Ten-Tec Scout 555 60m Band Module: IF Filter Redesign & Final Testing [Part 4]

This is the final episode in my Ten-Tec Scout 555 60-meter band module series:

- Part 1: Initial Conversion and Filter Design

- Part 2: Crystal Selection and Mixer Circuits

- Part 3: Field Testing and Troubleshooting

- Part 4: IF Filter Redesign (this post)

Today saw me finish the TenTec Scout 555 60 meter Band Module Project in the workshop, now it is on to phase 2, testing in the field… I had wanted to rework the filter in the IF stage as it was not great from before. I was able to get a signal out of the radio, but the filter shape left something to be desired…

The first chart image is of my filter after I got the radio to work. This is not what a bandpass filter should look like. The lower photo is of a factory TenTec VERY narrow bandwidth 40 meter bandpass filter. Turns out, this is almost impossible to replicate with discrete components that are not adjustable…as you will see. They used two tunable inductors to achieve this filter shape. I almost went to the trouble of adding some trimmer capacitors to mine, but as you will soon see, I felt it wasn’t needed. I soon learned what a filter Q is and why it matters when your trying to build a filter. I used some online calculators to get the component sizes for the filters and after messing around with a couple, I found one of the calculators had came up with numbers that were actually able to be made in my shop.

Armed with this information I decide to breadboard this filter first so I could tinker with it before soldering it to the band module circuit board. This, as it turns out, is a terrible idea if you plan to simply transfer the parts to the board and solder it all together. There is SO much stray capacitance and inductance at RF frequencies with a breadboard that you can build up a circuit, but when it comes to making the final item on perf board or Manhattan style, that you WILL use different values.

Looks good doesn’t it? Well look at the plot below! It actually looked great! It did have some insertion loss that I was not super happy about, but that plot looks great! Yes, the one inductor is hand made, I didn’t have one small enough in my little kits to work here I so wound one for the job. Notice how long the leads are on those components, that will come into play very soon…

I may have gotten a little aggressive with the passband width… I think it was set to 200khz or maybe 500khz, but it was not much. I think the values I put into the online calculator were 5.4mhz center frequency, 500khz bandwidth and 3 LC stages in a “T” configuration. The bottom line on the chart is -70dB which is unbelievable! This thing was incredible! So I get the 60 meter module out and take it apart and strip out the old filter I had previously built wholesale leaving a clean slate for the new filter…

After I had the old filter gone and cleaned up a little, I cut me a pad as the new filter layout and the TenTec layout are different so I wanted to not have to permanently alter the circuit board other then removing the old components. This pad became the central connection point for the three stages. A little hot glue and I was in business! I simply tinned the whole top of the pad so I could land part leads where ever I wanted on it. This worked really well…much better than the filter I am about to build as it would turn out.

SIDE NOTE ABOUT MODIFYING SCOUT BAND MODULES:

What I am pointing out in these two photos is a home made Via that I added. You see, when you remove the canned tunable inductors from the board, turns out you break the grounds. TenTec decided that since these things were going to be here anyway, might as well use them for something. So they apparently decided to make them connect one ground plane to another in a couple of places on the board. Once you remove the cans, you lose a critical ground path…or two. I took a finger drill and a very small drill bit and drilled a holed from one trace to the ground plane on the opposite side(I am confident these are only two layer boards), then I soldered a bonding jumper wire in the hole so both planes would once again be connected again and all the stuff would continue to work properly. You also get to see the old homebrew filter in the below photo as a bonus.

Below is the result of me simply attempting to move the parts over to the circuit board and soldering them in place… What a mess. I started with the exact parts you saw on the breadboard, but when I connected the nanoVNA to the filter I found the center frequency had moved up 500khz! It was now pushing my desired frequency out of the passband!!! I was seeing something like 16dB of loss at the desired frequency.

This is the world of ultra high Q passband filters made with inductors and capacitors I guess. Once I removed all the stray capacitance from the breadboarded system, all the capacitor values went down enough to make the filter not usable. That is why you see the mess below, I didn’t have room for the trimmer caps I have on hand so it was good ole trial and error method instead to get the passband in the right spot. To be fair, this mess actually worked, it just had something like 12 or 13dB of insertion loss in the filter which made the output power go way down. I used the amp meter to check it and the current draw on transmit with this filter was 1.8 amps and the current draw with the factory 40 meter band module was 9.3 amps. So I took a break and grabbed some food and went back to the drawing board…

I am going to admit something here that might make some people unhappy with me…I decided to use Claude AI to see what it would come up with for a filter design. I gave it all the parameters I had used earlier and (since it is a chatbot after all) had a conversation about what my goal was. I decided to move the bandwidth out to 1mhz giving me a decent shot at making a working filter without adjustable parts and using only what I had available on site.

This AI model came up with a “T” filter but instead of the center 2 LC components being in parallel, they were in series and so I built it. It tested really bad as far as rejection goes as it only attenuated about 8 or 10dB across the whole band and looked more like a poorly designed high pass filter instead. This did give me some inspiration though. So I went back to the online calculators that are there to figure part values for you and changed my approach. I went to this website and just made a low pass filter and then a high pass filter and coupled them together with an inductor. Why did I choose an inductor? Because all the capacitors I tried kept increasing the losses in the filter output is why. As it turns out, if you attenuate the IF signal in the radio the output power get lowered and the radio goes “deaf”…

Below is what I came away with finally. Yes, only 3 inductors and two capacitors.

Below is the filter shape for this simple filter design too. It is a lot better than what I started with and I was able to build it with parts that I had available to me at the time. Could the passband be narrower? Yeah, it could, but I am not going to throw too much shade at this filter. This is what ham radio is all about in my book, experimentation. The marker is at 5.64mhz which is 250khz above the highest place I can transmit so I know this filter wont attenuate my signal and the radio will work as it should.

The original values for the filter ended up being wrong, but I simply did what I did in the previous filter build and just used it as a jumping off point and started switching out parts one at a time with different values. This tactic coupled with looking at the nanoVNA to see what the new part did to the filter shape gave me the results you see. At first I didn’t write down what I had done, but it quickly became apparent I needed to note what I had done and so things progressed much more quickly once I did.

The above photo ( bottom of the second page ) is what I started with on this part and the lower photo is what I ultimately ended up with for the actual parts I used to build the filter. You can see where I was trying all sorts of stuff to get a working filter, these are not the only sketches either…

This is what I ended up with ultimately to get the filter shape above. The impedances are wrong, I am sure as I didn’t take them into consideration at all, but the radio is working correctly now (from what I can tell) and transmits almost at full power. Shoot, they could be right and I just got them right by accident, I really don’t know to be honest.

This is the output power draw now for the 60 meter module on the same radio as the 40 meter module previously compared. I can live with that! I checked it when I got into the shack at the house and it is a little over 35 watts into a 1.3:1 SWR load. That will work just fine for me. (Note: The current reading was into a 50 ohm dummy load at the workbench so I don’t know the exact current draw in the shack.) This radio will produce more output power too, I have it dialed back on purpose to about 40 watts for the 20 meter band module that I use with this radio.

This is what the final filter looked like when I was finished building it. This looks a lot better than that mess I started out with above. Not overly complicated and generates a respectable filter shape. The radio sounds good too. So I am happy with this.

Side note about the above photo, I had noticed in my original filter design that the orientation of the inductors made a difference in the filter profile on the nanoVNA so I started marking them with a paint pen to remember the orientation. This filter didn’t seem to care too much about that for some reason but the original cared ALOT.

The side view photo shows something else the I have not mentioned in these write-ups. The filter has space constraints. The filter can not be taller than that metal post or the other circuit board with hit the parts and the parts can not go past the post either as that is the space for the output low pass filter from the other board. So parts count matters…

It feels oddly satisfying to see the finished module and to use it in the shack. All I did last night was listen to some stations having a QSO and how well it could hear them. I also dialed up the webSDR again and listened to myself calling CQ for a bit with no takers on a different channel. I know the radio is working as the listening station in Utah could hear me…

Tonight I hope to find a chance to jump into a QSO with someone and see how it sounds to them.

I hope this has inspired you to do something with your gear whether it be build a kit or modify a device to do something new or different or even to repair something you have that is not working now…shoot maybe you will learn something you didn’t know before! That is what has happened to me on this journey, I have learned a lot about how filters work and what affects them in use. I am starting to understand impedance matching the stages as well as insertion loss. A lot of things I never understood before are now becoming more clear. I also have a much clearer understand about how the TenTec Scout 555 radio works as well…

Thanks for following along on this 4-part journey! If you build your own 60-meter module for the Scout 555, I'd love to hear about it. 73!

WK4DS - David