1x probe is connected at A, the resultant waveform
captured by tapping at any of points C,D or E, is UGLY. However, with
a coax input tied to point A, the waveform is pretty. The idea of the
circuit is divide by 2, 20 or 200. A 2nd problem, is that my freq
response is pitiful ie <10MHz.
Any ideas on how I can make this circuit immune /tolerate the probe
capacitance? and increase the 3dB freq response?
Regards,
Paul
