In the last SolderSmoke Podcast I participated in the subject of Homebrew Versus Kits was covered. One aspect discussed in that session was the totally built homebrew everything. Certainly, to be able to say ALL homebrew would be a high-water mark.
However, there is a huge chasm between something barely working to something that comes close to a commercial element built on a production line. Homebrew elements like LC VFO's, DBM's and Crystal Filters fall in this chasm.
A seasoned homebrew veteran has no qualms about saying homebrew the LC VFO, the DBM and the Crystal Filter. This veteran says DON'T given that this may be a first venture into homebrewing and the first time building something more than a Michigan Mighty Mite.
Now, to defend my position on those elements and along the way I will provide a solution to those new to the art. Creating that huge gap is tribal knowledge.
The Double Balanced Mixer is first up and remember I don't use Single Balanced Mixers. In a homebrew DBM the parts are sparse consisting of two cores and four diodes. Each core consists of three windings. Hey, what could be simpler? A lot is going on with those components and not so simple unless you know some things going in (Tribal Knowledge).
As it turns out how you physically build the DBM and what components you use drives a good DBM versus one that sort of works.
Cores are a starting place. One ham had no ferrite cores and popped in a couple of powdered iron units and his DBM sucked! A good basic starting place is to use the FT-37-43 Ferrite Core.
The next question is how many turns to use? That is a question driven by the frequency of operation. As a starting place 10 to 13 turns per winding will get you there with the 13 turns to cover the lower frequencies.
The next issue is how you wind the wire on the cores. Start by using three different colors of wire and twisting the wires to 8 twists per inch. The winding on the core itself must be spaced evenly around the cores. The whole thrust of different colors is to track the phasing of the wires and the twisting and "even spacing" is to assure a coefficient of coupling approaching 1.
I created three you tube videos on how to build a DBM and you can see the how it is done. The above video is part 2 of 3. The videos even show how to add a "balance pot" and how to use that feature to unbalance the DBM for tune up purposes.
The next performance degrader is the choice of diodes and are they matched. Some hams will suggest you don't need to match the diodes -- well if you want the maximum performance you do. The videos were created for the LBS project and used the readily available 1N914 -- if you want max performance use Schottky diodes!
Short leads and tight spacing between the components are another success factor. I actually found a way to lay out the diodes that are a ring, but don't look like a ring but keeps the leads short.
For a 1st time builder this is more than wind a couple of turns on a core and slap in a bunch of diodes. This suggests buying a couple of ADE-1's and the second unit you build can have the high school science project DBM.
Crystal Filters are the next subject area where you can get really wrapped up in your underwear. Building a high-quality filter is difficult if you have never built one. Difficult in the sense of having a great filter versus it sort of works.
There are a couple of basic parts to a filter that determine the final result. These include the Crystals themselves and how close the frequencies are to each other, the stability of the crystals in PPM, the number of crystals used in the filter, the coupling capacitors and the matching networks.
The coupling capacitors determine the bandwidth with larger values inversely affecting the width. The number of crystals affects the skirts (bottom width) and the matching transformers the ripple. It is only here that we want the ripple to look like a pair of 29A's and not 44DD's.
There is a rigorous process to determine certain parameters that when popped into software will give you capacitor values and impedances for the matching. Often a homebrewer new to building crystal filters skips over the small steps and the filter sounds like crap. Buy a filter for the 1st project and then you can play science project on the second rig you build.
Next is the LC VFO. A ham recently created a totally homebrew DSB transceiver using an LC VFO. He was excited by his 1st contact using DSB and low power only to be told your signal sounds good -- but you are drifting all over the place. Building a stable LC VFO entails a lot of tribal knowledge including the components you use, the physical layout, the physical construction including shielding for RF and temperature and in the end, you are limited to a single VFO range that often is in the range below 10MHz.
True there are some tricks like using a higher frequency crystal filter with a low range VFO, but you also have to carefully study the spurs that get generated. The LC VFO inherently is limiting in a single conversion design. So again, build a Digital VFO for the 1st outing and then you can try your hand at LC VFO's if only to say I have built an LC VFO.
Immediately you ask where I can buy the DBM, Crystal Filter and Digi VFO. You are in luck as not only can you buy these parts at one place but they come mounted to circuit boards so they can be easily interfaced into your build. Check https://www.mostlydiyrf.com Todd K7TFC has the goods where you can buy Mounted ADE-1's, the 8 Pole QER Filter and a Digi-VFO.
There is nothing like a bad experience to cause one to change hobbies like from ham radio to stamp collecting. Being a new homebrewer and attempting to build stable LC VFO's, Crystal Filters and DBM's that will result in a good sounding station is on par with winning a 3 Billion dollar lottery. Possible but not likely. So, to have a great 1st experience think home fabricated with a goal of totally an all-homebrew rig.
Them that know can make it go.
73's
Pete N6QW