10MHz GPSDO by "BH3SAP"

[2A03]

Hi, fredo_

I tried v0.1.11 and it seems fine.
Thank you so much.

[daddygo]

Two LED versions PCB with inv-buffer and sized above the rotarry...

[ME]

I am using ver 1.9 , It displays the correct time on power on but the time never changes thereafter untill i power cycle the unit, Is that normal ?, cheers.

[fredo_]

Hi @ME, no it's not normal, time should be updated every second.
Maybe uart speed is incorrect ? Is it set to 9600 ?

[daddygo]

Hi @ME,

Well, there's definitely something wrong here, the clock must work continuously, just like a normal desktop clock

-how many SATs does the unit see?


Quite a little-lot has happened since your version, try the latest FW first
https://github.com/fredzo/gpsdo-fw/releases/tag/v0.1.11

[stspies]

I am using ver 1.9 , It displays the correct time on power on but the time never changes thereafter untill i power cycle the unit, Is that normal ?, cheers.

Hi,
I added a FTDI232 to the STM connecting GND, RX, TX and 1PPS to DCD. Baudrate is set to 9600.
This is connected to my server running gpsd and chrony. It all works very well - except sometimes.

Sometimes it happens that for unknown reasons the clock stops counting. I have to send a "gpsctl -n" because for unknown reasons the gps module switched to binary mode. Maybe same situation for you?

[stspies]

A great project and the new firmware is awesome.

My plan is to add a 10MHz SMA connector to support my HackRF One (0-3,3V), a programming connector (maybe a 3,5mm TRRS) and a 50 Ohm 1PPS output.
Als an USB-C-Serial connector with a FTDI232 (my prefered RS232 chip):
As well as LEDs for GPS/GPSDO lock and 1PPS - similar to the suggestions from daddygo and a little switch to torn off the display light.

These are the ideas I have in mind and at the moment I'm struggling a bit with the FTDI, which sometime doesn't talk properly to the STM - but haven't gone into details yet.

BTW mine also has the chinese GPS module.

Steffen

[daddygo]

A great project and the new firmware is awesome.

My plan is to add a 10MHz SMA connector to support my HackRF One (0-3,3V), a programming connector (maybe a 3,5mm TRRS) and a 50 Ohm 1PPS output.
Als an USB-C-Serial connector with a FTDI232 (my prefered RS232 chip):
As well as LEDs for GPS/GPSDO lock and 1PPS - similar to the suggestions from daddygo and a little switch to torn off the display light.

These are the ideas I have in mind and at the moment I'm struggling a bit with the FTDI, which sometime doesn't talk properly to the STM - but haven't gone into details yet.

BTW mine also has the chinese GPS module.


Hi Steffen, @


Steffen

Hi Steffen, @stspies

these improvements are pretty much done, if you read back in the thread...

-there is a 50ohm 1PPS impedance matching buffer PCB + USB to TTL also on this PCB , sandwich mounted in the original location with GPS modul (any GPS module can be on top)
-there is a new NEO-M9N based GPS module with LTE notch filter and external LNA power supply (5V) + 5mAh battery for GSP hot-start
-and has front panel-mounted LOCK LEDs PCB

(on the 50ohm buffer PCB you can choose between the MCU generated PPS signal and the GPS generated PPS signal with a mini DIP switch )

only Lock LED panel and the new GPS unit is in PCB&PCBA production (JLCPCB) soon and I can measure and test them

if you listen to me  , forget the FTDI, WCH chip much cheaper with minimum external parts, I have used CH340K (built-in diode protection and many good features, built-in quartz!!!!, etc.....)
I attach the datasheet of the CH340 read it and you will see...

[stspies]

these improvements are pretty much done, if you read back in the thread...

-there is a 50ohm 1PPS impedance matching buffer PCB + USB to TTL also on this PCB , sandwich mounted in the original location with GPS modul (any GPS module can be on top)
-there is a new NEO-M9N based GPS module with LTE notch filter and external LNA power supply (5V) + 5mAh battery for GSP hot-start
-and has front panel-mounted LOCK LEDs PCB

(on the 50ohm buffer PCB you can choose between the MCU generated PPS signal and the GPS generated PPS signal with a mini DIP switch )

only Lock LED panel and the new GPS unit is in PCB&PCBA production (JLCPCB) soon and I can measure and test them

if you listen to me  , forget the FTDI, WCH chip much cheaper with minimum external parts, I have used CH340K (built-in diode protection and many good features, built-in quartz!!!!, etc.....)
I attach the datasheet of the CH340 read it and you will see...

Hi @daddygo,
I've read the whole thread already a couple of times because there are many useful informations and I didn't want to say that I want to develop all of it by my own.
I simply wanted to point out what my plans of modifications are. 

Looking forward when the PCBs are available.

FTDI vs CH340... you know it is like Intel vs AMD, VW vs Opel, Microsoft vs Apple... pro and cons on both sides and everyone has made different experiences.
I'm fine with the CH340, but for me the FTDI worked better in the past.

Steffen

[daddygo]

these improvements are pretty much done, if you read back in the thread...

-there is a 50ohm 1PPS impedance matching buffer PCB + USB to TTL also on this PCB , sandwich mounted in the original location with GPS modul (any GPS module can be on top)
-there is a new NEO-M9N based GPS module with LTE notch filter and external LNA power supply (5V) + 5mAh battery for GSP hot-start
-and has front panel-mounted LOCK LEDs PCB

(on the 50ohm buffer PCB you can choose between the MCU generated PPS signal and the GPS generated PPS signal with a mini DIP switch )

only Lock LED panel and the new GPS unit is in PCB&PCBA production (JLCPCB) soon and I can measure and test them

if you listen to me  , forget the FTDI, WCH chip much cheaper with minimum external parts, I have used CH340K (built-in diode protection and many good features, built-in quartz!!!!, etc.....)
I attach the datasheet of the CH340 read it and you will see...

Hi @daddygo,
I've read the whole thread already a couple of times because there are many useful informations and I didn't want to say that I want to develop all of it by my own.
I simply wanted to point out what my plans of modifications are. 

Looking forward when the PCBs are available.

FTDI vs CH340... you know it is like Intel vs AMD, VW vs Opel, Microsoft vs Apple... pro and cons on both sides and everyone has made different experiences.
I'm fine with the CH340, but for me the FTDI worked better in the past.

Steffen

Hi Steffen,

I immediately thought you'd read the whole thread, just thought I'd refresh a bit here on page 10 as well 
what's certain is that we were thinking about almost identical improvements, which is good because it filters out the real gaps to be improved, and in addition it is a force of comminity dev.

which I can assure you, although I understand what you mean (Intel / AMD, etc..), I'm also a fan of FTDI for many of my projects I use it exclusively because it has built-in win/linux driver support, but for a simple development like this, for me the CH340 was more reasonable

the CH340K is worth a try and gain experience, despite being full Chinese it is very thoughtful and well designed - well, works when pulled out of the box

I'll report back soon with a fully built full modded HW version, and let's not forget to thank the FW development: @fredo_ / @2A03 

[ME]

Is there any difference with this unit with the lock led indicator led and rs 232 output compaired to an off the shellf BG7TBL GSPDO apart from being 75% cheaper?.

[stspies]

Hi all, hi fredo_

before @fredo_ you are closing the case, I do have two suggestions, but not without mentioning, that it already is a great project and really appreciate your effort.
My two suggestions aren't anything critical or in any way neccessary - just nice to have.

- locator: as a HAM and because I like to use the device outdoors, too, I would like to know the locator I'm in. ("IARU locator" or "Maidenhead Locator System")
- date format (another one): dd.mm.yy or yy-mm-dd

But again it all works great as it is and I'm very thankful for it!

Steffen

[bingo600]

Could this be a suitable GPSDO BP replacement board ?
https://www.eevblog.com/forum/microcontrollers/new-bluepill-on-steroids-stm32h503cbt6-evt/

W. 250MHz clock it should offer more "PPS capture resolution" , and w. dual PLL's it should enable the use of both PLL1 250MHz timers and PLL2 240Mhz (/5 == 48MHz) USB Clock.
Even when clocked with a 10MHz HSE clock.

I haven't investigated  how the board should be modified, in order to receive the 10MHz OCXO Clock, maybe someone more experienced w. STM & Hardware could have a look.

Regarding porting the sw to the new board :
Unfortunately i'm not experienced enough to do the migration, especially USB is a "unknown blackhole" to me.

But if it's an easy route hardware wise, it might be a nice upgrade.


/Bingo

[daddygo]

Could this be a suitable GPSDO BP replacement board ?
https://www.eevblog.com/forum/microcontrollers/new-bluepill-on-steroids-stm32h503cbt6-evt/

W. 250MHz clock it should offer more "PPS capture resolution" , and w. dual PLL's it should enable the use of both PLL1 250MHz timers and PLL2 240Mhz (/5 == 48MHz) USB Clock.
Even when clocked with a 10MHz HSE clock.

I haven't investigated  how the board should be modified, in order to receive the 10MHz OCXO Clock, maybe someone more experienced w. STM & Hardware could have a look.

Regarding porting the sw to the new board :
Unfortunately i'm not experienced enough to do the migration, especially USB is a "unknown blackhole" to me.

But if it's an easy route hardware wise, it might be a nice upgrade.


/Bingo

Hi @ bingo600

Well, then I'm not the only one who's intrigued by this question
(mentioned somewhere earlier as an NTP server + GPSDO mix - better MCU performance is a must for NTP)

I've gone further, many STM32 would be suitable, but to avoid modifying the mainboard, I've ended up like you - STM32H503CBT6 /w BP PIN comp.

plus I would like to dev. an ethernet connection based on a WIZnet W5500 chip, or use the following module:
https://docs.wiznet.io/Product/ioModule/wiz850io

-this would be the next development, if @fredo_ he would have the time and he has the will to do so...

YES it's long gone beyond upgrading the original GPSDO, but it's also a useful pastime, everyone does it when they have the time 


BTW:
Meanwhile here is the LED(s) PCB with Molex PICOBLADE connector - I will not modify this any further


EDIT1:

I had an abandoned KiCAD project with this STM32H503CBT6, I checked it, it looks like I can make it 100% PIN compatible with BP...
-VDDA and PB2 issue resolved by default on this (VDDA - LC-filter)
-The two Vcap PINs are also correct and are connected to a caps not to breakout
-I'm still thinking about the WIZnet chip with a direct pin header mounted above the board optionally.

it's not BP compatible, but there are good things here
https://github.com/openedv/M48Z-M33-Mini-Board-STM32H503/tree/master/libraries/Drivers

It would be interesting to push this through, if you feel like it  @fredo_   ()

[stspies]

V1.12: thank you, @fredo_
works like a charm!

thanks a lot for the little details


Steffen

[fredo_]

Yes @stspies, v0.1.12 has been released :-): https://github.com/fredzo/gpsdo-fw/releases/tag/v0.1.12
Changes are:
- Added GPS decimal coordinates in GPS menu
- Added IARU Locator in GPS menu
- Added new date formats as requested by @stspies (dd.mm.yy and yy-mm-dd)

[stspies]

Am I the only one who‘s clock is one second late?

[fredo_]

Am I the only one who‘s clock is one second late?

Good find @stspies !
This is due to the fact that there is a slight delay between PPS pulse and GGA frame reception.
The firmware only updates the time display once every second, when the PPS pulse is received, actually showing the time received in GGA frame almost one second ago.
To preserve sync between PPS pulse and display update, I fixed that by adding one second to the time value received in the last GGA frame.
The fix is available in v0.1.13 : https://github.com/fredzo/gpsdo-fw/releases/tag/v0.1.13
Changes :
- Added one second to displayed time to compensate delay between PPS pulse and GGA frame from the GPS module.

[ME]

I think this now must have rabbit hole status! lol.

[bingo600]

Thank you 

Code: [Select]

Memory region         Used Size  Region Size  %age Used
             RAM:       20168 B        20 KB     98.48%
           FLASH:       59580 B        64 KB     90.91%
[100%] Built target gpsdo

And my GPS now has status: With local Courier

 

[ME]

I just flashed the latest fw, The time display is stuck at 19:40:36 when my time here is 10.53am,is it a bug?

[stspies]

if you turn it off/on, does it work again?

I experienced something similar, when the device is connected to my computer running GPSD.
For somereasons sometimes it switches to binary mode.
gpsctl -n does the magic and it works again.

Maybe same on your side?

Steffen

[bingo600]

Well got my new toy today 
Ended up at $58 w. coupons.

Has been opened & "fredo_ 'ised" 




Erasing

Code: [Select]

$:BH3SAP-bluepill-china-gpsdo/git/gpsdo-fw$ st-flash erase
st-flash 1.7.0
2025-04-04T12:26:10 INFO common.c: F1xx Medium-density: 20 KiB SRAM, 64 KiB flash in at least 1 KiB pages.
Mass erasing

Programming

Code: [Select]

$:BH3SAP-bluepill-china-gpsdo/git/gpsdo-fw$ st-flash --flash=0x20000 write gpsdo.bin 0x8000000
st-flash 1.7.0
2025-04-04T12:27:03 INFO common.c: F1xx Medium-density: 20 KiB SRAM, 64 KiB flash in at least 1 KiB pages.
Forcing flash size: --flash=0x00020000
file gpsdo.bin md5 checksum: da1ba0ac1ec7e104b2733aa421c1c9d, stlink checksum: 0x0058ce02
2025-04-04T12:27:03 INFO common.c: Attempting to write 59580 (0xe8bc) bytes to stm32 address: 134217728 (0x8000000)
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08000000 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08000400 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08000800 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08000c00 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08001000 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08001400 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08001800 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08001c00 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08002000 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08002400 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08002800 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08002c00 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08003000 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08003400 erased
2025-04-04T12:27:03 INFO common.c: Flash page at addr: 0x08003800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08003c00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08004000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08004400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08004800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08004c00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08005000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08005400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08005800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08005c00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08006000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08006400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08006800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08006c00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08007000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08007400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08007800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08007c00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08008000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08008400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08008800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08008c00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08009000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08009400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08009800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x08009c00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800a000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800a400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800a800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800ac00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800b000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800b400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800b800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800bc00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800c000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800c400 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800c800 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800cc00 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800d000 erased
2025-04-04T12:27:04 INFO common.c: Flash page at addr: 0x0800d400 erased
2025-04-04T12:27:05 INFO common.c: Flash page at addr: 0x0800d800 erased
2025-04-04T12:27:05 INFO common.c: Flash page at addr: 0x0800dc00 erased
2025-04-04T12:27:05 INFO common.c: Flash page at addr: 0x0800e000 erased
2025-04-04T12:27:05 INFO common.c: Flash page at addr: 0x0800e400 erased
2025-04-04T12:27:05 INFO common.c: Flash page at addr: 0x0800e800 erased
2025-04-04T12:27:05 INFO common.c: Finished erasing 59 pages of 1024 (0x400) bytes
2025-04-04T12:27:05 INFO common.c: Starting Flash write for VL/F0/F3/F1_XL
2025-04-04T12:27:05 INFO flash_loader.c: Successfully loaded flash loader in sram
2025-04-04T12:27:05 INFO flash_loader.c: Clear DFSR
2025-04-04T12:27:05 INFO flash_loader.c: Clear CFSR
2025-04-04T12:27:05 INFO flash_loader.c: Clear HFSR
 59/ 59 pages written
2025-04-04T12:27:08 INFO common.c: Starting verification of write complete
2025-04-04T12:27:10 INFO common.c: Flash written and verified! jolly good!
$:BH3SAP-bluepill-china-gpsdo/git/gpsdo-fw$


Now waiting for the OCXO to settle, while investigating the menus & how2 mount leds & serial nicely on the front/back plate

So far so good
Tbolt as ref on the counter - 10s sample time - 50R termination.
A bit of statisctics from Power On.



Thanx goes to all involved 

Edit:
This is not a timenuts tool (FLL) , but maybe a "good enough" for hobby product.
And the thought of using fredo_'s code (minimized) as a standalone pwm adjuster is nice.


.

[bingo600]

Is anyone running this box off a "Switcher module" @7.5v ??
If yes, can you post some info about the module.

We're "burning" 12-5 = 7v in the 7805CV .... 7*0.4 ~ 3w in heat.
Afaik 7.5..8v should drive it just as well , lowering the energy waste.
Well some might be "saved in OCXO heating, but .....

[ME]

now at 15:33 the time reads 00:32 after power cycling!.