wsprnet results, wsjt-x, and calibration success --> world domination V4 #firmware #calibration

Don - KM4UDX

Hello all. So after living with non-trivial frequency/calibration error, I finally found my boy-bits and went back to reading everything ever written on uBITX calibration (with CEC firmware and memory manager).  The result was extraordinary levels of fear and confusion. 

In desperation, I finally said "foo it",  calculated the error ratio @ 10.00 mhz, multiplied that ratio by the base calibration number to get ~~-5000, then added -5000 in MM, and hit reboot.  Then went to wsjtx's freqcal mode to measure my  new error relative to 3.330 (-7Hz) and 10.00 (-18Hz) reference signals. I was now close enough that I didn't want to touch anything. But I did. And every time I tried to reduce the error further by tweaking the calibration numbers, everything got worse. So I quickly realized I was as close as a mere mortal could aspire. 

But I did run the complete freqcal process in wsjtx, and that process reduced my uBITX freq error to like 1hz or less in wsjtx.  That wsjtx is wicked. If anyone wants help with uBITX wsjtx freqcal mode, I'm your dude.  
Here are my wsprnet results using my humble V4 uBITX.  Without any real effort other than stumbling in the dark, and GREAT help from fellow hams and wonderful persons Ding, KA8BRK, and Dan W3BU, I am ranked ~155 in 2 way WSPRnet spots after only a few days of auto band hopping. NEVER did I expect my little uBITX to get to Australia  and Antarctica.  All hale the mighty uBITX.




Would you consider a write-up or a youtube video on the process?

Don - KM4UDX

There was so much trial, error, fumble, confusion and general stupidity in my process that it is a model of what not to do. 

I distill till the process:
1) express the 10mhz  freq error as a percent delta
2) apply that percent delta to the weird calibration number in memory manager, reboot
3) remeasure the error (a) if close enough to step 4, (b) if still to far off go to step 1
4) run WSJT-X freq calibration process and apply proposed adjustments
5) if your WSPRnet spots still work at all, consider yourself lucky, if you get lots more spots consider yourself very lucky.

This process took me hours and hours to develop and implement.  I recall reading about this exact process in one of my late night uBITX reading deep dives.  But I didn't just implement it. No, I had to invent it on my own at great expense.  

So as usual, I have discovered what others have known all along. 


Jeff Karpinski (N0KAI)

This is right where I'm at now. Sooooo many configuration tips and absolutely none of them are the same. I especially loved the one about setting BFO by tuning to static till it sounded "good". 🙄

Also took me a while to finally realize that even though the config menu options are the same using CEC firmware, they work NOTHING like the factory setup. Finding a partially written manual for CEC clued me in. Using WWV is a great idea, except I live a stones throw from it and WWV is very much NOT NVIS. Gave me an excuse to finally order a proper signal generator.

Don - KM4UDX

I read that bit about BFO adjustments till static sounded "good".  i suspect that the "good" guidance works just fine -- as soon as you ear learns what good static sounds like. hahaha.  

You also have to figure that trial and error is a totally legitimate problem solving technique in this domain. So gird your loins and have at it. (Yes I had to look up the expression..hahah) 

I'll add to my summarization with this extra bit. 

if you want to do band hopping using the wsjt-x band hopper functions (which is wicked cool, says me), then you want an antenna contraption working on more than one frequency. My solution was my 80-10 OCFD run without an antenna tuner.   Yes the SWR is "variable" across the bands, but once you realize that a SWR change from 1.4 to 3.5 is small potatoes on the receiving end, you can embrace a multi-band antenna and have at it. The key here is to not actually look at SWR too often. Just run with it. 

Given that there will be a few no-go bands (no antenna seems to do it all...darn the physics), you do need to sweep your antenna as installed, find the bands with zip resonance, and be sure to EXCLUDE them from the wsjt-x band hopping click boxes.  

The dead zones for me includes 30m, where the SWR is approximately infinite. It wouldn't matter to me, but if you look at wsprnet activity reports, you see that 30m has a sizable percent of all WSPRnet activity.  See the pie charts below.  Turns out that 30m is the third most active band behind 40m then 20m then 30m.  

And my antenna is barren, lifeless, empty, massively indifferent to 30m. Think of all the action I'm missing (like in high school).  I could not let this be. So I've calculated adding  30m OCFD wings split at 29.5% of total (to match the existing 80-10 split point). This is like a fan dipole -- and they work, right?  I expect it will completely screw up the works. I will surely  regret ever trying to improve what is really great as is. But "good enough" can always be made "worser" with enough improvements. 

So i am off to safari seeking the 30m game. Solder at the ready and push onward!


Congrats on turning it loose Don, those are some really nice results.  I'm envious of your antenna doing so well without a tuner.  I hadn't noticed those pie charts before, but yes 30m is a pretty good band these days, for small bandwidth work at least.  

Don - KM4UDX

Thanks -- my goal with the uBITX was multiband ops with mini-effort and max coverage -- simplicity.  I wasn't smart enough to know what this means is you lose low SWR numbers. But then I wasn't smart enough to know that SWR below say 4 doesn't matter much on the receive end. So if you can get the big picture and understand the interaction of large numbers of errors, you realize the approach works great.  hahah

I'm sure the uBITX doesn't like high SWRs at some point, and i know there is no fold back or power reduction/protection circuity. But it seem to do just fine at 4 (or more...I'm just saying).  And yes I pay attention to the drive levels and stay within the envelope, but there is one drive setting for all bands.  I've seen WSJT-X config setting drive level per band, and I may play with that, but the uBITX takes care of itself with one drive setting for all the bands that I can use. 

Based on what I've read, the EFHW 80-10 antennas could do even better than my DIY 80-10 OCFD as they seem to have 30m out of the box. If I wanted to just by a good solution, I'd try these guys (or similar -- there are lots of vendors)

How can you beat 80/40/30/20/17/15/12/10m with no antenna tuner? Several guys in our local radio club have them, and they seem to work.

But given a easy, reliable, effective and efficient commercial solution like a EFHW-80-10, or a marginal self made mess, I will aways pick the mess.  My old boss used to say that when success is too easy, you should put obstacles in your way, just to make victory sweeter.