DIY 20 Meter Bandpass Filter for Ham Radio 

I built a full set of these filters. The trick is that you need to grid dip each of the LC pairs together as they recommend in the article. Remember, that is a tuned circuit. Just be sure that when you grid dip them that you have the leads of each pair soldered together and the far legs on each LC pair cut to the length that they need to be when they are going to be soldered in place on the circuit board. If you tune them via the grid dip meter with the leads long and then you cut the leads you have just detuned the LC pair. When you solder them in the circuit the filters will typically be a little low in their resonate frequency. When you get to that point you can put a 50 ohm dummy load on one end of the filter and then something like a RigExpert antenna analyzer or a Nano VNA on the other end. Set the analyzer for the center resonate frequency that you are shooting for and then start with the middle inductor and carefully move the windings closer together to see if you can closer to 1:1 SWR at the center frequency. If you don't then try and spread out the windings. If you are adjusting an inductor, one inductor may not get you to the lowest SWR. You may get lower SWR and then it will start back up the other way so it typically takes adjusting two inductors back and forth to get the lowest overall SWR. Carefully work each of the inductors to try and achieve 1:1 SWR. If you adjust an inductor and it's not making much of a difference, leave that inductor alone, don't tune it any further. When you get to 1:1 at the center frequency then look at the filter across the band. If all is correct you should have a nice drop on or near the outside edges of the band. I highly recommend not taking the shortcut of not using a grid dip meter. This is a very active circuit when it comes to each of the LC pairs working in unison. If you build the 80m version it is very narrow banded and also the 10m version in my experience does not cover the entire band but it's nowhere near as narrow as the 80m filter.

Filter design and debugging (LC cavity microstrip waveguide bandpass, low-pass, high pass) duplexer design 10 years of experience in HFSS simulation design and engineering, with a design frequency of terahertz. All designs are based on practice and guided by theory. Remote after-sales service is available, and long-term cooperation can provide on-site service!

Thanks for this very nice and instrutive build.I just built the 20m and 40m version.And with good advice from this comment section i was succesfull tuning thoses perfectly.Ready in time for the 2023 WFD ! Thanks again all!

Use the largest magnet wire that will work on those cores to reduce losses and improve Q in order to get steeper band skirts. Remember that while you must provide some space between toroids, (and keep them at 90 degrees as you did,) that every bit of wire used in the filter creates losses and hurts the Q and band skirts. The ground wires should have been heavy gauge and gone straight down to the filter ground plane, not dragged out to the center of the board on either side. Let the ground-plane do the work. At RF it has much lower inductance and lower resistance than those ground wires. If you feel the need, lay the bare lead down on the ground plane and solder along its length to where you want it to 'go'. But remember that RF follows the surface of the conductor, and the surface of solder has a lot higher RF resistance than copper does, which is why I resist that temptation for RF work. The capacitor lead lengths were much too long, and totally unnecessary. Attach them only between the edge of the island pad and the nearest edge of the ground plane, so there should be essentially no lead length on the caps or inductors if it can be managed. If a length of lead from a device to ground is physically necessary then make sure the length is on the ground end not the hot end. Leads are bad. While short lead length is essential at VHF, it still a lot for good performance at HF. Always build EVERYTHING as if you are working above 500 MHz, and you will always be happy with the results. It's a habit well worth getting into. Finally, if you are off frequency with your bandpass, try spreading or squeezing the turns on the tuned circuit toroids and watch the result on the SA. Spread them to raise the freq, squeeze them to lower it. or spread the coupling toroid turns to play with the bandwidth and minimize the filter loss. I usually let it be a little wide if that helps efficiency. Ultimately, in such a hand-made filter the meter measurements are only to get you close. Proximity effects of even accurate devices shifts their inductance and capacitance values. BTW, polystyrene caps are NOT power caps. They are low leakage caps, for sample and hold and some audio applications, things like that. I would never use them in RF power applications. Polystyrene plastic is low very temperature and it melts. I don't know why anyone knowledgeable about RF would suggest them for an rf power circuit. Silver mica caps are more accurate in their values, (closer to spec,)and more stable over heat ranges. They are all I use in RF power applications at HF. Of course I'm lucky to have a couple of thousand of them in my parts stock, ( an amazing $20 box of parts from under a table at a ham swap,) so I don't have to drop $10 every time I need a couple of caps. Ultimately, ceramic trimmer caps are the easiest way to tune a filter and they have a better Q than the fixed caps. And yes, you should definitely have built it out of the box, and rough tuned it first. You can't get short, clean leads when you are reaching down into a box with a soldering iron and your fingers. All you can get is burned. Have fun.

Great build Ape and great explanation of bandpass filters in general.

I would replace the fixed capacitors by silvered plated trimmers and use shorter ground wires or tinned braids to the copper clad board. Also try to make the two toroids are not touching each other to minimise stray capacitance. The side skirts and insertion loss can be optimised by adjusting the three trimmers.

Try, Try, and Try again that’s the FUN of the Game👍🤔. I’ve been Hamming around since 1955 and I learn something new and interesting about our hobby every day and you always do an excellent with your videos. 73. KV5P

Thank you for making this video I'm looking forward to version 2.

This was a good watch.. Looking forward to the 2nd attempt!

Awesome Ape! Can't wait for the kits to hit the bare wire guys

Nice prototype build. Next phase are refinements as suggested by other commentators, especially Mr. West. When building these RF filters, I do it under a magnifying glass which brings one really "into the work". IN the next version, build outside of the box (natch) and compact the physical size of the layout, beginning with reducing the pad sizes, narrowing the widths of troughs between them. Use as little solder as possible when making connections. Connections off the board's ground plane to the connectors via copper strap as short as possible, or copper foil tape (soldering to the non-adhesive side). To reduce lead length to connectors, raise the board up standoff insulators and use flathead screws on the other side of the box that are countersunk with an oversized drill bit.

After my first field day and a recent trip with multiple operators, I realize this is a subject I’m going to have to tackle. Thanks for helping with this video. 73 de Bill K0WHW

Thanks for sharing this project. Even when they don't work as you had hoped you still removed some of the mystery.

Awesome very informative thank you ☀️

Gotta hand it to you for the effort though. I could say I would not have been able to do that just yet with my experience in radio. I'll get there, just not yet. If you do it over, and are successful, hope to see a video that shows the process from the successful build, maybe with clips from this video showing what happened, and why it didn't work out this time. Thanks for the video though. Not too many people would put something out that didn't work exactly as they wanted it to.

Would like to see you revisit this. Also would like see how you would set up a NanoVNA to test it.

This looks like fun!

Thanks for the video, I have a group of hams that we have been working with a local GMRS groups for fire watch and neighbor hood watch. Several of them have issues with interference on the GMRS frequencies, partly because most GMRS radios do not have good GMRS filters, because they are Ham radios that they converted to sell as GMRS radios ( but do not alter filters to be specific) and we live in an area that has some competing commercial users that I believe are boosting their commercial radios to overcome the terrain issues. However I think this will be a good group project for them to build and help them with reception. Thanks for the video and the information.

Creative video, thanks for sharing :)

Loved this one...I actually have two sets of these from the same QST article from 80-10M, kitted out in silicon muffin tins. I wonder if the toroids being placed at right angles to one another would improve the filter characterization. Martin Jue has told me that this is one of his design/layout practices. It would require resoldering but you might try it to see. Thanks again for your work, Ape!

I went back to the MarkiMicrowave site and played with the cutoff frequencies. Using 10 and 25 MHz as cutoffs, it looks pretty good. The values of LC components are a bit different. It appears it is quite sensitive to component choice. I’ll give this a try. I have 2 sets of TXBPF’s here for SO2R, but you’re right about POTA (and FD) activations - I need to give these a go.

Son .... my calipers, don't use no stink'n batteries! Very nice presentation Ape! 73 - KF6IF

Can't learn if we don't try. Looking forward to part 2. Would it be possible to build out the board on the bench, then drop it in the box and only have to sold er the 2 wires to the bulkhead connector?

Sweet!

Making these now for our club!! Thanks for the video! Did you revisit the project and get it to work correctly?

Great video. I have built 2 x Low Pass filters. Works great. But my attempt at a 2 x ferrite ring failed. George ps video soon.

Close, good vid.

What A Wonderfdul Video, Can You make A Variable Bandpass Filter For 20meter Band ?

Hi Ape. I watched this again and I am now hearing that band pass filters can work differently depending on the direction of the filter. So, should the cap be towards that antenna or towards the transceiver? I get it is not supposed to matter but Mike (K8MRD) found that it did. Also, another question, where is the recommended spot in the "chain" of Antenna, Amp, SWR/Power meter, and transceiver. In that chain, where is the best place for the band pass filter?

Way over my head but thank you…👍🏻🇺🇸🇺🇸🇺🇸

Interesting video, I have the same experience that these filters are difficult to get right at the first time. One problem is the low values of the inductors. 280 nH is very low to get right. One trick is to increase the impedance of the filter like 800 ohms and use the toriods as transformers. Like 1 : 4 in that case the inductors values increase and capacitors decrease, your capacitors are 500V i think even with 100 Watt (283 Volt in 800 ohm) it should be ok. (in the book "building a transceiver" of EI9GQ this is explained). Also in Elsie you have to set the impedance for analysis It is in the video 0.01 ohm (a bit low). Did you fix it in the end?

Great Video. You could be a Electronics teacher. I was wondering if i can show this at my club meeting tonight? Thanks.

Get rid of those long ground wires - they are adding unwanted inductance. Copy what the ARRL article shows and use braided copper as short as possible from the connector down to the board egde.

Hi Ape, Nice attempt. Probably try outside of the box and see the results. At least it will tell if the box had an effect. Stay safe. 73 WJ3U

@SmokinApe this is off topic but you mentioned you would like a smaller gmrs radio. Notarubicon just reviewed a new 50 midland, the mic runs everything and the base mounts under wherever. Just thought I'd bring it to your attention. Cheers VA6GKT

I would lay the mid- inductor down flat . To reduce magnetic interference

Is it possible to connect multiple filters to have multiband performance (I.E a 40M and 15M Filter)?

One problem I found with this construction method is. Especially if on VHF freq. Not to use double Sided copper board as it acts as capacitors top to bottom (thickness) also bolting it or gluing down in a metal box also capacitance problems. Copper board needs spacing away from metal box. Cheers from old George

Great project! Keep at it! I wonder if the coils are too close and your getting some interaction? Just a guess. Looking forward to the follow up! 73 de Chris K2CJB

I put Butterworth on my pan cakes. Hi hi Great content KQ4CD

Hi Ape,....Your BP filter should have worked Much closer to the Calculated specs...For those 3 Extremely low value Inductors, I bet just a simple Inductor of a few turns of 18 ga enameled Air core (No Toroidal core) would do much better results....Plus, with Coreless construction Inductors the Filter is much Easier to Tune by simple turns Spacing manipulation....Diy low value Toroids can sometimes Wreck your expected results where a simple Air core Coil works Unexpectedly well....

When I Build Toroids, I Wall Them In Tin Walls, Walls You Can Just Twist And Bend However You Want :) ! I GHet Them Fro Food And Bear Cases, Aluminium, Steel, Tin.. Aluminium can be bolted, not recomended tho. aluminium sucks for rf buids

Thank you. I looked at the original photo. Shorter leads would help. Part layout and orientation changes would help. N0QFT

many mistakes made…like long grounding wires, close located coils and as a result not the best parameters achieved.

Ape, In the video you say the term "broadcast", that really is an incorrect term. Hams technically are not allowed to "broadcast". The proper term is "transmit". Broadcast is a one way transmission like a commercial radio or tv station does. Broadcast is a one way transmission with no expectation of a response or transmission from another station back to you. Hams " transmit" and "receive" because hams want and encourage others to respond to them and have a QSO. Hams operate to initiate a QSO where there is an exchange of information. W1AW at the ARRL "broadcasts" CW practice messages. They have a waiver from the FCC to pay a cw operator to send the CW practice messages. That is the only time a ham can be paid or in a paid position as a ham operator. It is a " broadcast" in that its a one way message or transmission and no response is required or needed to that broadcast. Hams in general do not "broadcast". Broadcast is done in a one to many model. See the first paragraph of the wikipedia page below that I cite. en.m.wikipedia.org/wiki/Broadcasting The U.S. Code of Federal Regulations, title 47, part 97 defines "broadcasting" as "transmissions intended for reception by the general public, either direct or relayed".[6] Private or two-way telecommunications transmissions do not qualify under this definition. For example, amateur ("ham") and citizens band (CB) radio operators are not allowed to broadcast. As defined, "transmitting" and "broadcasting" are not the same.

Perhaps trying the Double tuned Bandpass next time? I realy like it for receivers. It has much steeper skirts. Might have slightly higher IL though but you can easily adjust that by increasing the coupling capacitor. w7zoi.net/transformer_coupled_lc_bpf.pdf The Q of a T50-6 is around 200 at 14MHz