A look at my Symmetricom GPSDO / 10MHz reference (OCXO + Furuno receiver)

[Macbeth]

I think calling him a thief is a bit too much. I guess he doesn't have any idea what he is selling is working or not and he should list it as such. However, they are cheap - and there is a reason for it. The market price would appear to include that risk. There are many examples of how these boards turn up in this thread alone.

He could easily have more goodwill for sure if he was more honest, but ultimately he is at ebays/paypals mercy and a full refund is on offer. Clearly he has costed this into his business model selling scrap boards. I'm sure he thinks lots of his customers are thieves too, even me because I ultimately got mine for free (though not in my time, effort, and a couple of quid fixing the fucking thing  )

Seriously, when you are dealing with scrap boards coming from China you should have your wits about you and be willing to write off the transaction, even though you are guaranteed a full refund by AliExpress or ebay, etc.

The guy is not a thief. Just a typical Chinese businessman.

[kilobyte]

Can someone look what's the Part Marking on U22?

It's a 5 pin SOT marked Z32H. I reckon NC7SZ32

Thanks for the info Macbeth.
I will try it first with some singlegate 74ahc32 from a old scrap pcb. Should be good enough for a quick test.

I didn't powered the UCCM yet, because the U22 is missing.
If it will work then I'm happy and if it doesn't work then I can still ask for a refund.

[uski]

I think calling him a thief is a bit too much. I guess he doesn't have any idea what he is selling is working or not and he should list it as such. However, they are cheap - and there is a reason for it. The market price would appear to include that risk. There are many examples of how these boards turn up in this thread alone.

He could easily have more goodwill for sure if he was more honest, but ultimately he is at ebays/paypals mercy and a full refund is on offer. Clearly he has costed this into his business model selling scrap boards. I'm sure he thinks lots of his customers are thieves too, even me because I ultimately got mine for free (though not in my time, effort, and a couple of quid fixing the fucking thing  )

Seriously, when you are dealing with scrap boards coming from China you should have your wits about you and be willing to write off the transaction, even though you are guaranteed a full refund by AliExpress or ebay, etc.

The guy is not a thief. Just a typical Chinese businessman.

He agreed to send a new board. He lied and never did. Scrap board or anything else, it's the same, he lied, this should not be accepted.

I've been dealing with chinese sellers for years. I know how it works. I'm generally reasonable. But when someone promises you something and they don't fulfill their promise, then they steal.

It's simple...

I'd like to add that he increased the prices quite a bit. So he's making a very decent profit. We're not speaking about a poor guy making $0.1 per transaction. He's making at least $25 per transaction, considering the recent price increase. At that price, he should behave differently.

[Macbeth]

uski, your not wrong. I mean he had me with the same "ok - i send one out!". But I just used the ebay/paypal rules against him when he couldn't prove he sent a replacement (which he obviously didn't). I am very surprised you got stung and didn't actually get a refund.

But yes, this guy is right at the sleazy end of Chinese sellers. Most are very good.

[grumpydoc]

I got a replacement without much argument.

But it appears he then put me in his blocked buyer list 

[TheSteve]

I got a full refund when the trimble unit I ordered from him didn't show. It did arrive after 2 + months finally, and had a few small caps ripped off the board.

[NivagSwerdna]

I managed to get the UCCM-P, display and Raspberry Pi hooked up with the same kernel variant I used for my prior NTP server.
The Furano PPS is just tapped off and seems to work OK with the Pi (this goes in via GPIO 18)

Code: [Select]

root@ntpi:/etc# ppstest /dev/pps0
trying PPS source "/dev/pps0"
found PPS source "/dev/pps0"
ok, found 1 source(s), now start fetching data...
source 0 - assert 1461343553.000001559, sequence: 3585 - clear  0.000000000, sequence: 0
source 0 - assert 1461343554.000001046, sequence: 3586 - clear  0.000000000, sequence: 0
source 0 - assert 1461343555.000001533, sequence: 3587 - clear  0.000000000, sequence: 0
source 0 - assert 1461343556.000001019, sequence: 3588 - clear  0.000000000, sequence: 0
source 0 - assert 1461343557.000001506, sequence: 3589 - clear  0.000000000, sequence: 0
source 0 - assert 1461343558.000001994, sequence: 3590 - clear  0.000000000, sequence: 0
^C

I translate the $PFEC,GPtps messages which I am already receiving into $GPRMC ones and send to RPi (I have a spare UART in the Arduino2560).

Then use the NMEA refclock driver.... that's the theory anyway...

From clockstats...

Code: [Select]

57500 71382.283 127.127.20.0 $GPRMC,194942,A,0.0,N,0.0,W,0.0,0.0,220416,0.0,E*54
I don't know enough about NTPd to tell if this is good or not...

Code: [Select]

pi@ntpi ~ $ ntpq -p
     remote           refid      st t when poll reach   delay   offset  jitter
==============================================================================
oPPS(0)          .PPS.            0 l    3   16  377    0.000    0.001   0.002
xGPS_NMEA(0)     .GPS.            0 l   50   64  377    0.000   36.874   3.410
*ptbtime1.ptb.de .PTB.            1 u    2   64  175   47.091   -4.047   0.539
-pyxis.my-rz.de  36.224.68.195    2 u    7   64  377   35.558   -2.520   0.358
-cipher-code.de  129.70.132.32    3 u    4   64  377   35.602   -6.459   3.421
+ntp.atomki.mta. .GPS.            1 u   12   64  377   53.283   -5.236   2.095
+195.50.171.101  145.253.3.52     2 u   46   64  377   31.870   -4.044   0.236
pi@ntpi ~ $


Using config...

Code: [Select]

# pps-gpio on /dev/pps0
server 127.127.22.0 minpoll 4 maxpoll 4
fudge 127.127.22.0 refid PPS
fudge 127.127.22.0 flag3 1  # enable kernel PLL/FLL clock discipline

#NMEA
#server 127.127.20.0 mode 17
server 127.127.20.0 mode 65553 time1 0.7



[kilobyte]

So, I've repaired my Symmetricon GPSDO successfully with a SN74AHC1G32 from an old PCB and with some enamelled copper wire.
Without U22 it says something like "FPGA error" on startup.

Next step is to order some MCX to SMA/BNC pigtails

[Nuno_pt]

Well I'm jealous of you guys,mine still haven't arrive.
Let's see what damage it shows or what missing parts are not there.

My seller was "waikeionline_uk", he gave free tacking number and free shipping for 40€.
The tracking number shows that it arrive at Portugal on 15 of this month, so 1 week in the customs and counting, I hope I don't have to pay the VAT + Customs expenses.

[PA0PBZ]

My seller was "waikeionline_uk"

I ordered 2 more from waikeionline_us, and they both arrived broken.





So I sent him the pictures and he asked me if I wanted a refund or replacement. I told him I want a replacement if he can send unbroken working ones (like he says on ebay: working).
Then he asked me to send the broken ones back and I told him that it would cost me $20 ( eur 16.50) for registered mail. So he refunded me $20 from the original payment... hmmm, not exactly fair but ok. Let's see how things progress. Clearly they were sent broken because the packing was sealed an the missing part was not in the package.

[NivagSwerdna]

I'm still trying to get a refund from flyxy2015.  He asked me to send it back and we eventually managed to get the US$20 shipping transferred to me, still waiting for a proper refund or replacement... I haven't given up on him yet... it will either end up in a replacement or a refund.  I would be a bit worried with a partial refund... might be worth asking EBAY.

The damage appears to be consistent with other boards... i.e. it shows what happens when you brute force the board out.

I resoldered my L20... was hanging off like yours.  Likewise my C1 was a bit bent, not as bad as yours!

L3 is an inductor...  Coilcraft marked 333 on my board (33uH?)

Also check the bottom very carefully... I had a missing cap on the bottom.

I don't think it is unreasonable for them to get the board back if they are giving a replacement/refund but they should pay the postage (just make sure you file as Not As Described).

[PA0PBZ]

I looked at L20 thinking I could resolder it but couldn't find the wires, it ripped off badly.
And yes, one of the boards also has some damage on the other side:



I offered him $20 for the broken boards thinking that I could maybe fix one or both, but he didn't agree.

[Bryan]

Hi Daniel
Yes please that would be brill,i wish I could do that,
Thanks Dave 2E0DMB

Here it is. Please tell me if it works, this way I can send the fix to the original author.


Cheers.
---
Daniel

Re:
https://www.eevblog.com/forum/projects/a-look-at-my-symmetricom-gpsdo-(ocxo-furuno-receiver)/msg849732/#msg849732

Any change of getting the sourcecode , or at least a description of the differences against the below source 

Afficheur-GPS-V2.22-Source
http://www.f4ctz.fr/?page_id=692&aid=1282&sa=1

I'm using AVR not PIC

/Bingo

I sent the patch this week for the furuno support to F4CTZ, I think he will made it available soon (full source code or patch only, I don't know).

Cheers.
---
Daniel

Any luck in the author able to post the changes (source, changes), checked the site and didn't see anything new.

[NivagSwerdna]

Please could someone suggest how I could convert the single 10MHz out to a dual output so I can plug it into my Racal-Dana 1999 Frequency counter ans Siglent 1025 Function Generator simultaenously.  A Fan-out Clock Buffer?

[Nuno_pt]

Here is one http://www.f4ctz.fr/?page_id=891&lang=en
OZ2M make one with 10 outputs for about 100€
Mini Circuits also sells some
G4HUP have another http://www.g4hup.com/DA/DA1_4.htm
Another one http://www.stable32.com/An%20Octal%2010%20MHz%20Distribution%20Amplifier.pdf
DEMI also sells some

[uncle_bob]

Please could someone suggest how I could convert the single 10MHz out to a dual output so I can plug it into my Racal-Dana 1999 Frequency counter ans Siglent 1025 Function Generator simultaenously.  A Fan-out Clock Buffer?

Hi

Are you looking for something to build (a schematic) or something to buy (a rack mount distribution amp) ?

Bob

[NivagSwerdna]

Definitely build and I want it super simple.  I assume all I really need is a buffer amp, probably got something suitable in the junk box.

[edpalmer42]

On most test equipment, the External Reference input is NOT 50 ohm.  It's usually much higher - typically in the hundreds of ohms.  So you don't need any kind of splitter or distribution amp.  Just use a BNC T connector and daisy-chain one cable from one piece of equipment to the next.  Maybe finish off the cable with a 50 ohm terminator.

I checked your Racal 1999 and although I didn't see a spec, the schematic shows that the impedance will be over 500 ohms.  I couldn't find any info on the Siglent 1025.

Ed

[uncle_bob]

Definitely build and I want it super simple.  I assume all I really need is a buffer amp, probably got something suitable in the junk box.

Hi

Ok, super simple (but not a Tee):

Run the signal into the input of a NS7SZ125 or any similar single gate. AC couple it through a 0.1 uf cap and bias it with a pair of 10K resistors to put 1/2 Vcc on the input. You now have a (sort of) square wave at the output of the gate at 10 MHz. You can play with the bias a bit if you want 50/50 duty cycle.

Next run that into however many gates you need to drive your outputs. Put 100 ohms in series with the output and 100 ohms to ground (50 ohm source). AC couple that to a coax connector and you have the drive for your gizmo. If you want to get exotic, put a two inductor one capacitor Tee lowpass filter in place of the two resistors.

None of the parts are critical. None of the values are critical. Just use something as fast as 74AC or faster (no 74HC or 4000 series).

Bob

[SeanB]

Just look locally for a used video distribution amplifier. Those typically have 2 inputs, switchable termination, and you can swap out the terminator resistor to a 56R one to make it 50R input, and then replace the 75R output resistors with 51R ones to get a 50R output. You get from 2 to 4 outputs per side, or at least 4 outputs for a single input. While not technically rated for 10MHz, as video is really only going to 7MHz, they typically will do around 15MHz before the gain drops off to less than unity.

I have been tossing them out recently, though I still have one left in a 1U rack format, which has 4 inputs and each has 16 outputs. A little much for the OPAMP LABS potted modules to drive, I have had to fix 2 of them to replace burnt out power transistors, seeing as they use a BC237/BC337 pair inside. Replaced with 2N2219/2N2905 with push on heatsinks, using the original LM709 opamp inside, though the tantalum caps in it and some of the board did not survive the epoxy removal, but a bodge wire and some new resistors and diodes and it worked.

[Bryan]

Please could someone suggest how I could convert the single 10MHz out to a dual output so I can plug it into my Racal-Dana 1999 Frequency counter ans Siglent 1025 Function Generator simultaenously.  A Fan-out Clock Buffer?

The schematics are all here if you wanted to roll your own vs the kit.

https://www.tapr.org/kits_tadd-1.html

[f1rmb]

...

Any luck in the author able to post the changes (source, changes), checked the site and didn't see anything new.

Yes, code has been merged, unified version should be release sooner or later.

[NivagSwerdna]

Is there a +ve and -ve regulated supply on the board?

The reason I ask is that looking at the 10MHz output... https://www.eevblog.com/forum/projects/a-look-at-my-symmetricom-gpsdo-(ocxo-furuno-receiver)/msg913093/#msg913093 it is exactly the right shape so drive both my counter and my function generator. I believe the Siglent spec is 3v3 pp and the Racal counter something similar.

[Nuno_pt]

Hi all,

Mine just arrive this morning, board with oil, missing GPS connector, on the top, that I can see.

On the bottom missing C73, in the middle of U20 and U16, missing one C or R left of R89 and the U next to it.

The rest of the pads doesn't seem to ever been solder, the solder mask is intact.

Two pictures of Hi resolution, here < https://onedrive.live.com/redir?resid=A26220D7C8CFDD26!11112&authkey=!AILPETMicS4LO8A&ithint=folder%2cJPG >.

Can you guys spot anything else I might have overlook.
 

[NivagSwerdna]

Indeed you have C73 missing as I did.  No idea how they managed to lose that one!

In the section between C119 and R89 I also have an unpopulated chip spot so I don't think you have that part missing.

If you look super close it's quite easy to spot the genuinely missing components as the pads are rough

Your board is pretty typical, shame about the antenna connector but again par for the course.. I guess the soldering wasn't good enough to cope with the loving dismantling the board went through in its last hours.

(You should delete the out of focus images; my eyes are squiffy now)