[time-nuts] CS reservoir depletion

[Bob Camp]

Hi

While I agree with the idea of a "super GPS" as being a good standard, there are some reasons for needing something else:

1) You need something to compare *your* GPS gizmo to in your setting. Knowing that it might be working ok is not as good as knowing that it is working ok.

2) Without some fancy corrections, GPS can indeed drift. The period might be hours, it could be days. The net effects will cancel over a long enough time, but that time may be longer than your loop can suppress.  

3) Anything that creates a "fast" change in your local fly wheel standards will still show up on the output. It will walk out with time. Things in this category are stuff like a step change in the supply voltage creating a change in your OCXO output. 

All of that can be worked on. Much of it applies equally to having a single Cs in the basement. 

There are a couple of things the GPS gives you that the Cs will not:

1) It will tell you what time it is to within a few ns. Most of us "frequency guys" don't realize quite how hard that is to do without GPS. 

2) The long term stability (because it's steered) of the GPS is going to eventually beat anything else out there. If you do really long data runs (months, years) the GPS will always win.

3) Low cost to run. Even if you have a working Cs, how long until the tube goes? The GPS has essentially no wear out mechanisms. The Cs is full of strange stuff. The parts in the GPS are pretty cheap / easy to find. 

4) Easy to duplicate. Everything you are likely to use in a fancy GPS setup is commonly available in quantity. The components are "known good" and cheap. No gamble on a tube or other hard to replace stuff. 

5) Easy to run. There's not much mystery about what's going on. Nothing is hidden in a vacuum bottle. No hidden numbers in a rom. Everything is pretty much right in front of you. The only exception would be the disciplining software in the GPSDO if you choose to use it. 

I admit that the GPS makes a *lot* of sense. I sure wouldn't turn down a nice shinny new HP cesium if one showed up on my doorstep. 

Bob




On Jan 16, 2011, at 7:01 AM, EWKehren at aol.com wrote:

> Having two HP CBT's minus enclosure sitting on my window sill, allow me to  
> ad my two cents worth. Looking at the assemblies I see more art than  
> science and duplicating something like that would most likely end in  failure. 
> Comparing that to the previous H Maser discussions the collective know  how 
> and resources of the list could maybe result in a  Maser.
> As to refurbishing tubes some of us have discussed that off line and  in  
> my opinion with proper tools and equipment cleaning and replacing  cesium is 
> doable. Cynics say the manufacturers of tubes could do it but rather  sell 
> only new ones since they are the only source of the much more expensive new  
> tube. I think that is only half the story.
> The reason in my opinion why refurbishing the tube is commercially not  
> viable is you have to ask: when done what do you really have? You have not  
> eliminated some of the failure modes, in the case of the 5071 the data in the  
> EPROM does not necessarily reflect the tube and who could say how long the 
> tube  would last? A crap shoot.
> I have a HP 5061 B and a HP 5062 C but eventually want to replace them  
> totally with a Tbolt-Rb-HP 10811 combination using two digital loops that are  
> tailored to  the devices. I am now focusing  on the thermal management  in 
> order to get maximum performance. There is work going on by some members of  
> the list to develop more sophisticated digital loops. Lets face it, with GPS 
> properly  used, having a Cesium Standard will give you the warm feeling  
> that you have a primary standard.
> By the way that is why I repeatedly have asked the list if there is any  
> long term Tbolt data out there comparing the 1 PPS or the 10 MHz with a  Maser.
> I hope this is worth two cents.
> Bert Kehren
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.