#qcx #firmware 1.03 release (CAT, real time clock) #qcx #firmware


Hans Summers
 

Hi all

I have now officially released 1.03 firmware and we will use this version for all kits shipped by QRP Labs from this moment onward. 

The download is here: http://qrp-labs.com/images/qcx/firmware/T1.03.zip for upgrading if you have your own AVR programmer, or the chip is also available in the QRP Labs shop see http://shop.qrp-labs.com/firmware/qcxchip 

Firmware 1.03 contains:
  1. CAT Control (TS-480 subset)
  2. Real time clock option, in bottom right of display, enabled by menu "7.8 Show clock"
  3. Clock will be automatically set from GPS if it is temporarily connected; 
  4. GPS plugged in, puts the radio in Practice mode AND does not key it anymore
  5. When GPS is plugged and removed, Practice mode is automatically switched off again
  6. Setting speed to 0, puts the QCX in Straight key mode
  7. Remove spurious short TX enable at power-up
I have also added a new version of the QCX manual at http://qrp-labs.com/qcx which describes the CAT commands and the method to connect a serial port. 

The real time clock is a display with format HH:MM which can be enabled to show at the bottom right of the screen. It is enabled in new menu "7.8 Clock" and you can set the time in the beacon menu "6.9 Set time". It is not maintained when the QCX is powered down. To make things easy, you can now connect the GPS whenever you like; there will be no damage because the QCX will automatically detect serial data and put the rig safely in Practice mode (no transmission) within under 1 second. Then the QCX will start parsing the serial data and will set the real time clock. When you disconnect the GPS the QCX will automatically notice that and switch off Practice mode. In this way, it is easy to set the real time clock - in my QCX-40 in its nice enclosure, I plan to install an extra switch which simply connects the shack GPS serial data to the QCX. Then when I turn on the QCX in the morning I can just switch on GPS to set the time, within a couple of seconds. Easy! 

I enjoyed having the real time clock there, I operated 40m CW this morning for a little over an hour and worked YL Elvira ZA1EM, followed by E74LZ, UR5MUY, IK7XJA, LZ1HDA, OM3CAZ, R7BW, EW8CP and RW3KE. The real time clock was useful, at least to me... and I did not have any problems with the new firmware or any of the functionality. I haven't connected anything to the serial port as I don't use CAT. 

The other change is the CW speed: setting it to 0 (with a turn of the rotary encoder knob) will put the QCX in straight key mode and you can then use this for antenna tuning - all while still in the CW speed adjustment mode - then when you turn the speed back to your ordinary keyer speed you will be put back in your chosen Iambic mode. 

Finally, the TX keying enable, pin 11 of IC3, goes high 3 times when starting up the QCX... this is a topic I will mention in a separate email. In this firmware, I was able to remove one of these three times. This is also documented in this new article about my PA failure: http://www.qrp-labs.com/qcx/qcxmods/ptt.html

Note that from now on we will supply QCX chips with low fuse 0xD7 not 0xF7. The new recommendation will be 0xD7. The old F7 has a 65ms delay on start-up for slowly rising power line but this is NOT required when you are using a Brownout Detector bit since the brownout logic will hold the reset anyway until the power line comes up (ATmega328 datasheet). D7 removes the 65ms startup delay. 

73 Hans G0UPL
http://qrp-labs.com



Jim W8JD <jimd2002@...>
 

Great stuff Hans.  Mine programmed perfectly and I really like the 0 speed feature for tune ups and such.  I did notice that the transmit decoder just returns mostly "T"s if attempting to send as if in straight key mode.  Straight key mode works normally and I turn that mode off in 5.5 anyway as I find it distracting.  I only noticed it because it defaulted to on when I re-programmed the chip.   I'm not using CAT but you are correct, the clock is very useful and just looks in place displayed there.

Jim 
W8JD


Jim W8JD <jimd2002@...>
 

Also, In my case I actually like the fact that the clock resets to 0 because If I am operating on my battery it keeps track of the on time.  Very nice.

73s
Jim
W8JD


Tom N9GVP
 

Thanks, Hans.  The clock is a nice feature. 

I noticed that if you switch the VFO to split mode (using the right button), then the first digit of the time display does not show.  However, as soon as I turn the right rotary dial, the full time does display correctly.  Or, instead of turning the dial,  I can just wait until the clock updates to the next minute and then the full time is displayed correctly.  Not a problem.  Not sure if this behavior occurs for anyone else.


Hans Summers
 

Hello Tom

I did some tests here and at first I still could not see that. 

However after some further experimentation with settings, I could reproduce it. The left digit of the clock is overwritten when in Split mode, only when the CW decoder is active. The clock display gets corrected at the next minute, or when you tune the frequency (as you noted). Any new character that is added to the CW decoder display, also causes the left digit of the Clock to disappear. Only when in Split mode. 

In Split mode with the Clock displayed, the decoded CW text has nowhere to go. It was supposed to indeed, go nowhere... to not display at all... but apparently it didn't like being squeezed. Using:
  • Real time clock
  • CW decoder
  • Split mode
All at the same time is fundamentally incompatible anyway because the real time clock and the Split display (VFO B) don't leave any space for any CW decoder text. 

So this is a very minor thing really... I will fix it in due course because it is a bit ugly... but I think it is just a cosmetic annoyance that will hardly affect anyone.

Thanks for bringing it to my attention!

73 Hans G0UPL

On Wed, Dec 4, 2019 at 1:00 AM Tom N9GVP <tsnyder35@...> wrote:
Thanks, Hans.  The clock is a nice feature. 

I noticed that if you switch the VFO to split mode (using the right button), then the first digit of the time display does not show.  However, as soon as I turn the right rotary dial, the full time does display correctly.  Or, instead of turning the dial,  I can just wait until the clock updates to the next minute and then the full time is displayed correctly.  Not a problem.  Not sure if this behavior occurs for anyone else.


Andrea Massucco IZ1IVA
 

Il giorno 03 dic 2019, alle ore 12:35, Hans Summers <hans.summers@...> ha scritto:

Note that from now on we will supply QCX chips with low fuse 0xD7 not 0xF7. The new recommendation will be 0xD7. The old F7 has a 65ms delay on start-up for slowly rising power line but this is NOT required when you are using a Brownout Detector bit since the brownout logic will hold the reset anyway until the power line comes up (ATmega328 datasheet). D7 removes the 65ms startup delay.
Dear Hans,

does T1.03.zip incorporate the new fuse settings or do I have to manually fiddle with them when re-flashing an old chip?
I’m a total ATmega328 newbie running 1.00c firmware (OEM chip).

TNX
--
73 de Andrea IZ1IVA - cq at iz1iva . net - http://iz1iva.net



--
-- 73 de Andrea IZ1IVA - cq at iz1iva . net - http://iz1iva.net


Tom N9GVP
 

Thanks, Hans, for the info on the time display when in split mode with CW decode ON.  It really is just a cosmetic issue.  Not a problem at all.
-Tom


Gregg Myers
 

Has anyone else tried to calibrate their QCX crystal frequencies with GPS with firmware T1.03? It doesn't work for me. I get strange button responses, and the QCX never does update the crystal frequencies properly or at all. I reverted back to T1.02b and the GPS calibration works fine.

73,
Gregg W7GRM

On Wed, Dec 4, 2019 at 9:05 AM Tom N9GVP <tsnyder35@...> wrote:
Thanks, Hans, for the info on the time display when in split mode with CW decode ON.  It really is just a cosmetic issue.  Not a problem at all.
-Tom


Hans Summers
 

Hello Gregg

I have just tried a fresh QCX-40 with a fresh 1.03 chip, and the frequency calibration worked fine. Steps:

1) Go to menu 8.13
2) Plug in GPS
3) Wait some seconds until I see satellite lock
4) Turn the rotary encoder one click anticlockwise to 8.12 and wait 6-7 seconds for the Sys Osc cal
5) Turn the rotary encoder one click anticlockwise to 8.11 and wait 6-7 seconds for the Ref Clk cal
6) Unplug GPS
7) Go to menu items 8.5 and 8.6 to check calibrations ran fine

If you have a different sequence of steps and it is failing please can you specify, so I can try that? 

Are you sure your fuse settings are correct (no divide-by-8 fuse etc)?

73 Hans G0UPL



On Thu, Dec 5, 2019 at 7:54 AM Gregg Myers <gregg.w7grm@...> wrote:
Has anyone else tried to calibrate their QCX crystal frequencies with GPS with firmware T1.03? It doesn't work for me. I get strange button responses, and the QCX never does update the crystal frequencies properly or at all. I reverted back to T1.02b and the GPS calibration works fine.

73,
Gregg W7GRM

On Wed, Dec 4, 2019 at 9:05 AM Tom N9GVP <tsnyder35@...> wrote:
Thanks, Hans, for the info on the time display when in split mode with CW decode ON.  It really is just a cosmetic issue.  Not a problem at all.
-Tom


Hans Summers
 

Hello Andrea

The hex file does not incorporate fuse settings. You have to set fuses in your programmer software when you flash the chip. 

As I said on the PTT mod article though... this is unlikely to ever be an issue. Changing the fuses makes the startup 65ms faster. I suppose if you:

1) Used IC3 pin 11 to key an external amplifier AND
2) Did NOT perform the pullup resistor mod detailed at http://www.qrp-labs.com/qcx/qcxmods/ptt.html AND
3) Have some power supply anomaly that initiates a spike on power-up in some circumstances 

Then if ALL three of those match your situation, then there is a POSSIBILITY that changing this fuse to eliminate the 65ms delay at startup MIGHT reduce the probability that you don't blow a PA transistor with a self-destructive spurious spontaneous oscillation. 

There's a lot of conditions and IFs and MAYBEs so really changing this fuse is not important at all. I am mostly just informing people that it is harmless to remove this 65ms delay by changing the fuse and that we will supply QRP Labs programmed chips like this from now on. 

73 Hans G0UPL
http://qrp-labs.com

On Wed, Dec 4, 2019 at 12:53 PM Andrea Massucco IZ1IVA <ml@...> wrote:
> Il giorno 03 dic 2019, alle ore 12:35, Hans Summers <hans.summers@...> ha scritto:
>
> Note that from now on we will supply QCX chips with low fuse 0xD7 not 0xF7. The new recommendation will be 0xD7. The old F7 has a 65ms delay on start-up for slowly rising power line but this is NOT required when you are using a Brownout Detector bit since the brownout logic will hold the reset anyway until the power line comes up (ATmega328 datasheet). D7 removes the 65ms startup delay.

Dear Hans,

does T1.03.zip incorporate the new fuse settings or do I have to manually fiddle with them when re-flashing an old chip?
I’m a total ATmega328 newbie running 1.00c firmware (OEM chip).

TNX
--
73 de Andrea IZ1IVA  -  cq  at iz1iva . net  -  http://iz1iva.net



--
-- 73 de Andrea IZ1IVA - cq at iz1iva . net - http://iz1iva.net




Andrea Massucco IZ1IVA
 

Il giorno 05 dic 2019, alle ore 07:01, Hans Summers <hans.summers@...> ha scritto:

There's a lot of conditions and IFs and MAYBEs so really changing this fuse is not important at all. I am mostly just informing people that it is harmless to remove this 65ms delay by changing the fuse and that we will supply QRP Labs programmed chips like this from now on.
Understood. Thank you Hans!
--
73 de Andrea IZ1IVA - cq at iz1iva . net - http://iz1iva.net



--
-- 73 de Andrea IZ1IVA - cq at iz1iva . net - http://iz1iva.net


Gregg Myers
 

Hi Hans,

Thanks for testing this. I had used a slightly different sequence.
1. Power is off
2. Plug in GPS
3. Power on
4. Go to Alignment 8.11
5. At this point, I start to see the weird behavior... For example, the QCX automatically seems to get out of the menu without me doing anything (showing VFO A and blank line underneath without me doing anything). But actually, I find it is still in the menu settings if I turn the rotary encoder back and forth 
6. If I go back to 8.11 then hit select, It doesn't seem to calibrate. But just now I was playing with it and turning the rotary encoder between 8.11 and 8.12, eventually it starts to calibrate correctly.

If I use your sequence I still saw my menu item appear to blank out when I hot plugged the GPS, but trurning the rotary encoder back and forth to 8.11 and 8.12 made it work.

So the behavior still seems different compared to T1.02b, where I never had the blank screen and I never had to turn rotary encoder back and forth. But anyway, I did get it to calibrate.

I had similar behavior on two different QCX. I didn't change fuse settings on either unit. Maybe I have noise eminaiting from my GPS connection? But I never saw this on earlier firmware.

Anyway, I can see how I can get it to work now. I like the new features you added in T1.03. keep up the good work!

73,
Gregg W7GRM



On Wed, Dec 4, 2019 at 10:55 PM Hans Summers <hans.summers@...> wrote:
Hello Gregg

I have just tried a fresh QCX-40 with a fresh 1.03 chip, and the frequency calibration worked fine. Steps:

1) Go to menu 8.13
2) Plug in GPS
3) Wait some seconds until I see satellite lock
4) Turn the rotary encoder one click anticlockwise to 8.12 and wait 6-7 seconds for the Sys Osc cal
5) Turn the rotary encoder one click anticlockwise to 8.11 and wait 6-7 seconds for the Ref Clk cal
6) Unplug GPS
7) Go to menu items 8.5 and 8.6 to check calibrations ran fine

If you have a different sequence of steps and it is failing please can you specify, so I can try that? 

Are you sure your fuse settings are correct (no divide-by-8 fuse etc)?

73 Hans G0UPL



On Thu, Dec 5, 2019 at 7:54 AM Gregg Myers <gregg.w7grm@...> wrote:
Has anyone else tried to calibrate their QCX crystal frequencies with GPS with firmware T1.03? It doesn't work for me. I get strange button responses, and the QCX never does update the crystal frequencies properly or at all. I reverted back to T1.02b and the GPS calibration works fine.

73,
Gregg W7GRM

On Wed, Dec 4, 2019 at 9:05 AM Tom N9GVP <tsnyder35@...> wrote:
Thanks, Hans, for the info on the time display when in split mode with CW decode ON.  It really is just a cosmetic issue.  Not a problem at all.
-Tom


Gregg Myers
 

Hi Hans,

Slightly correction to my earlier email. When I tried testing your sequence, I started on 8.11 and not 8.13. I guess if I started n 8.13, I would naturally have to change the encoder to get to 8.11 or 8.12 and it would have worked that way even better.

73,
Gregg


On Thu, Dec 5, 2019 at 6:04 AM Gregg Myers via Groups.Io <gregg.w7grm=gmail.com@groups.io> wrote:
Hi Hans,

Thanks for testing this. I had used a slightly different sequence.
1. Power is off
2. Plug in GPS
3. Power on
4. Go to Alignment 8.11
5. At this point, I start to see the weird behavior... For example, the QCX automatically seems to get out of the menu without me doing anything (showing VFO A and blank line underneath without me doing anything). But actually, I find it is still in the menu settings if I turn the rotary encoder back and forth 
6. If I go back to 8.11 then hit select, It doesn't seem to calibrate. But just now I was playing with it and turning the rotary encoder between 8.11 and 8.12, eventually it starts to calibrate correctly.

If I use your sequence I still saw my menu item appear to blank out when I hot plugged the GPS, but trurning the rotary encoder back and forth to 8.11 and 8.12 made it work.

So the behavior still seems different compared to T1.02b, where I never had the blank screen and I never had to turn rotary encoder back and forth. But anyway, I did get it to calibrate.

I had similar behavior on two different QCX. I didn't change fuse settings on either unit. Maybe I have noise eminaiting from my GPS connection? But I never saw this on earlier firmware.

Anyway, I can see how I can get it to work now. I like the new features you added in T1.03. keep up the good work!

73,
Gregg W7GRM



On Wed, Dec 4, 2019 at 10:55 PM Hans Summers <hans.summers@...> wrote:
Hello Gregg

I have just tried a fresh QCX-40 with a fresh 1.03 chip, and the frequency calibration worked fine. Steps:

1) Go to menu 8.13
2) Plug in GPS
3) Wait some seconds until I see satellite lock
4) Turn the rotary encoder one click anticlockwise to 8.12 and wait 6-7 seconds for the Sys Osc cal
5) Turn the rotary encoder one click anticlockwise to 8.11 and wait 6-7 seconds for the Ref Clk cal
6) Unplug GPS
7) Go to menu items 8.5 and 8.6 to check calibrations ran fine

If you have a different sequence of steps and it is failing please can you specify, so I can try that? 

Are you sure your fuse settings are correct (no divide-by-8 fuse etc)?

73 Hans G0UPL



On Thu, Dec 5, 2019 at 7:54 AM Gregg Myers <gregg.w7grm@...> wrote:
Has anyone else tried to calibrate their QCX crystal frequencies with GPS with firmware T1.03? It doesn't work for me. I get strange button responses, and the QCX never does update the crystal frequencies properly or at all. I reverted back to T1.02b and the GPS calibration works fine.

73,
Gregg W7GRM

On Wed, Dec 4, 2019 at 9:05 AM Tom N9GVP <tsnyder35@...> wrote:
Thanks, Hans, for the info on the time display when in split mode with CW decode ON.  It really is just a cosmetic issue.  Not a problem at all.
-Tom


f1mqj
 

Hi Hans,
Trying GPS with 1.02b and 1.03, I get some small problems :
I've tried 2 GPS modules: VK2828U7G5LF integrated antenna (old), Neo-6M external SMA antenna,
Pb1 - I never got "f" data other than __ and datas displayed (A 3D) seems not very acurate (puting the GPS module to ucenter give a good 3D fix)
Pb2 - with the VK2828 i never gain menu 8.12 and 8.11 working (I plan to check the PPS output of this module to check if it's not too small -as you wrote somewhere)
Pb3 - the most important (but not a great problem if we are aware of it) :
the first time after startup I go to the 8.12 menu the calibration frequency show an erroneous number 20,032693 20,024373 20,018131 (the good one for me is 20,006124) and write it in the 8.6 data
rotating the selector to switch on 8.11 then, about 4s after, go back to 8.12 then it compute the (assuming) true calibration number
As the system frequency is only used for timing purpose, it seem to be very far from critical but i'd prefer to warn you
Pb3 is the same with 1.02b and 1.03
Best 73
--
FR4SW - F1MQJ - QCX 20m - not yet used


Z32ZZZ
 

Hi Hans,

Just to bring it to attention.
If Rx decode is enabled, and there is some decoded message in the second line, when frequency is tuned, decoded text is deleted until new text is updated/displayed. This happens in VFO A, VFO B and also in Split mode for the right part of the second row. This mean that decoded text is not lost, but only cleared and not restored on each dial change. Similar, decoded text is restored back also on each change of the VFO displayed, A, B or Split by using right button.
Nothing problematic, only cosmetic issues.

73, Zoran Z32ZZZ


Hans Summers
 

Hi Remi & Zoran

I will check these problems but I do not think either of them are related specifically to 1.03. The lost decoded CW text on dial change is a pre-existing thing and Remi, I think the GPS problems you mention would apply to all QCX versions. It is tested for compatibility with QLG1 but some other modules may be less successful... but I will have a look at that and see if anything can be done. 

73 Hans G0UPL

On Mon, Dec 9, 2019 at 5:01 PM <zj0369@...> wrote:
Hi Hans,

Just to bring it to attention.
If Rx decode is enabled, and there is some decoded message in the second line, when frequency is tuned, decoded text is deleted until new text is updated/displayed. This happens in VFO A, VFO B and also in Split mode for the right part of the second row. This mean that decoded text is not lost, but only cleared and not restored on each dial change. Similar, decoded text is restored back also on each change of the VFO displayed, A, B or Split by using right button.
Nothing problematic, only cosmetic issues.

73, Zoran Z32ZZZ