HFU HF Underground
Technical Topics => Equipment => Topic started by: IQ_imbalance on February 19, 2018, 2046 UTC
-
Hi folks. i'm putting together a notch filter to deal with the two BCB hotties within about 10km of me. I think i've done the math right and can figure out the necessary components. Since there are two frequencies I need to notch, i'm going to set up a series tuned trap for one and a parallel tuned trap for the other. I was thinking i would put these both in my extremely fancy project box (aka mint tin); are there any interactions or pitfalls i need to worry about? The fine tuning will be done with a shielded tunable inductor (one on each trap) instead of a variable capacitor mainly because i couldn't find a small variable capacitor in the range i needed....
Thanks!
R
-
Be sure the tuned networks are shielded from each other to prevent detuning. I would either use two 365pF AM radio plastic tuning caps for tuning, or use trimmers with parallel caps to get the value you need.
I would do two series tuned networks in parallel with some resistance between them to prevent detuning. I would need to muck around with it to get the response I need. A parallel tuned trap is going to act like a very narrow bandpass filter. If that's what you want.... I assume you want to pass everything but those two MW channels?
+-RH
-
No, trying to notch out 700kHz and 800kHz in order to deal with the high power BCB stations near my house that are overloading my receiver. If possible i'd like to target those two frequencies as they bleed through the high pass filter I already have.
Perhaps I could either 1) make two filters and separate w/coax (which would give the resistance you pointed out), or 2) put a metal shield between the parallel trap and the series trap and connect them with coax (or maybe a 50 ohm resistor if i can brainstorm a way to do that inside the can).
-
No, trying to notch out 700kHz and 800kHz in order to deal with the high power BCB stations near my house that are overloading my receiver. If possible i'd like to target those two frequencies as they bleed through the high pass filter I already have.
Perhaps I could either 1) make two filters and separate w/coax (which would give the resistance you pointed out), or 2) put a metal shield between the parallel trap and the series trap and connect them with coax (or maybe a 50 ohm resistor if i can brainstorm a way to do that inside the can).
Build two identical tuned traps in different boxes to maximize isolation, each tuned to a different frequency of reject. You can probably tune them by ear with a receiver and tune for the deepest notch. When you couple the filters together, you may need to tweek the tuning of each filter slightly to account for coupling.
More info here.
http://www.qsl.net/kp4md/mwfilter.htm (http://www.qsl.net/kp4md/mwfilter.htm)
https://www.dxing.com/tnotes/tnote06.pdf (https://www.dxing.com/tnotes/tnote06.pdf)
+-RH
-
Well my first attempts at building notch filters left much to be desired. Tried a series tuned wave trap w/ a tunable coil and a parallel tuned wave trap specifically for 820 kHz. Didn't see any attenuation, either alone or in combination w/ my BCB high-pass filter. Could be the cheap Amazon components (fixed value inductors & capacitors), could be my questionable wiring skills, could be the antenna is boosting the broadband signal coming in to a point where i can't notch it out. Oh well... I don't think i'll quit my day job just yet.
-
I too would rather build something, but I knew buying a well rated little box would be so much faster and considering my time way cheaper. Just get a Par BCST-HPF. It destroys everything under 1800. Toggle switch too.
-
Just for peoples' amusement. On the left, the parallel tuned wave trap for 820kHz. On the right, the series tuned wave trap w/ an adjustable coil. Neither did much, if anything ::)
-
Everything in RF needs tweaking. It's rare for something new to work flawlessly the first time you turn it on without optimization. This isn't house wiring where it either works or it doesn't.
Understand that there will always be a certain amount of (unavoidable) parasitic capacitance, inductance and resistance that can't be predicted easily. Where I am going with this is that they may be working, just not at the frequencies you intended. So when you follow a pre-printed formula for capacitor and inductor values for your ~800 KHz filters, you may not actually get what you had intended due to the formulas not taking into consideration parasitics and that may very well explain what you are experiencing. The circuit in a tin box has some long leads on (for example) the capacitor. Shorten those up. Those leads are represent parasitics. I see room for optimization in that box. The frequency of resonance may change when you put a cover on that box.
Since you more than likely do not have a vector network analyzer at your ready disposal :D, I suggest that one place to start to determine if they are working at all is by tuning around the whole MW band and see if you can notice a drop in the noise floor or drop in any other signals at any other frequency. If you have an SDR, use the display in wide band mode to try to gauge this.
For something like this, it's probably better to build with variable components, if possible, just because it will need tweaking.
-
Needs to be in a screened box
Signals can also propagate via the coax braid.
You need to use a VNA as suggested for any work with filters, at worst a sig gen and a scope.
Str.