Yet another DIY GPSDO - yes, another one

[Johnny B Good]

So no parts, no problems?

 More properly put: "No parts, no problems, no project."

 Sorry, I couldn't resist.

[nealix]

This project sounds interesting and fun.  I have ordered the needed parts.

Andrew, do you have an approx date when you plan to merge the software changes/fixes that
have been discussed in the last month, and update your GIT site with a new version?

Sincere Thanks for designing this project.  It has far fewer parts than some of the other
DIY GPSDO projects, including LARS.

Cheers,

Neal

[felixd]

Have you ever encountered issue like this with GPS?

Code: [Select]

$GPTXT,01,01,02,u-blox ag - [url=http://www.u-blox.com]www.u-blox.com[/url]*50
$GPTXT,01,01,02,HW  UBX-G60xx  00040007 FF7FFFFFp*53
$GPTXT,01,01,02,ROM CORE 7.03 (45969) Mar 17 2011 16:18:34*59
$GPTXT,01,01,02,ANTSUPERV=AC SD PDoS SR*20
$GPTXT,01,01,02,ANTSTATUS=DONTKNOW*33
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64
$GPTXT,01,01,02,ANTSTATUS=INIT*25
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64
$GPTXT,01,01,02,ANTSTATUS=OK*3B
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64
$GPRMC,,V,,,,,,,,,,N*53
$GPVTG,,,,,,,,,N*30
$GPGGA,,,,,,0,00,99.99,,,,,,*48
$GPGSA,A,1,,,,,,,,,,,,,99.99,99.99,99.99*30
$GPGSV,1,1,00*79
$GPGLL,,,,,,V,N*64

I have looked for last NMEA codes and it says it's Internal Error of GPS...
I start to look right now for an answer in Intenet, but I am asking here, maybe you had issues like this before

EDIT:

Looks like some !@#!@$ antena connection... ... 
It is working again. I am not touching them anymore


Cheers

[iannez]

Hi Felixd,
I've never seen that kind of mistake before.
I imagine that if you restart everything back to normal.
Have you verified the number of active satellites at the moment?
And the other parameters detected?

in the meantime I'm trying to understand the difference of
"Fix Time" between two GPS modules (Chinese ...)

This because it often comes out:

Fix Time 1001ms and even more

and this is not good because it clears the counter even if I have the FIX of the Counter PPS.

bye A.

[mendip_discovery]

Felix, I recently has issues with my Lightening detector. Often it's a power issue, either no power getting to the antenna or in my case the usb power supply dying. The data coming from the chip just shows as default data but no time fix.


I must get around to getting mine working, can't let Bob beat me to getting one up and running. ;-)

[AndrewBCN]

This project sounds interesting and fun.  I have ordered the needed parts.

Andrew, do you have an approx date when you plan to merge the software changes/fixes that
have been discussed in the last month, and update your GIT site with a new version?

...

I should be merging the various pull requests and my own further developments in the coming days.

[AndrewBCN]

Have you ever encountered issue like this with GPS?

...

EDIT:

Looks like some !@#!@$ antena connection... ... 
It is working again. I am not touching them anymore


Cheers

Yes, I think I wrote it before, but it's worth repeating: any GPSDO depends on getting a good signal from as many satellites as possible (10 or more, preferably), and for that you need a good antenna and a good view of the sky.

[bob91343]

What ever happened to your plans for a PCB?  Maybe I missed it.

[mendip_discovery]

What ever happened to your plans for a PCB?  Maybe I missed it.

I think it is in the pipeline but I did see AndrewBCN says he hasn't been well.

[felixd]

@AndrewBCN I have few questions. I have made some changes to code, some things were really confusing for me in the beggining so I decided to change some names to something more meaningful.

Changes that are done, are made on fit-gpsdo branch on my fork. (this is WIP - Work In Progress branch so it can be messy )

https://github.com/felixd/STM32-GPSDO/tree/fit-gpsdo

Idea is: GPSDO_FUNCTION_IDENT

How it looks right now on my side:

Code: [Select]

#define GPSDO_DISPLAY_OLED  // SSD1306 128x64 I2C OLED display
//#define GPSDO_VCTL_MCP4725         // MCP4725 I2C 12-bit DAC
#define GPSDO_VCTL_PWM  // STM32 16-bit PWM DAC, requires two rc filters (2xr=20k, 2xc=10uF)
//#define GPSDO_SENSOR_AHT10           // I2C temperature and humidity sensor
#define GPSDO_GEN_2kHz  // Generate 2kHz square wave test signal on pin PB9 using Timer 4
//#define GPSDO_SENSOR_BMP280      // SPI atmospheric pressure, temperature and altitude sensor
//#define GPSDO_SENSOR_INA219          // INA219 I2C current and voltage sensor
// TODO: When UART2 (Bluetooth) is activated less messages are shown on OLED and UART1 (USB serial)
#define GPSDO_UART2         // UART2 (Bluetooth) serial (HC-06 module)
#define GPSDO_ADC_5V      // Vcc (nominal 5V) ; reading Vcc requires 1:2 voltage divider to PA0
#define GPSDO_ADC_3V3     // Vdd or Vref (nominal 3.3V) reads AVREF internal ADC channel
// TODO: Below Calibration definition is not used..
#define GPSDO_CALIBRATION   // Auto-calibration is enabled
#define GPSDO_GPS_UBX_CONFIG   
//#define GPSDO_GPS_VERBOSE_NMEA  // GPS module NMEA stream echoed to USB serial xor Bluetooth serial

--

Vcc and Vdd were super confusing for me in the beggining (same on Schematics). It was not clear what Voltages I should expect so I've changed names to point what voltages we are working with. Same with variable names

--

GPSDO_CALIBRATION is not used at all. From my point of view it should work as it is working right now: Calibration is done everytime after restart. This option sould be removed from code.

--

GPSDO_GPS_VERBOSE_NMEA I have just dirrectly connected to Serial1 on breadboard to see all GPS<->STM32 traffic. That's much better in my opinion. You have proper NMEA messages which can be decoded by u-Center app. An for that, this port should be available as well on PCB

--

I have also formated code (autoformating in Arduino IDE v1 with indent-preproc-block option ON)

Example of what I am writting about:

Code: [Select]

#ifdef GPSDO_SENSOR_INA219
  #include <LapINA219.h>   // LapINA219 library library
  LapINA219 ina219(0x40);  // create object INA219 with I2C address 0x40
  float ina219volt = 0.0, ina219curr = 0.0;
  TwoWire Wire3(PB4, PA8);  // Second TwoWire instance for INA219 on SDA3/SCL3 (should be put somewhere more fitting but must stay global)
#endif                    // INA219

--

Once functions / block are recognized and corrected in GPSDO.ino they should be move outside. Example GPS should be moved to gps.h / gps.c file. Same with sensors

--

UART output format: I suggest to format data in Arduino IDE Serial Plotter format.
https://docs.arduino.cc/software/ide-v2/tutorials/ide-v2-serial-plotter

You can directly save that to CSV file and process data later on. You can as well parse data with an external application.

--

That's where I am slowly going to. Waiting for Your feedback.

BTW: My Racal-Dana 9514 (last calibrated 11/10/2012) was only 0.4 Hz off (it's running for more than 30 hours already). I have just adjusted it, which was a bit tricky but I have managed to do it 

https://www.eevblog.com/forum/testgear/list-your-test-equipment-score-here!/msg790750/#msg790750

Cheers, Paweł.

[iannez]

Hi felixd,
I also made various modifications to the code ...

GPSDO_CALIBRATION is not used now also because
To enable or disable it, just use the following:

force_calibration_flag = false;
ocxo_needs_warming = false;

So doing bypasses from loop entry into the functions.

I'm running the compiled sketch for more than 26 hours.
Only with serial and si5351 for frequency generation on I2C.
All counters are perfectly consistent.

The frequency no because I simply read it from the si5351.

I'm waiting for Aliexpress a VCXO, hopefully good.

Fix Time 997MS
Uptime: 02:04:51  <= counter reset 24hours
SATS: 6 HDOP: 1.28

Frequency Measurements Using 64-bit Counter:
64-Bit Counter: 938771955800
Frequency: 9999997 Hz
10s Frequency AVG: 9999997.8 Hz
100s Frequency AVG: 9999997.85 Hz
1,000s Frequency AVG: 9999998.017 Hz
10,000s Frequency AVG: 10000000.0028 Hz


I have to check if the fixtime changes with the change of GPS module, I think so but I have to try again, I don't remember it.

In any case if it takes more than 1000ms just put 2 in:


const uint16_t waitFixTime = 2;


A bit of things should be reviewed but let's see what AndrewBNC says.

bye, A.

[mendip_discovery]

I'm waiting for Aliexpress a VCXO, hopefully good.

Why go for a VCXO over OCXO?

[AndrewBCN]

...
I think it is in the pipeline but I did see AndrewBCN says he hasn't been well.

After managing to avoid the 1st, 2nd, 3rd and 4th waves of SARS-COV-2 here in France, I was finally infected with the Omicron variant 10 days ago (community transmission, I don't know who I got it from or who I passed it on to - also I am fully vaccinated + booster with Pfizer), spent a week in isolation coughing and sneezing, and today (Sunday 30/01) is the first day I felt good enough to go out and walk a little bit and ride my bicycle.

I still have quite a few things to work out but in the coming days I should be able to find the time to sort the various pull requests from Pawel as well as my own accumulated changes and prepare a new release by next Sunday (6/02), if all goes well. I think I will name the next release "Omicron" - hehehehehehe, cough, cough. 

Thanks all for your patience! 

[mendip_discovery]

...
I think it is in the pipeline but I did see AndrewBCN says he hasn't been well.

After managing to avoid the 1st, 2nd, 3rd and 4th waves of SARS-COV-2 here in France, I was finally infected with the Omicron variant 10 days ago (community transmission, I don't know who I got it from or who I passed it on to - also I am fully vaccinated + booster with Pfizer), spent a week in isolation coughing and sneezing, and today (Sunday 30/01) is the first day I felt good enough to go out and walk a little bit and ride my bicycle.

I still have quite a few things to work out but in the coming days I should be able to find the time to sort the various pull requests from Pawel as well as my own accumulated changes and prepare a new release by next Sunday (6/02), if all goes well. I think I will name the next release "Omicron" - hehehehehehe, cough, cough. 

Thanks all for your patience! 

I think it's just down to luck with this thing. I spend a lot of time in engineering firms all over the UK, lots of a near misses but so far I have stayed free of it. I'm one door handle away from catching it.

[iannez]

mendip I ordered this:

https://it.aliexpress.com/item/1005001402512471.html?gatewayAdapt=glo2ita&spm=a2g0o.9042311.0.0.79314c4d96bWX1

bye,A.

[iannez]

My partner took Omicron many days ago.
But after three days of influence is now fine.
Tomorrow should have control, we hope it is negative otherwise another 10 days of quarantine.

Rested AndrewBNC   

Bye, A.

[mendip_discovery]

mendip I ordered this:

https://it.aliexpress.com/item/1005001402512471.html?gatewayAdapt=glo2ita&spm=a2g0o.9042311.0.0.79314c4d96bWX1

bye,A.

That is a OX and not a VX, one is controlled by temp and seen and the better, the other by Voltage and is not as good.

[iannez]

To try it goes more than good.
I connect to the potentiometer with the filtered PWM.
It should work properly, i hope eheheheh

Bye, A.

[felixd]

[NOTE]
There is Temperature sensor on STM32 uC. Why not to use it by default?

---

Andrew, have you seen that?
https://github.com/stm32duino/wiki/wiki/API#analog
https://github.com/stm32duino/STM32Examples/blob/main/examples/Peripherals/ADC/Internal_channels/Internal_channels.ino

Edit: My breadboard had weak connections. That was causing all power related problems....

Code: [Select]

Title Count Deviation Unit Description
Lat 18862 0.00004179 ° Position LTP Latitude
Lon 18862 0.00004728 ° Position LTP Longitude
Alt (HAE) 18861 7.143 m Position LTP Altitude (above ellipsoid)
Alt (MSL) 18861 7.143 m Position LTP Altitude (above mean sea level)
X 18861 3.668 m Position ECEF X
Y 18861 3.090 m Position ECEF Y
Z 18861 7.744 m Position ECEF Z


[thinkfat]

The temperature sensor on the STM32 is not worth a lot for general purpose use. You mainly need it for ADC offset calibration IIRC. It is measuring the die temperature of the MCU, as soon as it gets "something to do" the temperature will rise, but that's only a very local effect. There is of course an influence from ambient temperature, but if you want to correlate that with frequency deviation of the OCXO, it's not really suited for that. I've tried that and then decided to add a separate sensor near the OCXO.

[AndrewBCN]

I have tested the temperature sensor on the STM32F411CEU6 and it is practically unusable, even with averaging ("smoothing") algorithms. The first (internal) builds of the STM32 GPSDO firmware actually had the code for reading the temperature sensor, then I commented it out, then I deleted it entirely.

But you can easily try it: just check the code where I read the Vref, with a few small changes you can read the internal temperature sensor too.

[Johnny B Good]

[NOTE]
Adding 300uF on 3V3 + 200uF on 5V helped a lot:

• GPS readings are much more stable now
• 3V3 (Provided by BlackPill ) is much more stable (below +/- 0.01V)
• There is Temperature sensor on STM32 uC. Why not to use it by default?

---

Andrew, have you seen that?
https://github.com/stm32duino/wiki/wiki/API#analog
https://github.com/stm32duino/STM32Examples/blob/main/examples/Peripherals/ADC/Internal_channels/Internal_channels.ino

---

Going back to 3V3 caps, GPS readings are now way much more stable. Before I was hitting even 100m circle.

Code: [Select]

Title Count Deviation Unit Description
Lat 18862 0.00004179 ° Position LTP Latitude
Lon 18862 0.00004728 ° Position LTP Longitude
Alt (HAE) 18861 7.143 m Position LTP Altitude (above ellipsoid)
Alt (MSL) 18861 7.143 m Position LTP Altitude (above mean sea level)
X 18861 3.668 m Position ECEF X
Y 18861 3.090 m Position ECEF Y
Z 18861 7.744 m Position ECEF Z


 Looking at those deviation map captures, it looks as though the antenna doesn't have a direct view of the sky.

 As an example of just how accurate a positional accuracy you can get out a uBlox M8N (fake or not!), I've attached a set of screen photos (1st three from when I was using a fake M8N module way back in 2019 just before raising the mag mount active patch antenna another half meter to peep over the ridge tiles by a foot or so) and the last 7 screenshots were from after raising the antenna into 'clear air' whilst I was surveying in a freshly acquired M8T I'd installed into my MK II GPSDO.

[iannez]

Good evening everyone,
Today I finally arrived the VXCO from Alixepress
I calibrated it with a precise 10mhz on board potentiometer.


Fix Time 991ms
Uptime: 01:02:32
SATS: 8 HDOP: 1.29

Frequency Measurements Using 64-bit Counter:
64-Bit Counter: 37454988627
Frequency: 10000000 Hz
10s Frequency AVG: 10000000.0 Hz
100s Frequency AVG: 10000000.00 Hz
1,000s Frequency AVG: 10000000.004 Hz
10,000s Frequency AVG: 0.0000 Hz


Now I wanted to understand how to connect the quartz control pin to the micro PWM without using external DACs.

Apart from the KICAD scheme do you have advice on how to connect point-point pin resistor cap
 and with the names of the pin?

tnx all, A.

[Badwater]

... it is an OCXO 

... follow the schmatic: PB9 - low pass (R2+3, C4+5) - Vctl - Vadj
Jumper3 1-2 connected, 3 open

... do not forget to set the correct defines in the code: uncomment the PWM defines, comment the DAC defines

BTW: alternative display 

[AndrewBCN]

I like that display! Fabulous!