Bearcat 980SSB "PLL Porn"

[brandon7861]

I have a decision to make. Having all A - H channel tables is too much I think. Since I have it rolling over after ch1 or ch40, going from band A to band H would be hard on the encoder. That's hundreds of channels.

If I did 40 up and 40 down, I could just use an LED light as an indicator. Without having touched the VCO slug, it craps out at 27.8MHz. I think even if I do a basic 40 up and 40 down, people will still need to back that slug out just a little.

I also got the fast rotate to do 10k. Rapidly spinning it counterclockwise drops it 10k, rapidly turning it clockwise brings it back to normal. Unfortunately, you'll need to go find your channel again lol.

I am trying to avoid cutting more traces and complicating this for people. Those three traces for the data are bad enough. I'm thinking two or three LED lights. Opinions?

edit: maybe red for 40 up, blue for 40 down and green for -10k. I could use a thin RGB light off one of those sticky light strips.

 

[TM86]

If the issue with tables is they're taking up too much program space you could just go with a single table for the FCC 40 channels. The other "bands" could then be based on a number (representing a starting frequency) with some simple plus/minus math for reading the input from the channel encoder. You're already detecting mode as well so SSB offsets shouldn't be that hard to calculate.

If the issue is the VCO just won't stretch that far, none of the above will help so feel free to disregard it as mindless blathering.

 

[SuperLid]

I have a decision to make. Having all A - H channel tables is too much I think. Since I have it rolling over after ch1 or ch40, going from band A to band H would be hard on the encoder. That's hundreds of channels.

If I did 40 up and 40 down, I could just use an LED light as an indicator. Without having touched the VCO slug, it craps out at 27.8MHz. I think even if I do a basic 40 up and 40 down, people will still need to back that slug out just a little.

I also got the fast rotate to do 10k. Rapidly spinning it counterclockwise drops it 10k, rapidly turning it clockwise brings it back to normal. Unfortunately, you'll need to go find your channel again lol.

I am trying to avoid cutting more traces and complicating this for people. Those three traces for the data are bad enough. I'm thinking two or three LED lights. Opinions?

edit: maybe red for 40 up, blue for 40 down and green for -10k. I could use a thin RGB light off one of those sticky light strips.

I hate to rain on your parade, but that overkill mod, is simple enough for most people. The 980SSB is locked down hard, so something is better than nothing. I really like what that guy did with his mod.

In respect to VCO range, there may be something one could do, with a different varactor swapped in.

In regard to your parade, I don't think any of the other 980SSB modders, have solicited opinions from users, so I suspect, the end result will be more suited to the masses. I have always like the 980SSB, but always thought it could be improved.

SL

 

[brandon7861]

If the issue with tables is they're taking up too much program space you could just go with a single table for the FCC 40 channels. The other "bands" could then be based on a number (representing a starting frequency) with some simple plus/minus math for reading the input from the channel encoder. You're already detecting mode as well so SSB offsets shouldn't be that hard to calculate.

If the issue is the VCO just won't stretch that far, none of the above will help so feel free to disregard it as mindless blathering.

With what is in the radio now, A-H + a custom table ~ 9 full tables with entries for each mode (120 entries per table), I am using 24% of program storage and 15% of dynamic memory. The issue is that with 9 bands and a 10k mode, I need 10 indicator lights. With the current 9 bands, I find that I get lost without looking at the serial.print(bandNumber) feeding back into the computer.

I wanted to make it change the screen color for each band, but that will require more trace cutting and make it more difficult because that data is not on this SPI.

Making the gestures for band selection better isn't the issue, its telling the user what channel table they are in. If I had an easy way to do that, I would gladly come up with better gestures than simply rolling over at the band edges.

 

[nfsus]

With what is in the radio now, A-H + a custom table ~ 9 full tables with entries for each mode (120 entries per table), I am using 24% of program storage and 15% of dynamic memory. The issue is that with 9 bands and a 10k mode, I need 10 indicator lights. With the current 9 bands, I find that I get lost without looking at the serial.print(bandNumber) feeding back into the computer.

I wanted to make it change the screen color for each band, but that will require more trace cutting and make it more difficult because that data is not on this SPI.

Making the gestures for band selection better isn't the issue, its telling the user what channel table they are in. If I had an easy way to do that, I would gladly come up with better gestures than simply rolling over at the band edges.

Instead of lights, does the lcd have an available line to use as a band indicator?

 

[brandon7861]

The LCD uses a 1/3-bias multiplexed drive that I know nothing about. Taking control of a single segment would likely require taking control of the entire display and that would require a better microcontroller and a whole lot of caffeine. I don't think that is something I want to try right now.

I could stick a buzzer in there.

 

[brandon7861]

I hate to rain on your parade, but that overkill mod, is simple enough for most people. The 980SSB is locked down hard, so something is better than nothing. I really like what that guy did with his mod.

In respect to VCO range, there may be something one could do, with a different varactor swapped in.

In regard to your parade, I don't think any of the other 980SSB modders, have solicited opinions from users, so I suspect, the end result will be more suited to the masses. I have always like the 980SSB, but always thought it could be improved.

SL

It was able to get down to 25.2, so there is enough factory varactor there for at least a couple of bands if readjusted.

In Overkills video, his 40 up did not include skips. Because I used the table from Walcott, all bands on mine have skips. I like the idea of removing them. Maybe I should do the same, only letting the 10k offset work where there are subs and have the 10k offset automatically turn off when leaving that channel.

I also like how Overkill gets to the subs, maybe I should rethink my fast rotate and try something like that since I already have code to track channel changes.

 

[SuperLid]

It was able to get down to 25.2, so there is enough factory varactor there for at least a couple of bands if readjusted.

In Overkills video, his 40 up did not include skips. Because I used the table from Walcott, all bands on mine have skips. I like the idea of removing them. Maybe I should do the same, only letting the 10k offset work where there are subs and have the 10k offset automatically turn off when leaving that channel.

I also like how Overkill gets to the subs, maybe I should rethink my fast rotate and try something like that since I already have code to track channel changes.

You are right, there is enough factory C, if the radio goes that low.

I think you will figure it out. Overkill's was simple but effective. The Brazil guy that's hooked up with Troy, is awesome but impractical and just vaporware. Looks like you are in the middle, which seems like a good place to be.

 

[brandon7861]

The code is done! 40 up and 40 down without skips. The 5 skips in the center band can be reached by starting on the channel above, going down one, then back up just like Overkill did (and leaving that channel turns off the -10k). I have three pins set for LED indicators. One pin for 40 down, one pin for 40 up and one pin for -10k skips. That should work good with a single RGB LED. Its pretty much identical to overkills but with 125 channels instead of 85.

Just gotta make the install video. Tomorrow.

 

[SuperLid]

The code is done! 40 up and 40 down without skips. The 5 skips in the center band can be reached by starting on the channel above, going down one, then back up just like Overkill did (and leaving that channel turns off the -10k). I have three pins set for LED indicators. One pin for 40 down, one pin for 40 up and one pin for -10k skips. That should work good with a single RGB LED. Its pretty much identical to overkills but with 125 channels instead of 85.

Just gotta make the install video. Tomorrow.

I didn't really think I would have had much influence on this, but it seems I was wrong. Now, I am in a position, where I will have to to the actual mods. Thing is, I don't use freeband, but it is cool to have.

Maybe a version to do 10m, with various steps?

 

[brandon7861]

I didn't really think I would have had much influence on this, but it seems I was wrong. Now, I am in a position, where I will have to to the actual mods. Thing is, I don't use freeband, but it is cool to have.

Maybe a version to do 10m, with various steps?

I hope to be around for a while, so anyone needing help altering the tables just has to ask. I could make a 10m set and comment it out so switching it to 10m would be as simple as moving a few /* --- */. Let me know what frequencies you want for an alternate table.

It would be super easy for someone to put their own frequencies into the existing channel tables assuming they make sure the VCO slug is in a position to allow it. There are detailed instructions in the sketch for that now. I plan to explain that in the video. I also have an excel (libreoffice) spreadsheet that takes AM frequencies and calculates the sideband tables, so making complete tables for other frequencies is going to be really easy. The spreadsheet already concatenates the comma and the arduino channel comments like " 27205000, //ch20" for each line in vertical tables to copy/paste into the sketch.

I didn't want to have the sketch calculate the SSB tables automatically (they are offset from AM by 2.5k) in case someone wanted to put freeband/10mSSB frequencies alongside AM channels. For example, if you didn't want to change the 40up table and wanted just a few 10m SSB frequencies, you could put those alongside ch1-20. Nobody uses those channels in SSB, so why not make it something else for LSB/USB? Its just another option.

I need to go swap a pitman arm on my plow truck while its warm outside first.

 

[brandon7861]

Well, that whole move the VCO slug idea, not so much. There isn't one. This can next to the PLL is for the 10.24 reference oscillator, not the VCO. The VCO inductor is that little blue one north of the crystal by the varactor. When you turn to ch40 in the upper band (27.805), it is so close to the edge of it locking that it warbles for a moment before settling. That mod will have to come later. Putting together my video now.

 

[brandon7861]

Ok, here it is. The final code and video explaining the install. Not the best video, but it will have to do. Arduino sketch attached.