Topics

The dreaded LSB and USB swap problem again

 

I hate to ask this again but, I have a brand new V4 board all hooked up on my bench and I cant get it calibrated or the BFO set properly. I've read the past threads regarding this and they just don't make sense to me. It receives great and transmits great except the LSB and USB are swapped, also the calibration is 1.5Khz high in frequency. I am presently running CEC V1.095 firmware. I have a 3.2 inch Nextion display all ready but don't want to add it to the mix until I can get it calibrated running the 16X2 display. Can anybody tell me a simple way to do this.

Joel
N6ALT

Jacob Farnes
 

When I sat down 3 weeks ago to figure the LSB/USB swap and offset out, I followed the following process. I had to follow Ian KD8CEC instructions on the frequency offset fix (actual setting name: "Master Calibration") and load the value via the memory manager.
http://www.hamskey.com/2018/05/how-to-calibration-ubitx.html

Using a RTL-SDR hooked to the Antenna port for a frequency counter for the internal RX measurement.

I then disconnected the two, put a dummy load on the uBitx, a VHF (SignalStick) antenna on the RTL-SDR and set them about 4-6 inches apart. Transmitted CW and took a reading using the various FFT of SDR# and fldigi.

I opened an Excel spreadsheet, labeled 3 rows for each of the Measurements, labeled the 4th "Offset" and punched in formula into the 4th row.

I opened the uBitx Memory Manager program, selected the correct COM port,
Opened the port,
read from the uBitx,
Decoded,
entered the offset value calculated,
Encoded,
Write to uBitx,
Close port.
Disconnect USB cable.
Turn off uBitx.
Turn it back on and check the frequency offset.

Repeat if off. It should be a check once process, but I'm using what I have on hand.

For whatever it's worth, my PPM calibration went fine, but when I that right my V2 uBitx has swapped LSB/USB. I went so far to debug the issue as swapping the Raduino with my Bitx40 as calibration is simpler. I'm able to get concrete PPM and BFO frequencies quickly out of the '40.
My original uBitx Radiuino had a PPM of 248. My Bitx40 had 17 ppm. W0EB was 38ppm. Calibrated quickly using amunters software.

So I plugged in my Raduino from W0EB with a known PPM of 38. USB & LSB were still swapped.

I even went as far as loading an EEPROM erase sketch since I had already messed up my stored settings.

Huh, so I know the problem is on the uBitx board. It's just not as obvious to me where.

Well I calculated out the offset as I think I was getting ready for field day. It about 50Hz off at the moment.

I had to set the offset aside in frustration for now. That's the steps and the relevant findings I had during the last attempt I made to find and solve the magical -2.4kHz VFO offset.

73 de Jacob AG7CT

W2CTX
 

If you have w0eb RadI2CinoI2C card why don't you try the firmware at www.woeb.com and see

if the lsb/usb issue goes away?

rOn

On July 25, 2018 at 10:50 PM "Jacob Farnes via Groups.Io" <kk6jtl@...> wrote:

When I sat down 3 weeks ago to figure the LSB/USB swap and offset out, I followed the following process. I had to follow Ian KD8CEC instructions on the frequency offset fix (actual setting name: "Master Calibration") and load the value via the memory manager.
http://www.hamskey.com/2018/05/how-to-calibration-ubitx.html

Using a RTL-SDR hooked to the Antenna port for a frequency counter for the internal RX measurement.

I then disconnected the two, put a dummy load on the uBitx, a VHF (SignalStick) antenna on the RTL-SDR and set them about 4-6 inches apart. Transmitted CW and took a reading using the various FFT of SDR# and fldigi.

I opened an Excel spreadsheet, labeled 3 rows for each of the Measurements, labeled the 4th "Offset" and punched in formula into the 4th row.

I opened the uBitx Memory Manager program, selected the correct COM port,
Opened the port,
read from the uBitx,
Decoded,
entered the offset value calculated,
Encoded,
Write to uBitx,
Close port.
Disconnect USB cable.
Turn off uBitx.
Turn it back on and check the frequency offset.

Repeat if off. It should be a check once process, but I'm using what I have on hand.

For whatever it's worth, my PPM calibration went fine, but when I that right my V2 uBitx has swapped LSB/USB. I went so far to debug the issue as swapping the Raduino with my Bitx40 as calibration is simpler. I'm able to get concrete PPM and BFO frequencies quickly out of the '40.
My original uBitx Radiuino had a PPM of 248. My Bitx40 had 17 ppm. W0EB was 38ppm. Calibrated quickly using amunters software.

So I plugged in my Raduino from W0EB with a known PPM of 38. USB & LSB were still swapped.

I even went as far as loading an EEPROM erase sketch since I had already messed up my stored settings.

Huh, so I know the problem is on the uBitx board. It's just not as obvious to me where.

Well I calculated out the offset as I think I was getting ready for field day. It about 50Hz off at the moment.

I had to set the offset aside in frustration for now. That's the steps and the relevant findings I had during the last attempt I made to find and solve the magical -2.4kHz VFO offset.

73 de Jacob AG7CT

Jacob Farnes
 

W2CTX,

Sorry if I wasn't clear about the second uBitx Raduino. I bought an original Raduino from W0EB that he decided he didn't need anymore.

I fried my original Bitx40 Raduino via 12V to the A3 orange wire. I hadn't had luck removing the original Nano on that board without risking the Raduino PCB as well. So I bought W0EB orginal Raduino to get my Bitx40 back on the air.

Jacob AG7CT

 

Jacob,

Thanks for pointng me in the general direction. I used the uBitx Memory Manager and loaded Ian’s example calibration values into my uBitx, so now I'm at least up and running. I’m 200hz high in frequency and the BFO setting is off a bit because the audio is very bassy but at least I can tweak it from here. I have to figure what value to enter to get it closer on frequency. The stock calibration routine is totally broken. I was able to use it on my first uBitx successfully but it totally does not work anymore.

Joel
N6ALT

 

I have another question regarding the uBitx Memory manager in the ongoing saga of calibration on the uBitx. Using the memory manager I was able to get the BFO dialed in perfectly, now I'm trying to get the master frequency calibrated. In the Master Calibration box is a number that is the stock configuration, what does that number represent? Is it frequency in Hz, mine has 161645 in that box. I've been making small changes to that number without seeing any change. Right now, I have to tune about 230 Hz lower to be able to hear a conversation normally. So what would I change that number to to change the frequency by 230Hz? Please help if you know.

Joel
N6ALT

 

Okay, through trial and error I finally have it calibrated, both VFO and BFO. Thanks for all the help.

Joel
N6ALT

Raymond Edge
 

Hi Joel,

I am having a similar issue and just can’t seem to hit on the sweet spot. I am right now only .050 off frequency but it is enough to make it annoying.

 

My master Calibration is 143000 at the moment for that setting to still be offset.

 

Any tips would be appreciated.

 

 

Raymond – W9KHP

 

 

This email was brought to you by the Numbers 1 and 0.

No trees were harmed in sending this email but a lot of electrons were seriously inconvenienced.

 

From: Joel Caulkins
Sent: Friday, July 27, 2018 12:15 PM
To: BITX20@groups.io
Subject: Re: [BITX20] The dreaded LSB and USB swap problem again

 

Okay, through trial and error I finally have it calibrated, both VFO and BFO. Thanks for all the help.

Joel
N6ALT

 

 

Raymond,

Nobody was ever able to explain to me what that Master Calibration number represents, so I never knew what number to add or subtract to nudge it one way way or the other, so I decided to change it by a large amount to see what direction it moved. My Master Cal number was 161675 to start with, so I changed it to 131675 and the frequency dropped by about 200Hz, wrong direction, so I added the same amount to 161675 which resulted in 191675 and that got it almost perfect, then I just nudged it up a little at a time until I settled on 191700. Now I'm within a few Hz. Sorry I cant be more specific, you just have to keep playing with the numbers until you get it where you want it. Really, that Master Cal number makes no sense to me at all.

Joel
N6ALT

dan magers
 

Where is this "master cal" to be found?  I have a uBITX V4.3 stock that is way out based in what I'm hearing.  
Perhaps the BITX20 is entirely different.  

I'm not very familiar with the Raduino so, go easy.  

Tnx,
Dan 
N5MRG



On Jul 30, 2018, at 3:34 PM, Joel Caulkins <caulktel@...> wrote:

Raymond,

Nobody was ever able to explain to me what that Master Calibration number represents, so I never knew what number to add or subtract to nudge it one way way or the other, so I decided to change it by a large amount to see what direction it moved. My Master Cal number was 161675 to start with, so I changed it to 131675 and the frequency dropped by about 200Hz, wrong direction, so I added the same amount to 161675 which resulted in 191675 and that got it almost perfect, then I just nudged it up a little at a time until I settled on 191700. Now I'm within a few Hz. Sorry I cant be more specific, you just have to keep playing with the numbers until you get it where you want it. Really, that Master Cal number makes no sense to me at all.

Joel
N6ALT

dan magers
 

I've measured the second LO, clk1 at test point 15 and found it to be 33 and 57 MHz, LSB and USB respectively.  Opposite of what the circuit description gives.  So, I assume the sidebands are reversed.  



On Jul 30, 2018, at 3:48 PM, dan magers <crpdan@...> wrote:

Where is this "master cal" to be found?  I have a uBITX V4.3 stock that is way out based in what I'm hearing.  
Perhaps the BITX20 is entirely different.  

I'm not very familiar with the Raduino so, go easy.  

Tnx,
Dan 
N5MRG



On Jul 30, 2018, at 3:34 PM, Joel Caulkins <caulktel@...> wrote:

Raymond,

Nobody was ever able to explain to me what that Master Calibration number represents, so I never knew what number to add or subtract to nudge it one way way or the other, so I decided to change it by a large amount to see what direction it moved. My Master Cal number was 161675 to start with, so I changed it to 131675 and the frequency dropped by about 200Hz, wrong direction, so I added the same amount to 161675 which resulted in 191675 and that got it almost perfect, then I just nudged it up a little at a time until I settled on 191700. Now I'm within a few Hz. Sorry I cant be more specific, you just have to keep playing with the numbers until you get it where you want it. Really, that Master Cal number makes no sense to me at all.

Joel
N6ALT

Jerry Gaffke
 

Dan,

I assume by circuit description, you mean this from
    http://www.hfsignals.com/index.php/ubitx-circuit-description/
"To invert the sideband between USB and LSB, the second oscillator is switched between 33 MHz and 57 MHz."

I'd say that's just a casual statement, not meant to be read by someone with a law degree.
(But should be corrected.  This has come up before.)

The second oscillator is normally at 45-12 = 33mhz for LSB, and 45+12=57mhz for USB.

The VFO is normally on the high side, at 45 + operating_freq, putting it on the other side of 45mhz
would flip the sidebands (and create lots of new birdies, not recommended).

The stock firmware holds the BFO at slightly below the crystal filter, perhaps 11996500 hz.
Moving the BFO to 11999500 hz would flip the sidebands.
I'm assuming the 12mhz filter passband is roughly 11997000 to 11999000 hz, 
that will vary depending on which sorting bin your crystals were taken from.

The BFO was made low side to avoid audio tones resulting from harmonics of the BFO
beating with harmonics of the 16mhz clock frequency of the Nano.
However, the 16mhz ceramic resonator of the Nano is not very accurate, could be off by 100khz.
I doubt it makes much difference if the BFO is high or low for most rigs. 

Here's a more complete description of what's going on in the stock firmware
along with an attempt at explaining why the sidebands flip.
    https://groups.io/g/BITX20/message/44515


Jerry, KE7ER



On Mon, Jul 30, 2018 at 01:53 PM, dan magers wrote:
I've measured the second LO, clk1 at test point 15 and found it to be 33 and 57 MHz, LSB and USB respectively.  Opposite of what the circuit description gives.  So, I assume the sidebands are reversed.  
 

dan magers
 

Thanks Jerry.  
I had no clk0 when I received it. Found a broken C4 off if pin 10 of the si5351.  
Looking at all the LOs with the Agilent SA is how I determined why I had no rx or tx.  
Now that's sorted out I discovered it sounds like it's way out of the 12mc filter passband.  Or, way down on a skirt.  
The vfo does not change when I try to set it.  Goes back to where I started when I exit.  
Also, the cw keyer works as it should using 2.2k and 10k resistors however, I do not get any straight key function.  
Plus,  ssb tx drops off from 10w to 1watt within a couple of seconds.  

I read where these boards were tested prior to shipment. Hmm...



On Jul 30, 2018, at 4:27 PM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

Dan,

I assume by circuit description, you mean this from
    http://www.hfsignals.com/index.php/ubitx-circuit-description/
"To invert the sideband between USB and LSB, the second oscillator is switched between 33 MHz and 57 MHz."

I'd say that's just a casual statement, not meant to be read by someone with a law degree.
(But should be corrected.  This has come up before.)

The second oscillator is normally at 45-12 = 33mhz for LSB, and 45+12=57mhz for USB.

The VFO is normally on the high side, at 45 + operating_freq, putting it on the other side of 45mhz
would flip the sidebands (and create lots of new birdies, not recommended).

The stock firmware holds the BFO at slightly below the crystal filter, perhaps 11996500 hz.
Moving the BFO to 11999500 hz would flip the sidebands.
I'm assuming the 12mhz filter passband is roughly 11997000 to 11999000 hz, 
that will vary depending on which sorting bin your crystals were taken from.

The BFO was made low side to avoid audio tones resulting from harmonics of the BFO
beating with harmonics of the 16mhz clock frequency of the Nano.
However, the 16mhz ceramic resonator of the Nano is not very accurate, could be off by 100khz.
I doubt it makes much difference if the BFO is high or low for most rigs. 

Here's a more complete description of what's going on in the stock firmware
along with an attempt at explaining why the sidebands flip.
    https://groups.io/g/BITX20/message/44515


Jerry, KE7ER



On Mon, Jul 30, 2018 at 01:53 PM, dan magers wrote:
I've measured the second LO, clk1 at test point 15 and found it to be 33 and 57 MHz, LSB and USB respectively.  Opposite of what the circuit description gives.  So, I assume the sidebands are reversed.  
 

 

Dan,

We are using KC8CEC's firmware 1.08 or above along with KC8CEC's uBitx Memory Manager to tweak the calibration settings. If you used the stock firmware to set your calibration you are probably out in no man's land. Mine was so far out I couldn't get control of it until upgraded the firmware and used the memory manager, then it's relatively easy once you figure out that the mystery Master Calibration number in the memory manager doesn't mean anything, you just keep changing it one way or the other from the the base settings until you get the results you want. Others have accomplished calibration by other means that i didn't understand, like the zero beat method which never worked for me. The calibration process is the weak link in the whole uBitx design, we keep adding all these awesome features to the firmware while the Cal process stays broken, IMO.

Joel
N6ALT