So, lots of interesting material to read through listed in BPFF Part 1, and much of it is available via the internet. Catch the whole series of w4kaz band pass filter musings.
This project is not really new for me but just something that was recently underscored by recent operations. Deciding which projects to tackle became a matter of choosing those that I thought are ‘possible’. For these purposes ‘possible’ is defined as
- the parts or substitutes are readily available
- the construction appears simple enough for clumsy fingers
- the final tuning is simple
These criteria knocked out the W3NQN filter and the N1AL because I thought tuning the multi-filar toroidal inductors would be enough to exceed my level of patience. The W3NQN design is well recommended functionally, I just thought the tuning procedures were going to prove too difficult. As a side note, the Toroid King sells a kit of the required toroids for the W3NQN filter, and the DL2BNU article describes a method of tuning that is simpler than using a VNA (that I don’t have). The N1AL design bothers me because of the taps on the multi turn inductor and the trial and error tuning. The ARRL sidebar on the N1AL project is interesting, stating that the filters interact to give nulls on the even harmonics. Curious.
A couple of years back, I built the K4VX filters. That project is relatively simple. Tuning can be done with an MFJ antenna analyzer or using a grid dip method. My own analyzer seemed sufficient, and I was able to build filters for 80m, 40m, and 20m fairly simply. The 160m was more difficult to tune and I never achieved a good SWR. It is now being used on the receive antenna to filter out nearby broadcast QRM from 680 and 850 on the AM dial.
The first real problem I ran across came when the 20m filter was mistakenly put into a high SWR load. FUBAR! The damage manifests itself as a high SWR into the dummy load, and the resonances have changed. I suspect the toroids were damaged by overheating. The capacitors used were silver mica’s, and their values seem unchanged.
Another filter project recently bubbled up through the dogpile stream of data. I ran across the NVARC “Ugly” filter project. Besides having a good write up, their project is billed as a “no tune” design.
So far, that seems to be mostly correct. I’ve had success with the two assembled so far. The 20m filter came in with a resonance a bit low, but the loss through it is in the area of about .5db. No tuning required. The 15m filter is actually centered right at 21Mc, and the SWR is dead flat at 1:1 across the entire 15m band. It’s showing insertion losses of about .7db. I have not yet given either of them any on-air trials, so that’s coming later, but the transmit smoke tests into the dummy load(three one minute intervals, 15 seconds apart) showed very little heating in any of the components. After cutting power, the hottest components were just barely warm to the touch. Sweet.
Not so sweet with the first attempt on the 10m version. It originally showed a resonance around 27.500 with an SWR above 2:1 at the lowest. The NVARC document had a discrepancy, which has since been corrected. Re-building the filter as specified in the updated instructions corrected the resonance. It is now good over most of 10 meters – the SWR minimum is around 28.500, and is about 1.4:1 at the bottom of the CW segment, but the insertion loss is low, only about .5db. A look at their VNA sweep shows that it is resonant at the high end of 10m and above. My rendition shows a double dip minimum in the SWR, one centered on 28.500 and a second broader dip at 31.500. Their VNA sweep of the SWR on their filter is very similar to my own SWR plot taken with an MFJ-259, so I declared my reproduction a success.
The NVARC filters are made with air wound coils, using common schedule 40 PVC pipe sections as the coil forms. The guidelines for winding the coils are fairly accurate, and I didn’t really have any problems winding nice, tight coils. The winding process IS a bit hard on the hands, but not impossible. It might be difficult for someone suffering from arthritis. Post winding testing of a set of the coils with K4CZ’s LCR meter showed the inductances on each to be quite close to the value specified on the Ugly Filter schematic.
It occurred to me that the NVARC filter designs might be built just as easily using toroidal cores. It should be simple enough to use an air wound set to come up with resonance values for each section. That will allow for easy rough tuning, similar to the K4VX method. I guess they could then be fine tuned for maximum smoke with an RF probe and a multi-meter. If the air wound design seems to have good performance, I may try one out with toroids. But credit must be given W1XP and the NVARC crew – these filters are easy enough to re-produce, and none of the three required any sort of tuning.
Ultimately I would like to have a pair of each type NVARC and K4VX on 40m and 20m to test forsuitability. For the time being, its a project on hold to gather resources. I gotta get on the 160 mod for the K2, not to mention re-painting thefasciathe roofers had to fix last month when installing the new roof on the QTH.
Note: There will be a bit of a time gap in this “BPFF” series. I have several other projects going on, and of the radio related items I really need to get on the K2 160m option.
First in series: Band Pass Filter Fever – The Tangled Web – Part 1
Next in series: Band Pass Filter Fever – The Guinea Pigs – Part 3.