Keyboard Shortcuts
Likes
Search
First hack at modifying Raduino code
|
Likely nothing too special, I plopped a version of the Raduino sketch in the "KA7OEI" folder ("raduino_ka7oei_20170117.ino") that has the following modifications: - Added oversampling and leaking integration to the tuning pot to both increase resolution of the tuning pot somewhat and reduce the 50 Hz "flickering" that occurs when the A/D can't decide between two steps which can cause an SSB or CW signal to seem to warble badly.? It works slightly better WITHOUT the 0.1uF cap on the tuning pot - at least on receive - as the light amount of noise helps "dither" the result to marginally higher effective resolution.? (If RF gets into it, put the capacitor back!)? The "leaky" integrator help to minimize very slow drift of the potentiometer value causing a drift in frequency, the side effect being a tiny bit of "front-lash" (as opposed to backlash) in the tuning. - Increased the display resolution to show 10 Hz steps now that additional resolution has been synthesized.? Note that not each and every 10 Hz step can always be adjusted repeatedly, but since the above adds a lot of steps between the previous 100Hz displayed steps it makes sense to show the user where they actually are. - Added an on-screen notice - and about a 1 second delay - when the end of the tuning pot is reached telling the user that it will start tuning in 10 kHz jumps.? The user has about 1 second to move the pot back if the didn't really want this to happen. To be sure, a rotary encoder is likely a bit nicer, but I believe this to be a large improvement over the original, default tuning scheme. At some point I'll establish a GitHub account and remember how to use it again, but for now the file is there for immediate use for those who find the 50 Hz "frequency" flickering and the sudden jumping in 10 kHz steps while tuning around to be maddening. 73, Clint KA7OEI |
|
clint, push the change, man! you could also try disabling the output during the frequency change to suppress the clicks. i have now gotten so used to the clicks that i have to look to see if the display is changing on other rigs. - f On 18-Jan-2017 10:33 am, "Clint Turner" <turner@...> wrote:
|
¿ªÔÆÌåÓýHi Clint. Jason 's si5351 library has changed and I had to tweek the arguments. Also, the line 2 printing is an issue. (garbage display)? But the underlying code works well! Arduino 1.8.1, si5351 library v2 73. Good Work.? Larry kb3cuf Sent from my Galaxy Tab? E -------- Original message -------- From: Clint Turner <turner@...> Date: 1/18/2017 00:03 (GMT-05:00) Subject: [BITX20] First hack at modifying Raduino code Likely nothing too special, I plopped a version of the Raduino sketch in the "KA7OEI" folder ("raduino_ka7oei_20170117.ino") that has the following modifications: - Added oversampling and leaking integration to the tuning pot to both increase resolution of the tuning pot somewhat and reduce the 50 Hz "flickering" that occurs when the A/D can't decide between two steps which can cause an SSB or CW signal to seem to warble badly.? It works slightly better WITHOUT the 0.1uF cap on the tuning pot - at least on receive - as the light amount of noise helps "dither" the result to marginally higher effective resolution.? (If RF gets into it, put the capacitor back!)? The "leaky" integrator help to minimize very slow drift of the potentiometer value causing a drift in frequency, the side effect being a tiny bit of "front-lash" (as opposed to backlash) in the tuning. - Increased the display resolution to show 10 Hz steps now that additional resolution has been synthesized.? Note that not each and every 10 Hz step can always be adjusted repeatedly, but since the above adds a lot of steps between the previous 100Hz displayed steps it makes sense to show the user where they actually are. - Added an on-screen notice - and about a 1 second delay - when the end of the tuning pot is reached telling the user that it will start tuning in 10 kHz jumps.? The user has about 1 second to move the pot back if the didn't really want this to happen. To be sure, a rotary encoder is likely a bit nicer, but I believe this to be a large improvement over the original, default tuning scheme. At some point I'll establish a GitHub account and remember how to use it again, but for now the file is there for immediate use for those who find the 50 Hz "frequency" flickering and the sudden jumping in 10 kHz steps while tuning around to be maddening. 73, Clint KA7OEI |
|
Hi Larry, At the time I modified the code I hadn't found the notes about the change needed to accommodate the newer '5351 code - it seems as though there is (or was) no obvious, single FAQ-type of document on the "hfsigs" site or in this group's archives (the Wiki is still empty!) so it took a lot of reading of past message just to put the various bits and pieces together before I got the sketch to compile properly.? Still, not too bad for 1-2 hours' work and scattered documentation! About the 2nd line being "garbaged up", I never had that problem, at least when using? < Version 2 '5351 code:? One would think that the generic LCD functions would be reasonably robust - unless there's some sort of blocking going on in the background. I've been fighting my 3D printer when trying to print an ABS box for the radio, so that - and a few other projects - must be done before I can get time to put the radio on the air:? Too many projects, too little time! TNX for the comments and 73, Clint KA7OEI |