Re: QSX radio feature requests
J68HZ
Hey Arv… Oh the binary approach does work. Eventually. My comment was more along the lines that it’s the most inefficient way to get to a match. I did a lot of research around algorithms used to solve the classical binary tuning problem years ago (8 discrete coil, 8 discrete capacitors… 65535 combinations.). The best hunting algorithm uses a newton gradient approach to finding the solution (calculate the derivative basis your swr readings). It’s also the fastest of the hunting algorithms. Smarter approach… one used by most military tuners… it to calculate the real and imaginary parts of the load at the frequency of interest. Then calculate the transform to give you exactly 50 ohms real… and implement that as a solution. It’s like plotting the load on a smith chart and drawing the twostep solution to get to 50 ohms. It’s obvious from the diagram what to do. Since there can be a bit of error in the calculations, these algorithms usually either stop and live with a tiny bit of SWR or they then switch to a halving algorithm to tough up the SWR by looking at reactive components on either side of the calculated solution. This is also covered in Sabin’s book. This approach is way faster than the binary/ hunting approach unless you start at the solution to begin with… at which time they both have about the same solution time!
Latching relays. I started looking for some good quality, small, latching RF relays capable of 100 watts. I found two types… a single coilreverse voltage to change state (must be magnetic), and dual coil models (flip flop). I eventually settled on the dual coil model as they are easier to implement than reversing the coil polarity. I found what I needed on Alibaba… It was a direct from the factory deal for several hundred. I asked for a sample to try first. When I was satisfied they worked, I bought several hundred.
Dr. William J. Schmidt  K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ
Owner  Operator Big Signal Ranch – K9ZC Staunton, Illinois
Owner – Operator Villa Grand Piton – J68HZ Soufriere, St. Lucia W.I. Rent it: www.VillaGrandPiton.com
Moderator – North American QRO Group at Groups.IO.
email: bill@...
From: QRPLabs@groups.io [mailto:QRPLabs@groups.io] On Behalf Of Arv Evans
Bill My own QRP ATU uses binary progression of inductances with a 0150 pf variable cap and a switch for adding another 150 pf fixed C. It doesn't seem to be too much of a stretch to have the C values also binary progression. Might take a bit of thought for those not used to programming in machine code but it does seem doable. I'm curious about how small latching relays work. My only experience in that area is with much larger units. They had two coils, one to operate the relay and another to release the lock. Do the small ones work that way. or is there some other methodology? Magnetic latch? Arv
On Fri, Aug 24, 2018 at 8:17 PM J68HZ <bill@...> wrote:


