Here is a pic of Rod Elliot's P52 notch filter I've already built. (works !)
http://i5.photobucket.com/albums/y177/M ... CN2753.jpg
Anyways......now I want to build another notch filter, but with only one fixed frequency....@....1 kHz.
I built a prototype that has a fixed notch around 1.034 kHz.
R1=R2 (16k)
R3 (8k)
C1=C2 (10nf)
C3 (20nf)
Some hurdles have to be overcome...
It must still be a slightly trim-able notch, because not all uber-low distortion test oscillators have an analog "fine trim" control....my Amber, and my Boonton, are digitally fine-tuned...about 5 Hz or 10 hz minimum per pushbutton step @ 1kHz......no in between steps.
How to trim this notch ? My problem (if you can call it that) is I have many near perfectly matched 16k Shallco wirewound resistors....most are 0.01% or better (16.000k) Paralleling two of them gives me the 8k value....this also keep them temperature stable, in relation to each other......so I'd like to keep the resistors fixed...if at all possible.
But matching the caps is more difficult. Trim caps usually trim only up to 50pf or so, so I need to match the caps within 50pf before paralleling any trim cap.....but how do I measure caps perfectly (polystyrene)? Solid-state cap meters have fairly low resolution.....An accurate cap bridge circuit is difficult to find (make one ? http://www.allaboutcircuits.com/vol_2/chpt_12/5.html )
It would be nice to trim this circuit with a large triple-deck air trim capacitor.
http://www.modernradiolabs.com/430 CAP test.jpg
If I remember correctly when using Rod's project 52, once R1 is trimmed to equal R2 (2k + 200 ohm trim pots), that left the majority of finding the notch freq...to the varaition/trim of R3 (5k, 500 ohm trim pots).
I wonder if the same is true for C3 if it's made variable?
Is there such a thing as active capacitance matching? or resistance ?
How does the HP 339 automatically find a notch ?
=FB=
http://i5.photobucket.com/albums/y177/M ... CN2753.jpg
Anyways......now I want to build another notch filter, but with only one fixed frequency....@....1 kHz.
I built a prototype that has a fixed notch around 1.034 kHz.
R1=R2 (16k)
R3 (8k)
C1=C2 (10nf)
C3 (20nf)
Some hurdles have to be overcome...
It must still be a slightly trim-able notch, because not all uber-low distortion test oscillators have an analog "fine trim" control....my Amber, and my Boonton, are digitally fine-tuned...about 5 Hz or 10 hz minimum per pushbutton step @ 1kHz......no in between steps.
How to trim this notch ? My problem (if you can call it that) is I have many near perfectly matched 16k Shallco wirewound resistors....most are 0.01% or better (16.000k) Paralleling two of them gives me the 8k value....this also keep them temperature stable, in relation to each other......so I'd like to keep the resistors fixed...if at all possible.
But matching the caps is more difficult. Trim caps usually trim only up to 50pf or so, so I need to match the caps within 50pf before paralleling any trim cap.....but how do I measure caps perfectly (polystyrene)? Solid-state cap meters have fairly low resolution.....An accurate cap bridge circuit is difficult to find (make one ? http://www.allaboutcircuits.com/vol_2/chpt_12/5.html )
It would be nice to trim this circuit with a large triple-deck air trim capacitor.
http://www.modernradiolabs.com/430 CAP test.jpg
If I remember correctly when using Rod's project 52, once R1 is trimmed to equal R2 (2k + 200 ohm trim pots), that left the majority of finding the notch freq...to the varaition/trim of R3 (5k, 500 ohm trim pots).
I wonder if the same is true for C3 if it's made variable?
Is there such a thing as active capacitance matching? or resistance ?
How does the HP 339 automatically find a notch ?
=FB=
...Because my Boonton oscillator will not fine-tune exactly to 1034 Hz..... only 1030, or 1040 (10 Hz steps).
I guess it's kind of a goal for me, an experiment...
But I realized my Amber (oscillator unit only) does tune in 1Hz steps. So my old notch filter is able to tune a 4 volt notch at 1110 Hz to read 2.9 mV rms at the output.....And at 1110 Hz, my Boonton reads 0.8 mV.
The reason for building a new filter is, I came accross some better parts...polystyrene caps, and some AD797 opamps.
It is my experience that the ability to fine tune the oscillator is just as important as tuning the notch filter, to get the deepest notch.
But I've been fighting what I believe is an un-winnable battle, trying to build this circuit with only one resistance pot on the front panel.
I was hoping to use small internal trim pots, but...
Temperature, humidity, (and other factors?) seem to necessitate the use of (at least) 2 multi-turn pots available on the front panel ....and maybe a cap trim if I an find a suitable one.
Last night I built the 1kHz filter section two times, but with poor results compared to my older working unit with several panel pots in which all of them were used to get below 3 mV on my analog AC meter.
(um.....I think my Amber is supposed to have lower dist than my Boonton.....but my results show different ?????)
=RR=
I guess it's kind of a goal for me, an experiment...
But I realized my Amber (oscillator unit only) does tune in 1Hz steps. So my old notch filter is able to tune a 4 volt notch at 1110 Hz to read 2.9 mV rms at the output.....And at 1110 Hz, my Boonton reads 0.8 mV.
The reason for building a new filter is, I came accross some better parts...polystyrene caps, and some AD797 opamps.
It is my experience that the ability to fine tune the oscillator is just as important as tuning the notch filter, to get the deepest notch.
But I've been fighting what I believe is an un-winnable battle, trying to build this circuit with only one resistance pot on the front panel.
I was hoping to use small internal trim pots, but...
Temperature, humidity, (and other factors?) seem to necessitate the use of (at least) 2 multi-turn pots available on the front panel ....and maybe a cap trim if I an find a suitable one.
Last night I built the 1kHz filter section two times, but with poor results compared to my older working unit with several panel pots in which all of them were used to get below 3 mV on my analog AC meter.
(um.....I think my Amber is supposed to have lower dist than my Boonton.....but my results show different ?????)
=RR=
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.