[time-nuts] New Subscriber, DIY GPSDO project (yes, another one)

[ew]

With all the recent discussions on GPSDO's allow me to add my two cents worth. Being a Ham and at the time subscribing to QST the July 1998 Brooks Shera GPSDO article was what I needed since I had for 10 years controlled my FRK with a 12 bit DAC and did have 60 KHZ Tracor M100 and Loran C. It started with emails with Brooks to adapt his circuit to Rb, changing the input to 100 MHz etc. Visiting him since I had friends in Albuquerque I did have a chance to meet him personally.  Over lunch we had a great time and when I offered to  sign a non disclosure agreement he laughed so hard saying with my lack of understanding of code not necessary. I was made aware of all code changes. After passing away from cancer, starting with melanoma going to his brain his wife gave me permission to share his work with time nuts. 

 

Parallel to that work I got to know Richard McCorkle when shortly after joining time nuts I declared I thought I could do a D/M along with counters but needed help on PIC Code, one response was that's why we call our self time nuts but Richard offered help. This developed in a long relationship with many applications including PIC TIC and GPSDO. Most the time response to a question because of the time difference to Alaska, was code. He was a genius when it came to code and we still do new designs with his code. We disagreed only twice one had to do with his GPSDO where he was using PIC TIC type interface to GPS, in my opinion an overkill. More to it later. My last message from him was that he had fallen down stairs and prostate cancer was detected. He would get with me when better, never did.

 

Lars and I connected also on his code capabilities. We discussed multiple projects but not much his GPSDO.  He never had an opportunity to get AV tests done. Magnus I do not know if he contacted you his health may have prevented it. We skiped regularly till 4 weeks before his passing. Early last year the subject came up and somehow Jim Hamman who had worked with Lars contacted me.  He had been in contact with Lars worked with him on his code and his input circuit. Jim shared his work with Nick Saya but also continue his refinement. Attached the three versions. I offered to run tests on his unit and he send me a picture. I was getting worried and asked him to make it more user friendly. He complied and I did post the excellent results on time nuts. We continue to work together with my focus mainly on board designs and procurement of some of the critical components. Recently we had a great time when Jim came to Florida for 3 weeks 20 miles from me. We call them Snow Birds.

 

>From these inter actions I learned a lot. I like to share with you. To me there are three major sections in a GPSBO. GPS input, Filter and Frequency source. I have piles of information but will only refer to two.

I call the GPS side the dirty side since the output of most GRS receivers have a lot of Jitter. Antenna and field of view is the main contributor to a good signal output. I spend a lot of time of that part of the circuit and it is mainly an improvement of Brooks with later circuits and 100 MHz OCXO. remember well Brooks explanation of sampling. 

 

The NIST paper 

TEST 76120S

Test Folder Number: 288073-16

Test performed for EndRun Technologies

 

tells me a lot about GPS , particular the last plot between NIST and USNO common view.

 

The filter is where you do the music. In reality that is where 90% of your contribution is. Can not say much about it because there is where I know the least. I consider for convenience the DAC part of the filter but for obvious reasons make it part of the Frequency source.

 

The third one is the Frequency source. There has been lately discussion on OCXO's but few have resources other than buying what the budget allows. I like to refer to SRS FS740 manual that clearly demonstrates what a filter can do. It can not improve on the source but maintains it capability. All Telecom GPSDO's diminish Frequency sources because the change frequency to correct for time. That is where the hump comes from and if you have equipment like we do you can see it. The NIST paper shows it.

 

Based on the January visit Jim is working on some refinements that will be part of the next board design, there may well be also a TDC7200 version and a performance comparison will be done. One thing we agreed on there will be separate board with OCXO and LTC1655 with some adoption of the page 11 circuit. Also regulator. Last year I published two universal OCXO boards, plan to add the necessary circuits to the one with output amplifier. There will also be a Rb board Juerg and I plan to use but also for my Maser performance work.

Attached is a sketch from Jim showing the three input circuits starting with Lars. Also his prototype design using cardboard but his final circuit.

Lars along with Brooks have left a legacy

 
Bert Kehren 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: TIC Circuits.pdf
Type: application/pdf
Size: 611681 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20200301/1e292340/attachment.pdf>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Jims GPSDO.jpg
Type: image/jpeg
Size: 406411 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20200301/1e292340/attachment.jpg>