($40 GPSDO working well now, my mistake)

[texaspyro]

I kept testing it from a cold start at lower and lower voltages. At 9.4V the unit starts no problem but the 10MHz waveform starts to become distorted. No issues at 9.5V. The digital portion works fine down into the 8V range but the OCXO is not working.

It's designed for 12V... USE IT!   It's rather stupid to expect it to work properly at such a low voltage.  It might start at 9.5V today but you know it's 0.1V away from failure.  How do you know how well it is working at such a bogus voltage?   OCXOs are temperamental and fussy little beasties.

[don.r]

I kept testing it from a cold start at lower and lower voltages. At 9.4V the unit starts no problem but the 10MHz waveform starts to become distorted. No issues at 9.5V. The digital portion works fine down into the 8V range but the OCXO is not working.

It's designed for 12V... USE IT!   It's rather stupid to expect it to work properly at such a low voltage.  It might start at 9.5V today but you know it's 0.1V away from failure.  How do you know how well it is working at such a bogus voltage?   OCXOs are temperamental and fussy little beasties.

Actually, the GPS module is designed for 13-15V supply. The OCXO actually has a 5V working voltage. Don't worry, I'll measure it over a few days at different voltages to find the lowest stable point. There appears to be a buck regulator at the input which supplies the board with its power so input voltage is not too important, as long as it works and is stable over time. I also have to test inrush current at turn on to make sure the PSU can handle that.

[texaspyro]

The system these were pulled out of run it at 12V... not 13V... not 15V... but 12V  Ignore what the Packrat docs say.

Just because it powers up does not mean it is working properly.  You would need to measure phase noise and ADEVs across a reasonable temperature range to verify it is OK... and something tells me you don't have the equipment to do that.

[don.r]

I have access to a spectrum analyzer and enviro chamber but that's a waste of time. There is a buck regulator and an LDO precision linear regulator between the supply input and the OCXO. Input voltage is a non-issue.

[don.r]

Worked great with a sub-$10 GPS antenna from China. Picked up half a dozen satellites in the space of a few minutes. The only worry is that the OCXO control voltage is already sitting at 4.00V and has a max of 5V. Hopefully she doesn't drift too much anymore. Looking at the specs of the Bliley, the power requirements are 5V +- 5% and the freq error due to supply is less than 1X10^-9 per %. Given that the LM1965 has a load regulation of around 0.1% should be no problem.

[cdev]

The performance of a GPS-DO should be substantially better than that of the base OCXO its disciplining..

My main concerns are - 1. is the 12v voltage optimal and 2. will putting it in a closed or semi closed case which may involve additional voltage regulation perturb the thermal characteristics in a negative manner.

[don.r]

The performance of a GPS-DO should be substantially better than that of the base OCXO its disciplining..

My main concerns are - 1. is the 12v voltage optimal and 2. will putting it in a closed or semi closed case which may involve additional voltage regulation perturb the thermal characteristics in a negative manner.

I would check out the specs of the two regulators involved with the OCXO. The first one is an LT1765 set to provide 5.78V to the next regulator. The input voltage goes anywhere up to the low 20V range, which is probably why we are seeing people destroy their GPS units with more than 22V. Here is the datasheet: http://cds.linear.com/docs/en/datasheet/1765fd.pdf

The next regulator is the LT1963 set to provide 5V to the OCXO from the 5.78V given to it from the LT1765. This output does not change regardless of input voltage to the unit. The LT1963 is very stable (0.1% regulation) and has only 40uV of noise. The OCXO has further power filtering inside (per Bliley website). Here is the LT1963 datasheet: http://cds.linear.com/docs/en/datasheet/1963fc.pdf

As for temp, look at the OCXO datasheet here: http://www.bliley.com/pdfs/Bliley_N47X_NV47X_OCXO.pdf and check out the best, good and easy figures for 10MHz for the various ranges. Even at 70C (remember the oven is temp controlled) the easy figure is +- 3X10^-8 The OCXO can take a +-5% variation in the power supply. I would not worry about it.

[don.r]

A few questions for the more experienced:

1. My survey time as returned by the $SURVEY response never changes from the set time. Is that supposed to happen?
   Works now. Just needed to wait for the right conditions.
2. Anyone know what state 19 is? It appears to be some form of wait state.
3. Mine also has a 10 year lag on the time. Can I simply add a 10 yr fudge factor to the time?
4. Will I experience much signal loss if I add an antenna extension cable?
5. What is the command to save location, satellites, training, etc? or is saving done by default?

WiFi works a treat, at least for monitor purposes. ESP32 is an incredible little MCU for $7. Webserver and GPS control code takes up about 640kb so far. About 3.5MB to spare...

[pigrew]

A few questions for the more experienced:

1. My survey time as returned by the $SURVEY response never changes from the set time. Is that supposed to happen?
   Works now. Just needed to wait for the right conditions.
2. Anyone know what state 19 is? It appears to be some form of wait state.

It's part of a series of wait states 19->20->21->22, that I call "Wait B". I think this sequence happens when entering state 1 (not completely sure). There is another series of wait states 15->16->17->18 that happens while entering state 0.

Look at the start of https://github.com/pigrew/trueposctrl/blob/master/trueposctrl/Src/displayTask.c to see what I've reasoned, based on guess and check.

3. Mine also has a 10 year lag on the time. Can I simply add a 10 yr fudge factor to the time?

That's what I did.

4. Will I experience much signal loss if I add an antenna extension cable?

It depends on what sort of cable you add. I saw that uBlox suggests using 75-ohm cables. The mismatch isn't too bad for reasonable cable lengths. If super-long, you'd need to add another line-amplifier (preferably at the start or middle of the cable).

5. What is the command to save location, satellites, training, etc? or is saving done by default?

It seems that my unit saves them automatically. texaspyro had written that "$UPDATE FLASH" would store something, though I've never used it. Use at your own risk.

WiFi works a treat, at least for monitor purposes. ESP32 is an incredible little MCU for $7. Webserver and GPS control code takes up about 640kb so far. About 3.5MB to spare...

Good to hear. I have mixed feelings about the ESP32. I'm disappointed by its ESP-IDF, ADC linearity, and documentation.

Would you post your code somewhere? I'm curious about how it compares against my control program for STM32. Is your goal to run a NTP server?

[don.r]

Thanks for the reply. I'll probably need no more than 6 feet of extra cable. I have yet to play with the gain so it may not be an issue. Currently finding 7 satellites with the antenna sitting on my window sill but lock keeps drifting in and out. I'm not going to risk anything by running the flash command yet. I'll post some code when its ready but it will be ugly, very ugly.

I'm OK with the IDF. I use Eclipse so much is hidden. C++ integration is lacking so I'm using C. Documentation is a rather sore point but I'm using Kolban's book as a guide. Its only been out a year so hopefully time will improve the software and docs. I have yet to characterize RF emissions so that may also be an issue. Haven't looked at the linearity of the ADC yet. Is it consistent enough to be corrected in software? A dual core 240MHz 4GB WiFi and BT enabled ucontroller with OTA update capability for less than $7 (less if you just need the module) is still impressive to me. For the cheap stuff there is always the STM32 or PICs for the real cheap stuff. 

Not planning on a time server, maybe?, but just don't want to have to hook a computer up to control and check things since the unit will sit high up on a shelf and somewhat inaccessible. I'll bring the RS232 out so I can hook up a PC if need be. My lab is in an outbuilding from my house so its also handy to check things out while in the house.... plus its excellent experience programming the ESP32.

[don.r]

Here is a sample screen shot of the info window. I have a separate window for commands.

[don.r]

Code now on github: https://github.com/don-r/TruePosition-GPS

[cdev]

If I use appropriate 'magnetics' is it possible to use a compatible ESP32 device or at least access its web server via TCP/IP via conventional wired Ethernet (and not via Wifi?)

(Or, maybe it could be ported to a Linux SBC that did have Ethernet?)

[pigrew]

Assuming you have + use the magnetics is it possible to use the ESP32 device UART and access the web server via conventional wired Ethernet and not via Wifi?

I would like to reduce my own use of wifi....  I know I can set up a closed network, just for gadgets but still.

Yeah, it'll work. It's the same principle I used to connect the UART to USB (just the coding is trickier since you'd want to support multiple streams).

You'll have to pair the ESP32 with a PHY and transformer.

You likely can find a prototyping board which already has the PHY and Ethernet connector on it, which could be connected to ESP32.

[cdev]

That's good to know.

[don.r]

I added the WebServer code. If you want to use ethernet, you should just have to change the ws_init and the wifi_event_handler routines. I tried to use SSL at first but its so slow (like 4 - 5 seconds to respond) that I just ended up with plain unsecured sockets. Not an issue for me as no one lives near enough to me to tap into my wifi but for others SSL may be required. All the code uses FreeRTOS as the OS which comes with ESP-IDF.

[GigaJoe]

arduino, use arduino mini to send $PROCEED
need a library alt soft serial ...

Code: [Select]


#include "AltSoftSerial.h"

AltSoftSerial altSerial;
// TX:  PIN - 9 (connected to pin 2 GPS)
// RX:  PIN - 8 (connected to pin 4 GPS)


void setup()
 {
  altSerial.begin(9600);
 }


void loop()
{
  char buf=0;
  char Nbuf=0;
  char c[100]=" ";
  int i=0;
  int c_r=0;
   
  Nbuf = altSerial.available();
  if( Nbuf != 0)
  {
    while(i < Nbuf)
    {
     c[i]=altSerial.read();
     delay(1);
     i++;
    }
    c_r=1;
  }
    if( c_r != 0)
     {
       if( c[6] = "R" )
        {
          altSerial.write("$PROCEED");
        }
     }
}


[cdev]

That's simple!

Must be the simplest interface out there

arduino, use arduino mini to send $PROCEED
need a library alt soft serial ...

Code: [Select]


#include "AltSoftSerial.h"

AltSoftSerial altSerial;
// TX:  PIN - 9 (connected to pin 2 GPS)
// RX:  PIN - 8 (connected to pin 4 GPS)


void setup()
 {
  altSerial.begin(9600);
 }


void loop()
{
  char buf=0;
  char Nbuf=0;
  char c[100]=" ";
  int i=0;
  int c_r=0;
   
  Nbuf = altSerial.available();
  if( Nbuf != 0)
  {
    while(i < Nbuf)
    {
     c[i]=altSerial.read();
     delay(1);
     i++;
    }
    c_r=1;
  }
    if( c_r != 0)
     {
       if( c[6] = "R" )
        {
          altSerial.write("$PROCEED");
        }
     }
}


[GigaJoe]

well , I need to start it, if no any fenzy thingi needs, then the code may acceptable; run it as standalone ....
will add: 2-line display, +6 pins ; + 2 LED    +2 pins
and buttons will do an analog way, each button send a specific voltage to ADC port, when you measure it you know the button ...

[cdev]

Great!


> if no any fenzy thingi needs, then the code may acceptable; run it as standalone ....

Yes, that sounds very good. SoftSerial=no USB-UART chip needed.... So any MCU that runs Arduino code is okay.

>"will add: 2-line display, +6 pins ; + 2 LED    +2 pins "

Just displaying the stream as it emerges? That will work fine!

[GigaJoe]

Alt soft serial works well ( seems a bit differ comparing to original soft-serial) , 2 com ports - very easy to debug, FW upload and so ...  I dont have MEGA that has more HW ports ...

[Ernie]

After power-up it starts to sending  via the RS-232 port the  $GETVER msg almost for ever  and unable to stop and make the $PROCEED  msg to enter..... is there any way to sop and enter some other msg..
like $SURVEY and so on??? After several hrs  the little green LED on the motherboard does not lit.
Please advise

[UglySteve]

After power-up it starts to sending  via the RS-232 port the  $GETVER msg almost for ever  and unable to stop and make the $PROCEED  msg to enter..... is there any way to sop and enter some other msg..
like $SURVEY and so on??? After several hrs  the little green LED on the motherboard does not lit.
Please advise

When I had this problem, I had to change some settings in Hyperterminal. I believe these settings fixed it for me.
ASCII set up." Local Echo" on, "append line feeds to incoming line ends" checked, "Wrap lines that exceed terminal width" checked.
Steve

[Ernie]

Hi Steve,
Thanks a lot for the quick reply ...I had the same Hyperterminal set-up but unable to stop the incoming $GETVER msg..... what is the keyboard entry to stop the  $GETVER  msg...?pleas note I am not a PC/software guru so need detailed info.....
 Thanks.

[UglySteve]

Hi Steve,
Thanks a lot for the quick reply ...I had the same Hyperterminal set-up but unable to stop the incoming $GETVER msg..... what is the keyboard entry to stop the  $GETVER  msg...?pleas note I am not a PC/software guru so need detailed info.....
 Thanks.

As far as I know, $PROCEED should do it. I think that if you type it, even if it doesn't show up on your screen, it should work. Just type over the incoming data.  Do you have the serial data port wired correctly? 

I don't have the rs-232 wired to mine at the moment, I used the Packrat  software on a micro controller, and made it a stand alone unit with a distribution amp, made from a modified RGB video distribution amp.  I need to cut a hole in the box to mount a db9 connector for the rs-232.
Steve