BG7TBL designed 10MHz GPSDO

[VK2XAX]

Hi all,

In recent months a Chinese ham designed and built GPSDO has appeared on eBay.

Being a curious chap I decided to buy one to see how well it works.

I'm pleasantly surprised to find that it works exceptionally well.

I've done a complete test & tear down of the unit and written it up.

You can download my doc for this here....

https://drive.google.com/file/d/0B9Oysj ... sp=sharing

regards

Tim

[VK6OX]

An excellent write-up Tim!
Trying to find a Thunderbolt these days is difficult to start with, and over $200 if one is found.
The BG7TBL job therefore appears to be an economical alternative.

[VK3ALB]

VK3RGL 2m beacon uses this very device as its reference. The only issue I found with it was the serial comms were easily disturbed if handshaking was employed. I fiddled around with a few settings but didn't have much luck. I ended up with just gnd and ser out connected to the PC and everything was fine after that.

[VK3BQ]

Do any of you have an "eBay seller name" or other site you sourced one from successfully?


Sent from my iPhone using Tapatalk

[VK2XAX]

Do any of you have an "eBay seller name" or other site you sourced one from successfully?
Sent from my iPhone using Tapatalk

This is where I bought mine...

http://www.ebay.com/itm/121530825744

regards

Tim

[VK2XAX]

The only issue I found with it was the serial comms were easily disturbed if handshaking was employed. I fiddled around with a few settings but didn't have much luck. I ended up with just gnd and ser out connected to the PC and everything was fine after that.

Hi Lou,

Understandably so. The serial out put from the uBLox module is 9600, 8bit, no handshake, 1 stop bit, an old industry standard. Turing on any form of handshaking would definitely upset things. Not only that but since the PPS comes out on pin 8 of the DB9, that pin is used for CTS ( Clear to Send ) and as such is used in hardware handshaking. So that would upset anything that is expecting a proper CTS signal rather than PPS.

Also since the serial out is true RS232 format the waveform goes negative so you have to be careful if your driving anything that expects TTL since it might not take too kindly to -5v being applied.

I've also added this information to the doc.

Nice to know one of these units is in a beacon

regards

Tim

[VK2XAX]

An excellent write-up Tim!
Trying to find a Thunderbolt these days is difficult to start with, and over $200 if one is found.
The BG7TBL job therefore appears to be an economical alternative.

Thanks

I was browsing around eBay the other day to get a feel for the current pricing of such things and I was dismayed to see some Thunderbolts north of $400US !!! This is not far off being half price of a new one. I suppose when the latest models start filtering onto the market we'll see the originals drop in price but it will probably be another 5 years before that will happen.

I'm glad there is this alternative and that it is performant, at least the data output is in the more common NMEA instead of Trimble's TSIP making it a little easier to interface to other things.

regards

Tim

[ZL2BKC]

Thanks for posting Tim. Good review.

I feel obliged to comment that the Phase Noise plots are superfluous as the phase noise of the HP 8566 is a *lot* higher than that of the MV89A oscillator. The BG7TBL GPSDO certainly looks to be a good performer compared to others on the market.

I have been doing some tests on the MV89A myself and they are good performers, but as mentioned in another post resently teh 5MHz output even though it is quite low (-50dBc) it is still enough to cause havoc with the VK3XDK transverter. See http://www.vklogger.com/forum/viewtopic ... 22&t=13080

73,
Wayne

[VK2XAX]

Thanks for posting Tim. Good review.

Hi Wayne - thanks.

I feel obliged to comment that the Phase Noise plots are superfluous as the phase noise of the HP 8566 is a *lot* higher than that of the MV89A oscillator.

Hence my statement "The phase noise as measured is so low that we're mostly seeing the noise floor of the analyser." I can make that clearer by including the noise floor plot of the analyzer in the graph.

The BG7TBL GPSDO certainly looks to be a good performer compared to others on the market.
I have been doing some tests on the MV89A myself and they are good performers, but as mentioned in another post resently teh 5MHz output even though it is quite low (-50dBc) it is still enough to cause havoc with the VK3XDK transverter. See http://www.vklogger.com/forum/viewtopic ... 22&t=13080

I had forgotten that you had done those tests and to be specific, this is with the XDK synth, not the transverters he's built. But it certainly reinforces the importance of a clean reference signal for locking DDS based LO's and taking time to source any spurious signal that turns up in in the output.

Did you persist with the MV89A after that ?

regards

Tim