Date   
Re: UBITX_CEC firmware upload and uBITX_Manager from a Mac #ubitx #radiuno #firmware

Jack Purdum
 

Tom:

You probably already know this, but the Arduino IDE wants the compile directory name to be the same as the INO file that holds setup() and loop() functions, ubitx_20.ino in this case. During development, I change that to something like ubitxV001.ino, so I get a directory structure like:

   E://JackAl/Software/ubitxV001/ubitxV001.ino  

(I use my E: drive for development) and the rest of the project's files are in the same subdirectory. I also renamed any other files either *.cpp or *.h as needed and not *.ino except for the file with setup() and loop() functions. This allows me to have type checking across compile elements. Using some kind of naming convention means that project names can have some meaning.

Jack, W8TEE

On Monday, May 21, 2018, 5:03:06 AM EDT, Tom, wb6b <wb6b@...> wrote:


Hi,

I've managed to get the CEC uBITX Memory Manager running and the uBITX_CEC firmware uploaded to my uBITX from my Mac.

uBITX Manager:

I have Microsoft Visual Studio Community (a free version) already install on my Mac, so Mono was already installed. When I first tried to run uBITX_Manager it crashed with the warning "The Carbon driver has not been ported to 64bits, and very few parts of Windows.Forms will work properly, or at all".

After a little Googling I found the solution was to run mono in the 32 bit mode with the following command:
mono --arch=32 uBITX_Manager.exe

The uBITX Manager UI runs painfully slow on the Mac. Scrolling is next to impossible. But, with patients it is useful. The buttons will be covered over often, but just hover the mouse over them to bring them back.

The manager crashed when connecting to the USB serial chip unless the "Linux" mode check box was checked.

Also, I already had the USB serial driver installed because I'd purchased a temperature logger from this company "Elitech" and the driver was included with their software. However, the Mac "Homebrew" installer now has an updated signed driver for the CH340 chip. I found this on Github: https://github.com/adrianmihalko/ch340g-ch34g-ch34x-mac-os-x-driver 

uBITX CEC Firmware Upload:

I decided the best way to upload the firmware was through the Arduino sketch IDE. I downloaded the firmware source code and compiled it. The only issue I had with comping the code was I named the base directory with the version number of the code as part of the name. The source code wants to be in a directory named ubitx_20. I installed this directory under the "~/Documents/Arduino/" directory where all my other sketches are located. I have a fair number of libraries already installed, so it is possible you may need to add a library here or there if the CEC uBITX firmware needs any other libraries.

Running the Firmware:

The firmware seems to be running just fine. I used the uBITX Manager to copy the memory values from my Radiuno and save them to a file. I'm particularly interested in using the CAT interface to control my uBITX from a Raspberry Pi.

This software and the fact that it exposes all the memory settings, may have already started to point to some of the mysteries involving my radio. I have noticed my receiver is about 100hz low on sideband signals. That is no big deal. However, it looks like my transmitter frequency is about 1khz low on CW. I can't quite wrap my head around the standard factory setup calibration method and if it could result in an unwanted difference in the transmit and receive frequency calibrations (or fix it). 

I wanted to try the IF shift to get a handle of where my radio may be operating in the 12Mhz crystal filter bandpass. Interestingly, tuning the IF shift up a kHz makes the received SSB signals much clearer. So, maybe my uBITX is out of calibration. 

There truly are a lot of numbers and such to digest in the uBITX Manager, but I'll read the posts (and the synthesizer chip's app notes) on how the calculations are done and then I'll know for sure when my radio is calibrated. At least now, rather than not quite knowing what the assumptions are in the factory calibration, I can directly experiment with the values to understand them and see what works. 

Tom, wb6b


Re: Arduino Nano dds

VE7CWS WRSeiler <waltrseiler@...>
 

Thank you! 

WRS

Re: JackAl Board Debut

Jack Purdum
 

We'll keep you posted! Good seeing you...

Jack, W8TEE



On Monday, May 21, 2018, 2:07:49 AM EDT, Craig Thibodeaux <km4yec@...> wrote:


JackAl, 

Interesting presentation Al did at the Show and Tell.  Last minute trip to Dayton for the Hamfest and sneak a ride over to Club Night and Show and Tell, and a JackAl…

Eagerly waiting for the chance get one.

Craig
KM4YEC


On May 20, 2018, at 11:17 PM, Jack Purdum via Groups.Io <jjpurdum@...> wrote:

All:

The FDIM conference was a good time and the trip to nearby Xenia for the Hamvention was also great. The rain didn't help, but it was better than last year, and I'm sure it will be even better next year.

Friday night was a sort of Show-and-Tell at FDIM and we used that opportunity to show our JackAl board in action. This photo shows a little more about what it is and does:

<FDIMDebut001.JPG>

At the show, Al hooked up a noise generator to the µBITX to show how the filters work. (There are 4 preset filters for CW and 4 for SSB.) In addition, you can customize one CW and SSB filter to the bandwidth you desire. You might be able to see that the skirts are pretty steep for the filters on the scope in the background. The setting of the CW custom skirts are set differently, in that you pick a center frequency (e.g., 700Hz in the shot below), press the encoder, and then you see this:

<IMG_0468.JPG>


In this case, turning the encoder CCW increases the bandspread (i.e., the 440 red number above, although it looks orange in the photo) by simultaneously moving the skirts (480Hz and 920Hz) further apart. If you turn the encoder CW, you narrow the bandspread. Most CW users will probably center the bandpass on their favorite sidetone frequency, which centers the bandpass on that frequency.

The demo used a 5" display, although a 7" display is also available. The third knob on the front is for a second encoder that we use for everything from setting the CW keyer speed to adjusting the filter skirts. You can see some of the plots on the panel at the rear of the picture above for some of the board's features (e.g., filter responses, compression, etc.) Those will be included in the documentation when the (downloadable) manual is finished.

The JackAl board has the following features:

    5" or 7” touch screen 800x480 TFT color display
    Dual VFO's
    RIT
    S meter
    RTC
    CW keyer, 5 to 50wpm (we could go up to 100wpm, but...really?)
    Up to 50 CW preset messages, selectable at runtime...perfect for contest messages
    Touch screen function and control selection (e.g., band changes, RIT, mode, VFO, VFO increment, LSB/USB, etc.)
    Automatic LSB/USB selection based on frequency (overrideable)
    One touch frequency increment changes (1Hz to 1MHz in multiples of 10...the white underscore in the frequency window)
    Dual encoders (frequency, features)
    EEPROM storage of user preferences (one-click reset to "factory" defaults)
    Uses Teensy 3.6 processor (1Mb flash @180MHz) and companion audio board
    Support for 3 external CW push button switches (NO) for sending stored CW messages (e.g., contesting)
    Hardware AGC using IF take-off
    Audio AGC with adjustable threshold
    Mic compressor with adjustable threshold
    8 band audio equalizer
    Receive audio filter:
        48dB/octave (8 pole equivalent DSP filters)
        4 CW presets (150, 300, 400, 600, [or none] Hz 3dB bandwidth) + 1 user-defined knee frequencies (at runtime!) filter
        4 SSB presets (1500, 1800, 2200, 3000, [or none] Hz 3dB bandwidth) + 1 user-defined knee frequencies (at runtime!) filter
        Variable Notch filter, encoder adjustable, use specified Q
    7 watt power amplifier

The board will be distributed with all (surface mounted) parts in place. The user must supply the Teensy 3.6 ($30), its supporting audio board ($15), and the 5" ($34) or 7" ($44) touch screen displaying (using the RA8875 controller chip, BuyDisplay.com). We expect the JackAl board to sell for $50. We may need to adjust this price as we have only received "ballpark" cost estimates for the board since we only have the Gerber files for the Beta board. Currently, we are using less than 20% of the available flash memory (out of 1Mb) and less than 15% of the SRAM (256K), so there is plenty of memory resources available for adding "stuff". The board also brings out a number of I/O pins to help your experimentation. With the exception of removing one SMD resistor on the µBITX board and soldering two wires to those pads, all interconnections are via existing connectors.

Our best guess is that after finishing the modified Gerber files, production, Beta testing, and writing support manuals, it will be probably two months before we begin distribution. We will announce its availability here as soon as we can. BTW, if anyone knows a high-quality PCB manufacturer who also does pick-and-place at reasonable prices, we are getting quotes and would like to know about them.

Jack Purdum, W8TEE
Al Peter, AC8GY


<FDIMDebut001.JPG><IMG_0468.JPG>

Re: Volume potentiometer

Ralph Mowery
 

I used a fuse holder like that.  However I put it external to the ubitx in line with the positive lead of the power cord.  That way if you install the reverse polarity diode across the power connector and hook the wires up in reverse, the diode will often short our and blow the fuse before any damage to the ubitx boards.  It looks like you installed the fuse after that diode.


On Mon, May 21, 2018 at 12:25 AM, W7PEA <patrick@...> wrote:

Love it !! 

I added this to the uBITX assembly page as well.



Thanks!
W7PEA


Re: JackAl Board Debut

Jack Purdum
 

Some options to fix the "scoot" problem:

   some non-skid rubber feet
   a 10-penny nail through the case
   a SPDT CO switch wired to the TUNINGENCODERSWITCH pin

Jack, W8TEE

On Monday, May 21, 2018, 1:07:59 AM EDT, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:


Pressing down on a switch handle is easier then pressing in on that encoder shaft,
With the latter, the rig scoots across the desk and often as not the frequency changes a little bit.
Having an up direction on the switch gives the user interface a lot more options.

Though not a requirement.  I can live with a single switch on the encoder shaft.


On Sun, May 20, 2018 at 09:35 pm, Jack Purdum wrote:
What's the SPDT CO switch for?
 

Re: JackAl Board Debut

Terry Morris
 

You still have one of those Doug?


Terry - KB8AMZ
Brimfield Twp, OH  USA
Linux User# 412308, Ubuntu User# 34905
OSs: LM18.2 64bit, Ubuntu 16.04, tahrPup64 64bit, Raspian
Orgs: PCL70-FOP, NTHS, ALUG, ARRL, PCARS#78, NAQCC#6668, NO-QRP-C, QRP-ARCI#8855, SKCC#14195, USN 1965-1969 AG3

I chair the PCARS CW/QRP SIG and Linux for Hams SIG, second and fifth Tuesday

my computer, my opinion

On Sun, May 20, 2018 at 11:57 PM, Doug W <dougwilner@...> wrote:
very nice!
how sweet would that look in this...




--
www.bitxmap.com


Re: JackAl Board Debut

Terry Morris
 

Some options to fix the "scoot" problem:

a large 'C' clamp
strapping from a chimney antenna mount



Terry - KB8AMZ
Brimfield Twp, OH  USA
Linux User# 412308, Ubuntu User# 34905
OSs: LM18.2 64bit, Ubuntu 16.04, tahrPup64 64bit, Raspian
Orgs: PCL70-FOP, NTHS, ALUG, ARRL, PCARS#78, NAQCC#6668, NO-QRP-C, QRP-ARCI#8855, SKCC#14195, USN 1965-1969 AG3

I chair the PCARS CW/QRP SIG and Linux for Hams SIG, second and fifth Tuesday

my computer, my opinion

On Mon, May 21, 2018 at 10:34 AM, Terry Morris <terry.kb8amz@...> wrote:
You still have one of those Doug?


Terry - KB8AMZ
Brimfield Twp, OH  USA
Linux User# 412308, Ubuntu User# 34905
OSs: LM18.2 64bit, Ubuntu 16.04, tahrPup64 64bit, Raspian
Orgs: PCL70-FOP, NTHS, ALUG, ARRL, PCARS#78, NAQCC#6668, NO-QRP-C, QRP-ARCI#8855, SKCC#14195, USN 1965-1969 AG3

I chair the PCARS CW/QRP SIG and Linux for Hams SIG, second and fifth Tuesday

my computer, my opinion

On Sun, May 20, 2018 at 11:57 PM, Doug W <dougwilner@...> wrote:
very nice!
how sweet would that look in this...




--
www.bitxmap.com



Re: JackAl Board Debut

Jerry Gaffke
 

Here's an option to fix undesired tuning changes when pushing the encoder switch.

Within function read_enc()  which returns the number of encoder ticks that have occurred,
use the Arduino millis() call to determine how long it has been since the non-zero return value.
If there has been a pause in activity of more than 1000ms, then return a result of 0 encoder shaft ticks
but retain an internal copy of the total number of accumulated ticks since the last non-zero return.
When that total number of accumulated ticks exceeds some threshold, then go ahead and
give a non-zero return value, but only return the final encoder read, not the total accumulated ticks.
Giving only the final encoder read value and discarding the accumulated ticks allows me to
advance the knob by just a few hz without overshoot. 

And it doesn't involve any 10 penny nails:

Jerry, KE7ER


 

On Mon, May 21, 2018 at 07:14 am, Jack Purdum wrote:
Some options to fix the "scoot" problem:
 
   some non-skid rubber feet
   a 10-penny nail through the case
   a SPDT CO switch wired to the TUNINGENCODERSWITCH pin
 
Jack, W8TEE
 

Re: JackAl Board Debut

Jack Purdum
 

We don't do polling on the frequency encoder, we use an interrupt. When you're cruising along at 180MHz, it's pretty responsive to what you're doing. If you have set the increment to 1Hz, that's what you get.

Jack, W8TEE

On Monday, May 21, 2018, 11:01:21 AM EDT, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:


Here's an option to fix undesired tuning changes when pushing the encoder switch.

Within function read_enc()  which returns the number of encoder ticks that have occurred,
use the Arduino millis() call to determine how long it has been since the non-zero return value.
If there has been a pause in activity of more than 1000ms, then return a result of 0 encoder shaft ticks
but retain an internal copy of the total number of accumulated ticks since the last non-zero return.
When that total number of accumulated ticks exceeds some threshold, then go ahead and
give a non-zero return value, but only return the final encoder read, not the total accumulated ticks.
Giving only the final encoder read value and discarding the accumulated ticks allows me to
advance the knob by just a few hz without overshoot. 

And it doesn't involve any 10 penny nails:

Jerry, KE7ER


 
On Mon, May 21, 2018 at 07:14 am, Jack Purdum wrote:
Some options to fix the "scoot" problem:
 
   some non-skid rubber feet
   a 10-penny nail through the case
   a SPDT CO switch wired to the TUNINGENCODERSWITCH pin
 
Jack, W8TEE
 

Re: boosting the power on 28 MHz #ubitx

atouk
 

Since it's always nice to have someone doublecheck your work, would this be the proper place to add the cap if you mount it on top of the board?

Henry
N2VFL

Re: Headphone jack retaining nut

Tom Frobase
 

I have some if you have not found one already ... tom, N3LLL

On Sun, May 20, 2018 at 12:33 PM, brad martin <emclinux@...> wrote:
Anyone know where I can get the retaining nut for the headphone jack? Mine
was missing in the bag of parts. I know I can get them by getting some more
of the jacks but I only need one and really have no need for a bag of 10
just for the little nut.

Brad

Re: RF power chain mods and improvements..

ajparent1/KB1GMX
 

That might work in the Q90 spot, maybe a pair in the predriver.  Try them and let me know.

I hadn't considered that one as I'ts been maybe 35-40 years since I've put them in circuits.
I'm surprised its not long extinct.  I'd have to check to see if there is a reliable supply anymore.
I checked the big three and its likely unobtainium ( I have 9 of them in my stock, that proves it!).

If I had to put up a list of maybe this would work of all the possible devices it would number
in the hundreds if not more.  There are over 8000 2nxxx parts alone to pick from then
2SC, 2SD, MRF, SRF, BLF...  Not enough hours or days.


Allison

Re: JackAl Board Debut

Kees T
 

Very nice Jack and Al. It was good to see a wooden breadboard/plaque as the base to give it the "down to earth" perspective.
In anticipation, I had, but didn't need, my brown paper bag (to prevent hyperventilation). 

Great use of a color TFTs 

EasyEDA does not only boards but also will do assembly work.....but that may be "handing the design to them". Advanced Circuits also does assembly work as do some local business here in Austin. You might ask where Midnight Solutions (George Herron) has his NUE-PSK boards assembled today.

73 Kees K5BCQ  

Re: boosting the power on 28 MHz #ubitx

a.vision
 

Looks correct 



Sent from my Samsung Galaxy smartphone.

-------- Original message --------
From: atouk <atouk@...>
Date: 21/05/2018 16:13 (GMT+00:00)
To: BITX20@groups.io
Subject: Re: [BITX20] boosting the power on 28 MHz #ubitx

Since it's always nice to have someone doublecheck your work, would this be the proper place to add the cap if you mount it on top of the board?

Henry
N2VFL

Re: Headphone jack retaining nut

Bo Barry <wn4ghv@...>
 

tirIf you still need one let me know. I bought a bag of replacements since soldering them twice usually melts them. I have leftovers from the original.  I rewired
A little confusing since the replacements had a different wiring pattern.
Bo W4GHV since '54


On Mon, May 21, 2018 at 11:17 AM Tom Frobase <tfrobase@...> wrote:
I have some if you have not found one already ... tom, N3LLL

On Sun, May 20, 2018 at 12:33 PM, brad martin <emclinux@...> wrote:
> Anyone know where I can get the retaining nut for the headphone jack?  Mine
> was missing in the bag of parts. I know I can Get get them by getting some more
> of the jacks but I only need one and really have no need for a bag of 10
> just for the little nut.
>
> Brad
>



Re: pin 2 of raduino board seems to have gone to CW TX full time after boot message.

Arv Evans
 

I have two Nano's here in my hand, one from a US dealer via Amazon and the
other from a Chinese seller via Ebay.  Looking at both with 10X magnification
I see no difference.  Even the lettering is identical.  This makes me believe that
either the Chinese are buying US made Nano or US dealers are buying and
reselling Chinese made Nano.  Another possibility is that all of them may be
made in some 3rd country, maybe Mexico.  In today's global manufacturing
environment it is possible that all the Arduino Nano devices are made in one
place and sold by a myriad of dealers all over the world.

Arv
_._


On Sun, May 20, 2018 at 9:32 PM, Mike <msmith@...> wrote:
Best solution:

They're inexpensive enough to keep a half dozen Nano's or Mini's on hand. 
I've destroyed many by tinkering.

Also, my experience is it doesn't matter much who the vendor is - they're all made with same quality from the same gerbers and components. It's just of matter of waiting on a Chinese vendor or paying an extra dollar to get them more quickly from a US vendor.


Re: RF power chain mods and improvements..

ajparent1/KB1GMX
 

Since I had the 2n2369 and 2n2222A I put them in.

Q90 is 2n2369A
The rest (6 in all) are 2n2222A...

The test was done this way set the pot for 80M and 10W.  test all bands with no voltage or pot changes.
FYI this unit on 80m and 13V does about 13W and nearly that much 11.7W on 40M with the pot at max.
However my goal is to get 20m and up to better than 5W with the pot set to 10 at 80.

Results:
80m (pot set for 10W)  (this made not significant change)
40M 9.8w    Significant improvement over the prior parts (6.5W)
20M 5.7W   disappointing, no improvement
10m 1.9W   Better than the previous 1.4, but not much!

The 2n2222A is very much similar to the 2219A (same die) and I've used a pair pushpull
as an instrumentation power amp with a fairly flat 300mW output to 30mhz so it indicates
there are other factors at work.  I am eyeing the Driver and output transformers. 

I'll make up some different transformers and try them as I'm sure to not see the ordered
parts for a week.

One thing I wish the software had is an easy band change.too much cranking the encoder to get
from 80 to 10.


Allison

Re: JackAl Board Debut

Steve Black
 

Why over think this one. Duct tape. Now in matching colors! Steve kb1chu



Sent from my Verizon, Samsung Galaxy smartphone

-------- Original message --------
From: Terry Morris <terry.kb8amz@...>
Date: 5/21/18 10:45 AM (GMT-05:00)
To: BITX20@groups.io
Subject: Re: [BITX20] JackAl Board Debut

Some options to fix the "scoot" problem:

a large 'C' clamp
strapping from a chimney antenna mount



Terry - KB8AMZ
Brimfield Twp, OH  USA
Linux User# 412308, Ubuntu User# 34905
OSs: LM18.2 64bit, Ubuntu 16.04, tahrPup64 64bit, Raspian
Orgs: PCL70-FOP, NTHS, ALUG, ARRL, PCARS#78, NAQCC#6668, NO-QRP-C, QRP-ARCI#8855, SKCC#14195, USN 1965-1969 AG3

I chair the PCARS CW/QRP SIG and Linux for Hams SIG, second and fifth Tuesday

my computer, my opinion

On Mon, May 21, 2018 at 10:34 AM, Terry Morris <terry.kb8amz@...> wrote:
You still have one of those Doug?


Terry - KB8AMZ
Brimfield Twp, OH  USA
Linux User# 412308, Ubuntu User# 34905
OSs: LM18.2 64bit, Ubuntu 16.04, tahrPup64 64bit, Raspian
Orgs: PCL70-FOP, NTHS, ALUG, ARRL, PCARS#78, NAQCC#6668, NO-QRP-C, QRP-ARCI#8855, SKCC#14195, USN 1965-1969 AG3

I chair the PCARS CW/QRP SIG and Linux for Hams SIG, second and fifth Tuesday

my computer, my opinion

On Sun, May 20, 2018 at 11:57 PM, Doug W <dougwilner@...> wrote:
very nice!
how sweet would that look in this...




--
www.bitxmap.com



Re: RF power chain mods and improvements..

Skip Davis
 

Good tests Allison in the band changing that is where Ian’s firmware shines, two presses of the encoder and your able to jump from band to band. 
I didn’t get time to play but I pulled out two binocular cores for winding new transformers.
I’ve got 2N2222A’s in SM package to try in at Q90, I’ll let you know the results. 

Skip Davis, NC9O

On May 21, 2018, at 11:39, ajparent1/KB1GMX <kb1gmx@...> wrote:

Since I had the 2n2369 and 2n2222A I put them in.

Q90 is 2n2369A
The rest (6 in all) are 2n2222A...

The test was done this way set the pot for 80M and 10W.  test all bands with no voltage or pot changes.
FYI this unit on 80m and 13V does about 13W and nearly that much 11.7W on 40M with the pot at max.
However my goal is to get 20m and up to better than 5W with the pot set to 10 at 80.

Results:
80m (pot set for 10W)  (this made not significant change)
40M 9.8w    Significant improvement over the prior parts (6.5W)
20M 5.7W   disappointing, no improvement
10m 1.9W   Better than the previous 1.4, but not much!

The 2n2222A is very much similar to the 2219A (same die) and I've used a pair pushpull
as an instrumentation power amp with a fairly flat 300mW output to 30mhz so it indicates
there are other factors at work.  I am eyeing the Driver and output transformers. 

I'll make up some different transformers and try them as I'm sure to not see the ordered
parts for a week.

One thing I wish the software had is an easy band change.too much cranking the encoder to get
from 80 to 10.


Allison

Re: pin 2 of raduino board seems to have gone to CW TX full time after boot message.

ajparent1/KB1GMX
 

Based on volume prices I cannot see why they would use scrapings.
The 328 in the 32pin smt is a dirt cheap part, even digikey get only 1.245$ at 100pcs and
it gets under a buck at higher volumes (5000 piece reel).  Digikey is not the cheapest around.

The actual part is assembled elsewhere so boards are likely assembled elsewhere.
Its not like there is engineering involved as its already defines and the bootloader
is out there too.

I picked up a while back via amazon 10 pro-minis for about 2.35 each! 
If I have to I can substitute the pro-mini with a bunch of wires to put the right mini pins
at the Raduinos nano's locations for them.  I don't need or want USB cat and both
boards use the exact same chip.  I have a proto Raduino using a pro-mini,
Adafruit Si5351 board and a cheap backlit 2x16 LCD.  Words exactly the same
save for I use a FTDI RS232TTL-3V3 to program it instead of a usb cord and set 
IDE for the demilove rather than nano.

Allison