($40 GPSDO working well now, my mistake)

[tkamiya]

It is mounted up-right.  For some reason, when uploaded, it looks sideways.

I have no idea what is inside.....  I was curious but it is glued shut.  I doubt it is a kind that references earth as none of the metal screws touch the inside. 

[cdev]

Your camera must have attached the wrong metadata to the image, if you could rotate it and re-upload it that would be good.

Don't worry, if you can find a spot where your GPSDO can simply get three or more sats 90% of the time you should be more than okay as far as time and frequency accuracy. Even less than that will almost certainly work okay, once its gotten a decent start.

The beauty of a good GPSDO is that it should be able to maintain a good time and 10 MHz signal even with a fairly sketchy signal if it needs to, once its had an opportunity to do its survey and it knows its stationary.

[tkamiya]

I hope my house doesn't move....  I am in Florida USA where it is not that uncommon for sink hole to develop under your house!   

[cdev]

Everything on Earth is moving, and its complicated.

Lots of different things are going on. Where I live, there was a huge glacier which finished melting around 10,000 years ago and because land farther to my northwest is now going up, freed of all that weight, where I live is now apparently going down, just a little bit. And of course, 'sea level' is rising too, an arbitrary water level of the oceans, not meaning that mountains are getting shorter, just the water level is rising. And of course there is plate tectonics.

Sinkholes might be happening because of removal of a supply of water or pumping of water out of an aquifer? Ingress of fresh water likely keeps the ground up to some extent, even though that sounds a bit strange. If the ground is porous and the water that supported its volume was removed, some parts of it might get smaller. Just a guess.

I'm sorry, I don't know what i2c address the Packrat Arduino code wants for its LCD display because Ive never used that code personally.

[tkamiya]

I will probably end up writing my own code anyway, but I thought I'll give Pakrat a try.  It seem to be well received.

There is a location not too far from my house where a HUGE sink hole showed up in front of a house.  It swallowed a Porche that was parked there.  Florida is basically a mound of sand.  Pumping of ground water and torrential rain washes off sand and the result is a big hole.

Eventually, I'd like to have a small LCD on GPSDO to monitor and have a connection to a PC with Lady Heather.  That will be ideal.  Anyway, thinks for sharing information.  That university website was kind of interesting.  (I read Japanese)

[tkamiya]

Now, I have Lady Heather version 6.0x talking to it.

Other than yellow line is increasing, (OCXO temperature?), blue and green are horizontal lines.  Some column shows number but they are unchanged.  When "D mode" was 1, there was a red line in graph.  (I think Texas Pro mentioned this earlier?)  Is this supposed to be this way?

Edit:  Blue changed later on but green was still straight line.

[cdev]

You just turned it on. You have to let it run for quite a while. Have you done a survey?  You should ask texaspyro questions about his software because I have not used it with the TP.   I'm on Linux and although I have a controller its still caseless.

So I have mostly spoken to my GPSDO via the command line.

[mark03]

I am little concerned that we don't know what status 1 mean....  Mine flips between 0 and 1 fairly often.  If it means locking and synchronizing is compromised, it is unusable.  I read TexasPro commenting it probably doesn't mean fully unlocked.

I'm still concerned by this as well.  In LH I have a trace (green in the screen grab I attached on the previous page) which purports to show the OCXO EFC voltage.  We don't know if that is what it is, but it does stay relatively flat until one of these status 1/0 events happens, whereupon it makes a huge step and oscillates like crazy as it settles back down.  Whether this is holdover or something else, it looks pretty serious to me, but I don't have a second unit, nor a TICC, that I could use to evaluate the effect on the output.

I haven't built my board into an enclosure yet either.  I have an Extron dist amp for this purpose but mine is a slightly less tall version (ADA 6 300MX) with essentially no room on the back panel to place the GPS antenna connection.  (The rear PCB occupies the entire back panel---no room to sneak past it---and there aren't any unused portions of the PCB either.)  So that is a bit of a hassle...  I will get around to it, once some higher-priority projects settle down

[cdev]

The signal quality is much more important in the timing and surveying contexts than it is for locating oneself within the typical non-RTK GPS spec's level of accuracy.

One good way to analyze the quality of a signal from your antenna is to get any RTKlib-capable GPS (there are Ublox M8N copies on ebay for under $20 which work with it) and attempt to do either RTK with a stream of local, free base station data (such as a station on the CORS network of stationary receivers of known locations) if one is available or one of the several kinds of PPP. (accurate positioning using a single receiver over time to resolve the ambiguities)

Doing this one can get a fairly good idea of how good in terms of multipath any given GPS signal is. (so you are really testing the antenna/location combination)

Also, although LH's polar SNR display is not quite as much data as RTKlib with its chart of signals and residuals and cycle slips, its still a huge help, because one can see in one glance where in the sky the problems are. If a portion of the sky that would normally have arcs going across it has none or just a few sketchy ones after a day or so of capturing data, you'll know that your antenna could be moved - to get a better sky view, and should be.

For example, a tree or building blocking satellites can often be seen in the polar plot as an area without signals.

Ublox's-ucenter also can give you a similar body of info.

Despite the lack of manufacturer data, when one looks at the large number of settings and configuration parameters we already do know about, its a fairly rich set of info.

[pigrew]

When in state 1, I have noticed the reported control voltage to vary more than when in state 0. I have my antenna indoors, so only get 4 or 5 SV normally.  I don't know if it is the lack of good reception, or an issue with the control loop, that causes it to go into state 1. I don't have time to check the data presently, but you should be able to calculate the frequency deviation during state 1 by multiplying the control voltage shift by the OCXO gain.

[cdev]

Timing antennas have a pattern thats optimized for the higher satellites, (typically the ones with less multipath).

If one is put outdoors with a good sky view it should do well with the TP.

tkamiya, you definitely would have far better results if you put your antenna outdoors. Or even just moved it away from that beam to the side. Assuming the outdoors is behind and above where its now, that beam is between it and almost the whole sky in that direction.

Its a miracle that you're getting a fix at all with it like that.

[tkamiya]

I don't have a data to show you, but my current antenna works well feeding a pair of Nortel GPSTM.  I get full lock and decent SNR covering most of the sky.  I get a thin occluded area.  I can put up an outside antenna but didn't have a need up until now.  I might just do it for purpose of testing this board.

Once I put it in a case, I can run 24x7 and get a decent survey.  I don't want a heat source running unattended.

[cdev]

If you don't feel its safe by all means, wait until you do.

Get a good case for it. Its worth the time and energy putting it into a nice case. I would also either use a separate power supply or take care to not have the heat producing voltage regulators inside the case with the OCXO.

Their heatsinks if they need them should be mounted on the outside.

Take care to put it somewhere thats thermally stable. My GPSDO is not being used right now because i haven't put it in a permanent case yet.

I plan to use an Extron video distribution amp which I have which is ideal. It has a big case and lots of room which has additional cutouts which can support four more BNC connectors.  This particular video distribution amp also has a "TTL" channel which appears to work fine with the 1PPS. So I should end up with two devices in one case, the GPSDO and the VDA, which will give me an additional 9 10 MHz outputs and 3 1pps outputs that can either be switched on or left off.

You should post the temperatures its giving you. There are a range of different boards, if your temperatures are abnormal somebody here could probably tell you. Mine (with a Bliley NVG47A1282 are usually in the 40s-centigrade) but as the ambient temps can be higher with the oven still remaining functional, the oven's operation is likely over a pretty wide range of temps. Some people here have had temps that are more than 10 degrees higher.

[rbm]

Would you or someone else tell me what I2C address Pakrat code is expecting?  Not all LCD boards use same address and source is not available.

The Packrat software is expecting a PCF8574AT-based I2C display at 0x3F. A Sainsmart LCD will work; a Ywrobot display will not because it and many other 16 x 2 LCDs use the PCF8574T backpack instead. They listen at 0x27.  If you have an I2C LCD which you can swap the backpack for the right one, you should be good to go with Packrat SW.

[rbm]

I will probably end up writing my own code anyway, but I thought I'll give Pakrat a try.  It seem to be well received.

There is a location not too far from my house where a HUGE sink hole showed up in front of a house.  It swallowed a Porche that was parked there.  Florida is basically a mound of sand.  Pumping of ground water and torrential rain washes off sand and the result is a big hole.

Eventually, I'd like to have a small LCD on GPSDO to monitor and have a connection to a PC with Lady Heather.  That will be ideal.  Anyway, thinks for sharing information.  That university website was kind of interesting.  (I read Japanese)

I am working on modifying code that was published by Ben Hall, KD5BYB over on time-nuts.  He distributes through his web site. His code depends on a 20 x 4 LCD, rather than a 16 x 2 one.  I have the same aspirations as you, to have RS-232 connection for LH as well as an Arduino-based controller, and to build it into a box containing a distribution amp.

[tkamiya]

Thank you!  That is very helpful.

I have 20x4 but the case intend to use is 1U.  So I just ordered 2x16.

I'll probably roll my own knowing nothing out there fits my need exactly.  My plan is to make my GPSDO independent with minimal control and display to boot.  Then for the rest, use LadyHeather.  I am not sure now, if I want to include distribution amp.  Making it in a separate box will allow me to change just the timing device and not remake the whole thing.  I'll sit on this idea and see what else comes up.

We should keep in touch then. 

[tkamiya]

I decided to start with a minimalist system.

The TruePosition board and Aduino (the smallest one) on a raw PCB
Meanwell power supply and 16x2 LCD.  (the one on photo is NOT)

Antenna input
10Mhz output
1pps output
Enough coding in Aduino to generate "$PROCEED"
Serial port

This should be a nice development setup.

This case is some kind of telecom equipment and had a long haul interface in it.  All of that was stripped out.  I often visit surplus dealers and buy cases when re-usable ones show up.

[cdev]

Does anybody know the manufacturer and part number of the plugs that connect to the large two rowed connectors on the True Position and UCCM GPSDO boards?

In the pic below: I mean the two big multiple connectors on the left.

UCCM GPSDOs typically have one, the TruePosition board has two.

[cdev]

Found this while looking for videos about the GPSDO

[tkamiya]

I am a bit confused of existence of RS232C and TTL ports.  As I understand it, they are either-or proposition controlled by a jumper between pin 4 and 5.  When jumped, it enables RS232C.  When open, it enables TTL ports.

My initial plan was to use both simultaneously.  I wanted the unit to start up and run.  When needs arise, hook up a PC via RS232C.  I guess this is not possible.

Pin 6 is curious.  RX only??

[Miti]

Where do you buy one of these?

[tkamiya]

Some progress.....

Put one of my board into a metal case with its own power supply, and added an Arduino (for future use).  The case was some kind of telco equipment in its former role.

As is, it works nicely with Lady Heather but does not have an ability to work independently.  I have not decided how much stand-alone capability to build and how much UI to use.

[tkamiya]

OK, now that it has been running for a few days, it is starting to stabilize.

Since I have 3 of these boards, I wonder if it is possible to somehow take an average?  Of course not because there is such need, but because I might be able to complicate it unnecessary a bit more?

Right now, Arduino is there but not hooked up.  Lady Heather actually sends "$PROCEED".  I really don't feel like writing a software so I might keep it like this for a while.

Writing software - feels too much like work....

[cdev]

'Average' of what? Compared against what?

[cdev]

The usual suspects.

Where do you buy one of these?