Twin-T Filter On-line Calculator:
http://sim.okawa-denshi.jp/en/TwinTCRtool.php

Only select BODE Diagram (frequency response)...one or more of the other choices blows up.
Enter 1 pF as 10e-12 F

Observe the VERY BROAD null bandwidth (nearly the entire UHF Band).....

You can quickly see the HUGE impact if your resistors are off by say 1% or if stray capacitance changes
those 1 pF capacitors into (much more likely ) 3 pF capacitors [YIKES!!! NOT EVEN CLOSE!!!]

the frequency response of that calculator is hard to read, and the zero values don't seem to reflect the frequency chart when I change values.


is there a better frequency chart so I can more easily see when frequencies would be effected?
or is that circuit just too broadband?

what values would be best to knock out ch 33?

you can give this a whirl too...
I'm pretty sure if ya follow Tools, in the menu
you can play around with protoype filter synthesis, etc.
its for Windows.
RFSim99 (http://electroschematics.com/835/rfsim99-download/)

This one is similar, but for linux, and is found on sourceforge.
Qucs (http://qucs.sourceforge.net/)

On a somewhat related note, the High Pass Filter design I was playing with in Posts 1176, 1177 does work.
http://www.digitalhome.ca/forum/showthread.php?p=1201087#post1201087

Just, I don't think it's useful for anything here. As trash from occasional
electrical noise in the neighborhood will still pass thru it in the passband.
A diplexer connected as a high pass filter (UVSJ) will do the same thing.

It did knock down the FM band pretty well though. Think I measured like ~30-35 dB. Could be built better no question, and probably has some excess insertion loss but wasn't bad considering it was just a quick hack in the basement (never did find the caps I really wanted). I haven't gone thru the hassle of installing it in a shielded bud box yet either. Still was fun playin around with it though.

3) how many adjacent channels does a stub filter severely affect?

When I was a teenager, we used to make a stub from a length of 300 ohm twin lead, that would attach to the rear of ur tv set's 300 ohm terminals (of course ya won't find that anymore more on today's sets). But the trick was to wrap like a 1/2" wide strip of aluminum foil around it, and slide that up & down till you found a null. Worked very well in VHF HI.

thanks for the info.

if making a coax stub, can I use a splitter or do I need a real T?
I've already made one using a splitter but I didn't see much affect aside from the 3.5db attenuation. I have a real F type T ordered and am wondering if I will see a difference?

Should work with a real T (Zo=75 ohm). But I think will be trickier to adjust the stub with coaxial cable. i.e. - can't do the alum foil trick with coax, so make it longer than ya calculate initially, and keep trimming it little by little between tests, I guess. BTW, how are you going to measure it? And out of curiousity, why are ya trying to notch ch 33? is it an analog or digital ch 33?

Be aware of the velocity factor for the type of cable your using in the calculations. Might vary between cable manufacturers. Ya should have varying results depending exactly where ya put it in relation to the source (the antenna). I think I would try to shoot for placing it a 1/4 wavelength (at ch 33) down the line from the antenna for best results. N multiples of a 1/4 wavelength should be optimal too if placing it inside the house is more convenient.

my problem is 33 real is a local LOS and 30 real and 36 real are 2nd edge

I'm trying to knock down 33 so I can then add more amplification to keep the other two locked.

there is no 33 jointenna at warrens, so I am looking at either getting the stub to work, or fork out the cash for a custom ch33 notch filter.

http://www.tvfool.com/?option=com_wrapper&Itemid=29&q=id%3d3cf49abd5adfa0

A 1/4-wave stub filter is NEVER going to provide the selectivity of a multi-pole notch filter,
which would be inserting only a small amount of attenuation 3 channels away:
http://www.tinlee.com/Graph_CR7E.php?active=1
http://www.tinlee.com/Graph_CE7_Spec_Order.php?active=1

hmm, what's the rest of ur system's details? what antenna? using an amp now? if so where and what model is it? downlead length? splitter?
Off the top, I just don't think that's too strong a signal to warrant notching it out. my strongest LOS is like -24 dBm, and my 2 desired "1 edge" are only like +3 NM. And I dont seem to have any dynamic range issues with them, amping here.

So with ur 2 desired 2 Edge at such a stronger signal than mine here, the reason for being 2 edge must be the surrounding terrain? Just guessing, it might be hilly around there, am I right? If terrain is a little rough, is it possible your seeing some multipath? Do you have an antenna rotor?

2 91xg vertical stack, to combiner to uvsj with Y10-7-13 to Kitztech model 200 at mast. then 40-50 feet of quad rg-6 to grounding block to 5 feet indoor 3 way splitter. each leg is approx 40 feet.

TVFOOL is way off for my specific area, I can't get any 2nd edge channel without an amp. all those stations are at the same heading, yet where 50 (6.1 FOX) comes in the best, I loose 2 bars on 36(68.1 Myn) so I am assuming that 50 is far enough away from 33 that I can point right at it, but 36 and 30 are too close in frequency and the tuner is saturated. So I have to aim off from the optimum angle to try and keep 33 in a null, but it isn't effective enough.

A 1/4-wave stub filter is NEVER going to provide the selectivity of a multi-pole notch filter,
which would be inserting only a small amount of attenuation 3 channels away:
http://www.tinlee.com/Graph_CR7E.php?active=1
http://www.tinlee.com/Graph_CE7_Spec_Order.php?active=1
which model would you recomend?

How is reception of the 2 problem channels when connected to 1 TV with the splitter removed? Is it the same or better?

agree with hoopitup's test suggestion. I would try and test each tuner that way also.
If none of them get any better, then no amount of further amplification is gonna improve 'em.

before going to any more extremes, i'd really wanna prove to myself that there's a "signal strength" issue limiting my desired edge channels and not a "signal quality" issue. The Kitztech Amp, do they include an FM Trap?
The 91-XG stack, are they facing the same direction, or are you aiming them
at those two different transmitter sites?
Sometimes tilting ur antenna array up or down a few degrees, will improve things too.

yes, vertically stacked, pointed same, screens almost touching.

hooked to a tuner at the end of a 40' down-lead, ch 30 only shows up during tropo.
but it may be below tuner threshold.


I moved my antennas, from 25' to 6' and I got stronger signal. who knew?
I'm going to move it around some more in case there is a magic spot for 30.

It's all about the sweet spot when dealing with difficult UHF stations. As few as 3-4 inches can make a world of difference. You can go with much less than 3 feet of vertical separation between the VHF-HI if needed to find the desired UHF sweet spot from my experience.

vertically stacked, pointed same, screens almost touching.
I don't stack antennas here but screens almost touching sounds too close to me.

I think u should play with the vertical seperation between the two, should probably be at least a 1/2 wavelength apart (at the lowest freq of operation, i.e. ~ 24" at ch 14 = 1 W.L. should be even better) so they don't adversely affect each other.
Also make sure the jumpers between the two antennas and combiner are identical in length.
Check out the OTA FAQ, specifically Post # 16 (http://www.digitalhome.ca/forum/showpost.php?p=675713&postcount=16)where it talks about stacking antennas. Lot's of good info in there. There is also an entire Stacking, Ganging, Combining TV Antennas (http://www.digitalhome.ca/forum/showthread.php?t=95148) thread. Plenty good info in there too.

I just saw an interesting piece, two 300 Ohm to one 75 ohm, better than 2 baluns to splitter/combiner to 75 ohm do you think, for ganging 2 antennas?

In my potential application would need about 6" of 300 ohm each side to hook it up.

where 50 (6.1 FOX) comes in the best, I loose 2 bars on 36

The other factor to consider is the XG91 is optimized for somewhere around channel 68 and the gain gradually drops for lower channels, so it will provide more gain for channel 50 than it will for 36 or 30.

the booms are at least 24", screens extend up and down at bit.
maybe 1" between screens.
After I moved the antenna from 25' to 6', I gained more signal. Now I just need to walk the lawn and find a sweet spot.


I've thought about using 300 ohm line to joint to 2 91xg, and use a half wave balun in the middle of that line. If I can separate the fan dipoles from their internal balun connections I might try it this summer.

ok, back to filter and signal gear talk.

so after ordering a ch 19 jointenna and awaiting shipment, I remember I had ordered some notch filters for another experiment, so I decided to see what I could do.

my channel 19 antenna is some commercial brand GH. the main array is 2 91xg and a y10-7-13.

filters I had on hand were lpf 550, a notch 14-19, a notch 24-29 and a notch 19-23 as well as a UVSJ.

so from the GH, I go to a kitz 200 then out to a UVSJ with vhf terminated to the notch 24-29 to the lpf 550 then finally to the splitter.

from the array I go to a kitz 200 then to the notch 14-19 then to the splitter.

IT WORKS!!!

anyone know what kind of loss the jointenna has? I bet my solution has at least 4-5db insertion loss including splitter, maybe more.