Does anyone recognize this board, is it a gpsdo?

[grumpydoc]

For the gps connector would it be a smb female?

Yes, it's SMB male on the board.

[edpalmer42]

Looks like the seller found some more units.  Since you guys got the GPSDO, I decided to take a small chance and bought one.

If I had to guess why he said that he was selling the small board in the first picture, I'd guess it was more of a language issue than anything.  He might have been saying that the value was for the OCXO and the board with the GPS receiver just happened to be there.

Once this thing has been sorted out and is running good, does anyone have any good equipment to confirm it's performance?

[Vgkid]

Thanks for the reply Grumpy, I will pick up a sma to smb pigtail.

[grumpydoc]

It's quiet here - I hope you guys are having fun with your boards

GPS time of day is available on the main connector (and top of PCB) 4800baud 8N1 LV TTL data levels, one update per second.

Has anyone got anywhere with the RS232 interface?

[grumpydoc]

Gosh it is quiet

It would be great to evaluate this board, i.e. holdover performance. It would take probably a couple of days of measurements to see if the board output drift is just what OCXO specs say or if they implement predictive control.

To an extent the problem is that you need something more stable/accurate to check against - however I have it running against my rubidium unit at the moment.

Currently my Racal counter is showing between 0-1mHz (yes, I do mean small "m") difference in frequency between the FE5680A and the NEC board, very similar to my Rapco 1804. That is 0.1ppb (1 x 10^-10) and is right at the limit of what I can measure directly with the counter.

So all I can say is that the short term variation over the 20second gate time is 1 x 10^-10 or better and the absolute frequency accuracy is also 1 x 10^-10 or better.

Quick update: switching to an indirect measure (observing the relative change in phase over time) gives me a figure of about 3x10^-11 for the absolute difference in frequency between the NEC board and the FE-5680A, again similar to the Rapco - at this point I do not know which device is "out", just that it looks as though all my frequency standards agree within a few parts in 10¹¹.

In any case the short term variance is going to dominated by the OCXO - pretty much anything less than 100 seconds - sadly the datasheet does not seem to give the figures .

[lukier]

Gosh it is quiet
To an extent the problem is that you need something more stable/accurate to check against - however I have it running against my rubidium unit at the moment.

Currently my Racal counter is showing between 0-1mHz (yes, I do mean small "m") difference in frequency between the FE5680A and the NEC board, very similar to my Rapco 1804. That is 0.1ppb (1 x 10^-10) and is right at the limit of what I can measure directly with the counter.

So all I can say is that the short term variation over the 20second gate time is 1 x 10^-10 or better and the absolute frequency accuracy is also 1 x 10^-10 or better.

That's great news. I'll have decent frequency standard on the cheap

I don't know if I had figured that out correctly, but what thought was to see how NEC board without GPS compares to another GPS constantly synchronized reference. This way we would see how MCU+FPGA controls the OCXO when there is no GPS. Then after some time we would see the difference between them (e.g. Figure 12 here http://tf.nist.gov/general/pdf/2297.pdf).  Would that be just OCXO datasheet aging rate or something rather clever.

Thanks for figuring out GPS time of day UART. My experiments with this board are hindered by the fact that in my lab I don't have GPS reception and at my desk I do, but I don't have any tools

[Deathwish]

Someone reads the forum , the price of that ocxo has shot up in price literaly overnight to almost twice what it was before.

http://www.ebay.co.uk/itm/111709755330?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

[grumpydoc]

Someone reads the forum , the price of that ocxo has shot up in price literaly overnight to almost twice what it was before.

http://www.ebay.co.uk/itm/111709755330?_trksid=p2055119.m1438.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Well, 1.5x - he was asking $40, now $60

The number available seems to have gone from 10 -> 5 as well, without selling any on ebay.

[Velund]

Looks like it is a module from some NEC microwave data link or maybe WiMax base station.

Unfortunately, I do not know anyone who service this equipment to ask...

[edpalmer42]

Looks like it is a module from some NEC microwave data link or maybe WiMax base station.

Unfortunately, I do not know anyone who service this equipment to ask...

Yes, it appears to be from their Pasolink line.  That's the only place I found part numbers of the form NWA-05xxxx.  Unfortunately, I wasn't able to find any detailed manuals for that equipment.

[Velund]

Now that I have an antenna connected X8 initially flashes once per second missing every 10th second. Then both flash with X8 continuing to miss every 10th beat. I assume that means a basic GPS lock. After a while X8 stopped missing the 10th beats so I presume that means a better quality lock.

After a couple of hours running with a GPS antenna hooked up X7/8 are flashing in tandem at 1PPS and X4 is flashing rapidly - at this point I'd expect the oscillator control to maybe have a coarse lock so we will see if anything changes over the next 24 hours.
......

Well, after another hour or so X4 went steady and X5 starts flashing at 0.25Hz, I guess that means it locked - at least to some degree. At that point I turned the board over to look at the outputs and it looks like IC29 and it's associated fan-out buffer carry the 10MHZ and IC30 has a 1 PPS signal. Sadly turning the board pulled my dodgy power connection off so that's reverse engineering finished for now.

It would be interesting to determine meaning of LEDs...

I do not see any backup battery on the board, so every time GPS should begin from cold start. So, basic lock and valid time, then self-survey to precisely determine coordinates... If I remember correctly, there should be something like "dashboard" in uBlox software, so it is possible to hook input of RS232 buffer or USB-serial converter to TXD of module and observe process, noting LED state changes.

Then, some LED may reflect T-RAIM status...

[Vgkid]

My module came today i should have ordered that adapter sooner than today.

[Vgkid]

Since I'm waiting on my connector, is there a pin for the 12v input power. Or should i just apply it to the ocxo?

[Vgkid]

Sorry for the double post:::
smb female to sma female came today...
It works

[grumpydoc]

Since I'm waiting on my connector, is there a pin for the 12v input power. Or should i just apply it to the ocxo?

The power  is on the bottom connector, there's a bit of decoupling/LC filtering but it does go straight to the OCXO 12V pin.

[lukier]

I did some more reverse engineering on this board today.

I traced most signals on the main connector.

Pin	Signal	Comment
1	12V
2	12V
3	12V
4	12V
5	GND
6	GND
7	GND
8	GND
9	GND
10	GND
11	MAX3243 DIN1
12	PPS (3.3V)
13	MAX3243 ROUT1
14	GND
15	GND
16	IC28 Q1P	LVDS PPS
17	MAX3243 DIN2
18	IC28 Q1N	LVDS PPS
19	MAX3243 ROUT2
20	GND
21	GND
22	IC28 Q2P	LVDS PPS
23	MAX3243 DIN3
24	IC28 Q2N	LVDS PPS
25	MAX3243 ROUT3
26	GND
27	GND
28	IC28 Q3P	LVDS PPS
29	MCU P31 (RXDA0)	115200 bps 8n1
30	IC28 Q3N	LVDS PPS
31	MCU P30 (TXDA0)	115200 bps 8n1
32	GND
33	TOD UART TX - FPGA pin 50	4800 bps 8N1
34	IC27 Q1P	LVDS 10MHz
35	GND
36	IC27 Q1N	LVDS 10MHz
37	NC?	UNKNOWN
38	GND
39	MCU FLMD0
40	IC27 Q2P	LVDS 10MHz
41	NC?	UNKNOWN
42	IC27 Q2N	LVDS 10MHz
43	MCU pin 28	P33/TIP01/TOP01
44	GND
45	MCU pin 29	P34/TIP10/TOP10
46	IC27 Q3P	LVDS 10MHz
47	MCU pin 30	P35/TIP11/TOP11
48	IC27 Q3N	LVDS 10MHz
49	GND
50	GND

I got my TL866 adapters recently so I've desoldered M29W320 flash memory and dumped the contents but there isn't much there. I could recognize serial/model number and the rest is FPGA bitstream noise. MCU has its own flash and it looks like there are pins on the main header for ISP (UART0 and FLMD0) but I don't have any tools for this arcane architecture (or disassembler)

TOD UART funny enough comes from the FPGA and without GPS (my current condition) it says "TOD_NOT_VALID;" every second.

Main MCU UART works at 115kbps and at boot prints:

Code: [Select]

OK;
CONFIG_FPGA;
CONFIG_FPGA_OK;
CONFIG_GPS;
CONFIG_GPS_OK;

and nothing more, doesn't respond to random keys.

P33/P34/P35 are timer pins so I suppose they might be used to time external events. I've tried poking them with GND or 3V3, but nothing came up on the main UART.

It seems that MAX3243 hasn't much to do with the GPSDO function, it takes UART signals from the main connector, translates them to RS-232 and they are provided on the white pin header. Maybe some front panel interface?

I bought nice Extron distribution amplifier recently: http://www.ebay.co.uk/itm/201366115087
and both GPSDO and RPi for NTP server should fit nicely:



Distribution amplifier uses OPA3691 for video signals, 74AHCT04 for digital outputs (and NJM360 comparators). Need to replace 75 Ohm resistors

Power supply there provides +5 and -5V. 5V seems to handle 2A without trouble (just a quick test with Agilent 6632B as a load, need to check with thermal camera as well). GPSDO requires 12V so I'll use one of these cheap XL6009 step-up modules. It draws 1A @ 12V when heating the OCXO but later it is < 0.4A, maybe less if I enclose OCXO in styrofoam.
I plan to wire 10 MHz sine to R/G/B inputs, 10 MHz square to H and 1pps square to V. Separate PPS signal and TOD UART will go to RPi for NTP server functionality.

I've desoldered the main connector from GPSDO and the pin pitch is 0.8mm so I thought a SMD SATA connector would fit there, just enough to connect to these nice fast rising 1pps and 1 MHz LVDS and I thought about using SATA cable from GPSDO to my breakout board at the front of the case that would break out from SATA connector into 4 SMA connectors. In theory I could use these fast signals as a pulse generator to test things (I don't have free_electron's Jim Williams Pulse Gen). I'm worried though that I'll loose these nice rise times, not so much in the SATA cable, but when breaking them out into SMA connectors on an etched board (also LVDS impedance control and termination perhaps necessary?). 

[Vgkid]

So my modules been powered up for 50 minutes. X3,4 are blinking rapidly. With x8 blinking at 1pps.

[Vgkid]

I left it powered up all day, no change. Checked the antenna with my 6T module, and 2 sats are above 30dB, with 5 above 20dB.
I wonder if the 5T module on board is bad, it looks to have had either been dropped, or something hit it.

[lukier]

Check what the UARTs say, the main MCU UART (115200 8n1) on power up ("CONFIG_GPS_OK;" message) and Time Of Day UART (4800 8n1) that should output either UTC time (when it has a lock), something like that "$GPZDA,215112,28,07,2015,00,00*47" or "TOD_NOT_VALID;" otherwise.

Also, you could probably tap uBlox 5T UART coming from the MCU and see what is going on there.

[Vgkid]

I think the 5t is fried. Should I contact the seller. I have 3 days.

[Vgkid]

Checked it this morning , still no lock. Or anything else that matter.

[Vgkid]

I emailled the seller a few minutes ago, I really hate to complain.

[grumpydoc]

Remember that these were not sold as GPSDOs but just the OCXO,

[Vgkid]

I will see if I get a favorable response back. Good thing is that the 6T is pin compatable,though I wish the solderpads around the 5t were longer. I would have wired my 6T board to the 5T.

[edpalmer42]

I got my board today.  I'm trying to make it talk to me before I hook up an antenna.  I'd like to see if it has a stored location.  It's interesting to see where these things came from.  No discoveries so far, but I have removed a few possibilities.

I noticed that the communications between the LEA-5T and the micro is at 19200.  Add that to the TOD at 4800 and the MCU UART at 115200 and who knows what speed the RS-232 runs at.  I've confirmed Grumpy's statement that all the RS-232 outputs are turned off until at least one valid input is present.  I've checked the three TX-RX pairs 2-3, 5-6, and 9-10 in the white connector at speeds of 2400 - 115200 N-8-1.  I guessed that the 'language' might be SCPI without the normal command prompt so I sent :SYST:STAT? and SYST:STAT? but didn't even get gibberish for a response.

Things to check - not necessarily by me.  Feel free to jump in at any time!     If you'd like to try something, tell us so we're not all doing the same thing.

1.  There's no particular reason why that TX-RX pairing is correct.  Maybe another combination is correct.
2.  Maybe the language isn't SCPI.  Maybe it's U-blox's native language.  Try their u-center software.
3.  Try NMEA commands - although that seems unlikely.

The Quest continues......