Author Topic: bg7tbl gpsdo master reference (Read 371200 times)

kj7e · « **Reply #675 on:** March 14, 2019, 04:35:36 pm »

Just going to add this here since there is very little info on this site elsewhere regarding the BG7TBL 10Mhz distribution Amp and I use it in conjunction with the GPSDO. Originally, mine was supplied with the DAPU 055F-ASDD OCXO, which was very drifty. I changed it out the DAPU oven for a Trimble 49422 which was a huge improvement, but the EFC Vref was sourced from an onboard voltage regulator and not the most stable. Recently I replaced the Trimble OCXO for a UCT 108663 so I could use the internal ovenized Vref for the EFC, just need to move R34 to the R35 pad. The UCT's Vref is very stable being ovenized, further increasing overall stability.

BG7TBL 10MHz distribution amp with the UCT OCXO;

Bottom;

EFC Vref source selection, move the 18 Ohm resistor from R34 pads (onboard) to R35 pads (OCXO);

Mrt12 · « **Reply #676 on:** March 18, 2019, 12:22:09 pm »

Hi Guys,

I also got one of the STAR4 GPSDOs via eBay.

https://www.ebay.com/itm/Oscilloquartz-Star-4-GPS-Empfanger-10MHz-1PPS-GPS-disziplinierte-Uhr-GPSDO/292905940612

I found out that the sine output of the GPSDO is quite distorted and noisy and has a large DC offset, so I opened the unit and probed the OCXO directly with my 'scope. The OCXO signal looks good and has no offset, so I think there must be something wrong with this daughter board. Perhaps I will make my own one with an integrated distributor.

I measured a few of the signals on the 26pin connector and found out that the 10 MHz sine signal is missing, it is only available on the coaxial connector. Is that correct, or did this eBay seller modify other things on the STAR4 board, besides changing the coax connectors?

Dwaynej · « **Reply #677 on:** April 25, 2019, 02:32:25 am »

Hey all, I have been following this thread for weeks and just ordered a unit for my radios...

https://www.ebay.com/itm/Latest-10MHZ-SINE-WAVE-Sinewave-GPS-DISCiPLINED-CLOCK-GPSDO-with-LCD-Display/112700839490

I was expecting to open it up and quickly identify the components so that I could make some tweaks if needed... A surprise for me with both the Crystal and GPS unlabeled. Any idea based on packaging what they are?

Thanks!

kj7e · « **Reply #678 on:** April 25, 2019, 03:13:15 am »

Looks like the standard Trimble 65256 OCXO, which are very good.

kj7e · « **Reply #679 on:** April 25, 2019, 05:02:17 am »

@Dwaynej

Kindly post some more photos of the top and bottom of the board, there are some changes I see, like the chip in front of the OCXO. I would like to look at the 2018 version you have closer.

Dwaynej · « **Reply #680 on:** April 25, 2019, 05:52:44 am »

Incoming...

Dwaynej · « **Reply #681 on:** April 25, 2019, 06:07:38 am »

More... Note the board SN 2019-01-07.

Dwaynej · « **Reply #682 on:** April 25, 2019, 06:08:24 am »

More...

Dwaynej · « **Reply #683 on:** April 25, 2019, 06:09:42 am »

Last pics... Note the L (for Locked?) on the display after running for a number of hours. Also, black plug is where ribbon cable plugs into main board to display board.

Dwaine · « **Reply #684 on:** April 26, 2019, 06:29:28 am »

Quote from: Mrt12 on March 18, 2019, 12:22:09 pm

Hi Guys,

I also got one of the STAR4 GPSDOs via eBay.

https://www.ebay.com/itm/Oscilloquartz-Star-4-GPS-Empfanger-10MHz-1PPS-GPS-disziplinierte-Uhr-GPSDO/292905940612

I found out that the sine output of the GPSDO is quite distorted and noisy and has a large DC offset, so I opened the unit and probed the OCXO directly with my 'scope. The OCXO signal looks good and has no offset, so I think there must be something wrong with this daughter board. Perhaps I will make my own one with an integrated distributor.

I measured a few of the signals on the 26pin connector and found out that the 10 MHz sine signal is missing, it is only available on the coaxial connector. Is that correct, or did this eBay seller modify other things on the STAR4 board, besides changing the coax connectors?

I get this with the tap for the square wave. Terminated to 50ohms....

Electro Fan · « **Reply #685 on:** April 27, 2019, 02:33:47 am »

I just installed a BG7TBL. Thanks to everyone on this thread for helping me get up to starting speed.

FWIW, I found that simply putting the puck antenna on a baking sheet pan the SNR of the satellites nearly immediately (30 seconds or less?) jumped across the 14 satellites reporting SNRs an average of about 10dB per satellite. In two cases the difference was negative and in two cases there was no difference; in the other 10 the improvement was significant . The numbers represent a rapid capture of the new SNRs so it's possible the other four satellites might dial-in better if I take some more looks. This was just very fast, first impression but in one case a SNR increased 35dB and another increased 28dB. Long story short, just giving the magnetic puck antenna a ground plane is worth a try.

The LCD on the GPSDO is reporting 1,000,000.0012 Hz after about an hour - but that might have taken less if I had stared with the pan as a back plane.

The only thing I can't quite figure out is why the Lady Heather program doesn't render the satellite graphics without hitting Z and then some of the commands. I don't know why I can't get the standard view that everyone shows with the text/number info, the clock, and the satellite views to render all on one screen. Might be something to do with my graphic settings?

Thanks again to all the EEVers who have pioneered this stuff and shared your findings. EF

Electro Fan · « **Reply #686 on:** April 27, 2019, 02:40:18 am »

An update. When I first looked at the Relative strength vs azimuth without the pan ground plane it looked like a very narrow view - maybe 60 degrees or less (I just remember it was narrow). With the pan ground plane it looks like I can see either the full 360 degrees or maybe I'm missing a few spots that might total up to about 60 degrees. I'm still trying to interpret what I'm seeing. This is kind of fun.

Electro Fan · « **Reply #687 on:** April 27, 2019, 03:01:36 am »

Another update. While I was connecting the GPSDO to three counters (of which I have no way of knowing if any are accurate), I found that the GPSDO now shows 5 bars (it was 3) and Lady Heather has now rendered the satellite map in the upper right hand corner.

Electro Fan · « **Reply #688 on:** April 27, 2019, 11:14:10 pm »

The BG7TBL is running a-ok with Lady Heather when connected via a RS232 cable to a standard RS232 port on a PC. With a USB to RS232 cable I can get some communications between Lady Heather and the BG7TBL but I can't figure out how to get the basic receiver command going. Might be a problem with the pin mapping on my USB to RS232 cable (I've tried two), or maybe I just need to figure out the command. The auto /rx command doesn't work. Is there a way when I'm connected via the RS232 cable to interrogate something to figure out what the preferred command is so that when I reconnect with the USB cable that I can tell Lady Heather what it needs? Or maybe there is something amiss with the USB cable in case anyone has a link to a USB cable known to work (on Amazon, etc.)? Thx

Electro Fan · « **Reply #689 on:** April 27, 2019, 11:44:46 pm »

In case this helps anyone understand what is coming and going on what pins at the BG7TBL RS232 port in your own projects, or in case anyone has any suggestions on how to make this work with a USB cable, that would be cool. Thx

texaspyro · « **Reply #690 on:** April 28, 2019, 12:58:52 am »

You need to tell Heather what serial port number that the OS is mapping the USB converter to (use the Windoze device manager). To set the com port number use the /# command line option where # is the com port number... also note that Linux,etc is a bit different). If you don't specify a port number Heather will use COM1.

Electro Fan · « **Reply #691 on:** April 28, 2019, 03:25:17 am »

That was the first thing I did with the USB/RS232 cable - I set the port to Com1. I verified that the command was working by changing it to Com2, and then back again to Com1. Lady Heather displays the Com port very prominently. I also set the speed, etc. Those steps gave me some confidence that the BG7TBL and the PC were at least communicating at some minimal level with the USB/RS232 cable.

In the eBay ad it says:

4.3 Connect PC
GPSDO-DB9 PC-DB9
PIN2--------------------PIN2
PIN3--------------------PIN3
PIN5--------------------PIN5
PIN8--------------------PIN8
GPSDO-DB9 PIN2,PIN5,PIN8 is connect to PC-DB9 PIN2,PIN5,PIN8,
Baud rate is 9600BPS,Baud rate is fixed.

Normally, I would expect Pin 2 (RX) to talk to Pin 3 (TX) but I confirmed with the seller that the pins assignments are as specified in the eBay ad. The seller refers to his pin assignments (above) as "direct connect". In fact I confirmed that the cable I am using is "direct connect" by checking continuity from one connector end of the cable to the other connector.

(And I'm certain that with the PC's RS232 port the BG7TBL is working fine with Lady Heather - so I'm very confident that the software and the GPSDO play well together).

In case it helps, in the photos I posted above there is green activity on the 2 RX, 3 TX, and 8 CTS pins - as the eBay seller specified in the ad. 4 DTR and 7 RTS are permanently red. That leaves 5 GND (which the seller acknowledges is in use) plus no activity on 1 DCD, 6 DSR, and 9 RI - none of which we would expect would be used.

So I'm guessing that the USB to RS232 cable that I'm using is wired differently than as per the "direct connect." I don't quite get the USB architecture and how it maps to some or all of the pins on a DB9 - but with the USB to RS232 cable some commands (such as port, speed, etc.) are getting through - but the /rx auto receiver command isn't working and none of the other commands to set the receiver type are working (I tried them all I think).

Either I need to figure out the command to set the receiver using the USB/RS232 cable (similar to how I set the Com port and speed), or I need to get a better/different USB/RS232 cable. (The RS232 to RS232 connection is fine on the current PC but most other computers are not going to have a DB9 RS232 port and will need a USB port).

Quote from: texaspyro on April 28, 2019, 12:58:52 am

You need to tell Heather what serial port number that the OS is mapping the USB converter to (use the Windoze device manager). To set the com port number use the /# command line option where # is the com port number... also note that Linux,etc is a bit different). If you don't specify a port number Heather will use COM1.

Electro Fan · « **Reply #692 on:** April 28, 2019, 05:02:01 am »

Ok, problem solved.

Sometimes you have to learn older technology before you can deploy newer technology.

It turns out there were two problems.

As mentioned above some of the data was being conveyed between the PC's USB port and the BG7TBL's DB9 port - such as the Com port command. For example I could change between Com1 and Com2, but other commands such as the /rx auto receiver command did not communicate from Lady Heather to the BG7TBL.

The first issue is that Windows mapped the USB port I was using to Com8. I should have found this earlier/quicker. Lady Heather was showing me that I could change to Com1 or Com2 - so it was displaying the port commands but obviously that wasn't fixing the problem because the PC had assigned Com8 to the USB port I was using. However, this still didn't fully fix the problem as Lady Heather still couldn't find a specific receiver even when I set Lady Heather to Com8.

So, second/next I found a null modem adapter. With some continuity testing I had found that various (standard?) RS232 cables do in fact map pin 2 (TX) on one end to Pin 3 (RX) on the other end. And as I mentioned earlier, the cable that I initially used to go from the DB9/RS232 port on the PC to the DB9 port on the BG7TBL had a different straight through wiring such that Pin 2 talked to Pin 2 and Pin 3 talked to Pin 3 (and I think 8 talks to 8, etc.). So, on this cable I tested the null modem adapter (see photo) below and found that it preserved the straight through continuity of my working (but maybe non-standard?) RS232 cable. So then I put the null modem adapter (see photo) on the USB to RS232 cable (at the DB9 end of course), and voila, all is good.

When I open Lady Heather with the USB to RS232 cable with the null modem adapter connected to the BG7TBL I have to first tel Lady Heather that I'm on Com 8 with /8, and then I give it the auto receiver /rx command and everything works.

With the RS232 to RS232 "direct connection" cable Lady Heather starts completely on it's own because Lady Heather starts at Com 1 (where my RS232 cable happens to be plugged in) and then Lady Heather automatically goes into the /rx routine (or some similar default).

So, if you happen to have a RS232 port set to Com1 on your computer and you have a "direct connect' (straight through) cable with a DB9 on each end, this is the easiest way to start up the BG7TBL. If you have standard RS232 cable you will probably need a null modem adapter or redo the pin connectivity to make the wiring"straight through". Likewise, you might also need to do this depending on your USB to RS232 cable is wired (if you are going from USB to DB9).

Long story short, I think that BG7TBL thought that the "direct connect" wiring approach had some advantage over conventional RS232 wiring but when people see a DB9 they assume RS232 and I'm betting this has been the source of some head scratching - both for people who have been using DB9/RS232 to DB9/RS232 cables, and for people using USB to DB9/RS232 cables.

Electro Fan · « **Reply #693 on:** April 28, 2019, 08:47:50 pm »

Another rookie question:

Is the accuracy of the 10 MHz output from the BG7TBL dependent on the quality of the composite satellite signal, or is it binary?

In other words, if the satellite signals are good enough to establish a Lock condition then the internal oscillator is going to be the determining factor? Or does a relatively stronger GPS input signal make a relatively more accurate 10 MHz output signal?

For example, I see the LCD on the BG7TBL move around in the range of 999,999.9997 Hz to 10,000,000.0000 Hz. I could just be imagining things but it seems that when I have the antenna positioned for a relatively better view of the sky which produces higher SNRs the LCD is more likely to stay stable at 10,000,000.0000 Hz, and when I move the antenna to a lesser view and get lower SNRs the LCD is more likely to be at 999,999.9997 Hz or so.

Thx

kj7e · « **Reply #694 on:** April 28, 2019, 08:56:26 pm »

More tracked satellites and a high SNR will always give you the best and most accurate lock and lowest jitter. However, the accuracy does not diminish significantly with a mediocre SNR and only a few satellites tracked, there will be more jitter and short term stability will be reduced slightly.

Electro Fan · « **Reply #695 on:** April 28, 2019, 09:23:26 pm »

Cool! Thanks!

Quote from: kj7e on April 28, 2019, 08:56:26 pm

More tracked satellites and a high SNR will always give you the best and most accurate lock and lowest jitter. However, the accuracy does not diminish significantly with a mediocre SNR and only a few satellites tracked, there will be more jitter and short term stability will be reduced slightly.

Electro Fan · « **Reply #696 on:** April 28, 2019, 11:15:31 pm »

A little off thread topic, sorry - but this might help others who have been on the fence about adding a GPSDO.

After thinking about GPSDOs for awhile (couple years?) and now finally having one for a weekend, I'd say go for it!

I had a hunch installing a GPSDO would be a learning experience - and it has been on a two levels: 1) learning about GPSDOs and Lady Heather, and 2) seeing what happens when you can see frequency and time with better accuracy.

Having said that, this might be the gateway to TimeNuts land (said affectionately).

I'm now pretty sure that my HP 5315A is good to within 1 Hz at 10 MHz. My 5327B is out about 6 Hz at 10 MHz but that's 1 Hz better than I previously thought - and discounted for the cool factor of Nixie Tubes I'd say it's spot on (just kidding, but it was mostly acquired because I like looking at the Nixie Tubes, go figure.)

Anyway, that's mostly the good news. I'm now pretty sure that my 33522B could be off about 1 Hz on 10 MHz. That's kind of disappointing but the good news on that is that when I drive it with the BG7TBL it snaps into time alignment. So, the BG7TBL has (I think, I'm pretty sure) tuned up my reference generator.

Back to the good news or the bad news depending on how you look at it. The Tek 2247A was not very close on on frequency; I had a hunch but now I'm sure. The good news is that it has an input for 10 MHz and with the BG7TBL it snaps into what looks like an accuracy of ~1 Hz at 10 MHz. The 2467B wasn't out as much as the 2247A but it doesn't seem to have an input for an external 10 MHz clock - if anyone knows different, please say.

So, now the question is can I get by with the BG7TBL or is a distribution amplifier needed? It's a slippery slope. YMMV but I have found the BG7TBL to be a very good investment in $ and (haha) time. EF

kj7e · « **Reply #697 on:** April 29, 2019, 12:27:39 am »

Yes, get the distribution amp, I have found it alone cleans up the output from the GPSDO even more. The 2247a has a standard crystal for the counter, not a TCXO, so only expect a few PPM or 5x10 -6 and its going to be temp sensitive. The good news, its easy to adjust, trim cap next to the crystal, rear of the scope on the top board. But dont expect any long term accuracy out of it.

Electro Fan · « **Reply #698 on:** April 29, 2019, 02:22:06 am »

Thanks for the tip on the 2247A, I was sure the control was in there somewhere but figured it was buried - sounds pretty accessible. And thanks for the heads-up on what to expect. Knowing the threshold of what is good enough can save a lot of time finding it

Do you have a link to the distribution amp you like? Thx again.

Quote from: kj7e on April 29, 2019, 12:27:39 am

Yes, get the distribution amp, I have found it alone cleans up the output from the GPSDO even more. The 2247a has a standard crystal for the counter, not a TCXO, so only expect a few PPM or 5x10 -6 and its going to be temp sensitive. The good news, its easy to adjust, trim cap next to the crystal, rear of the scope on the top board. But dont expect any long term accuracy out of it.

kj7e · « **Reply #699 on:** April 29, 2019, 02:40:18 am »

The BG7TBL distribution amp works very well and if your going to use the GPSDO to feed it there is no need for the OCXO in the distribution amp. I read someone who found the combination of the internal OCXO while being fed an external ref caused increased IMD.

https://www.ebay.com/itm/10MHz-Distribution-amplifier-frequency-standard-8-port-output/121714097936?epid=869843952&hash=item1c56b9bf10:g:fjEAAOSw23NZoQQO


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: bg7tbl gpsdo master reference (Read 371200 times)

Share me