[time-nuts] Thunderbolt reception problems

WarrenS warrensjmail-one at yahoo.com
Tue Nov 24 07:27:28 UTC 2009


> Magnus said
> Actually, that was not my point.
Yea I know, but I hoped it was close enough to what we do agree on.

>My point was that the signal levels is  lower than what is the normally 
> recommended level and worse performance may be expected otherwise.
That is a good example of speculations with possible truth.
Myself, I have not seen the recommended signal level, especially for OUR application.
And to say the performance will be worse, Maybe, Maybe not. Have you tested it to know where it starts degrading?

My thought is JUST because it is a factory default, 
That  is a long way from saying it is the BEST or even a recommended setting for OUR application (which is not a cell site).
If you take the idea that the "Default is the recommended best" logic a bit further then It would seem to say that:
A TC of 100, A damping of 1.2, a Dac_startup voltage of zero, NO location stored after a site survey, a Dac gain of  -5Hz/V, a survey of 3000,  and ...  ARE all the best thing to use,    ...  I hope you know none of that is not true. 
Changing any and ALL of the above INCLUDING AMU, and performance can be improved "in our application'. 


> I rather viewed it as, when things isn't as optimal... lower the level 
> to get more sats to play with and thus a more stable situation.
Close, BUT not quite correct.
Has more do with the Tbolt's  control loop design and how it handles when satellites switch in and out, 
which unfortunately is not optimized for Our application, and the low AMU setting is just a Band-Aid to help that..  


> These things isn't really in conflict... one is being aware that you 
> left stable ground and the other is how to best handle that situation.
Yes, I think we ALMOST agree here, 
IF you change the 'You left stable ground' statement to "You could made it even better".. 



> Lowering the AMU limit would hopefully get sufficient sats in place, but 
> it can be used to find a balance so that the effective constellation 
> doesn't change, so weak potential dropouts can be cleared off while many 
> reliable (although maybe just a thad weak) remain in the solution. The 
> AMU limit is a two-edged sword... at least.
Close but only partially true. Its pretty much a single edge ... 
Lower the AMU and raise the elevation and things will get better for all fix location signal conditions.
Only real question is what is the optimum setting of the AMU when there is a good strong signal
Answer is probable about 3.

One of the faults with your statement, 
"so weak potential dropouts can be cleared off while many  reliable remain in the solution". 
That assumes there are MORE than 8 sat available for it to choose from, which is seldom the case.
And during the few times that there are more than 8 to choose from, your statement would indicate that the way they are choosing which to use is not at all based on signal level directly or indirectly.  Myself I do not know what their criteria is, 
I did not see an independent setting like on the Motorola Oncores that say take the highest 
and strongest ones when in fixed position mode and take the lowest ones when tracking position.
I'm going to give them some credit and say they probable got the basics right.

> 
> I was trying to warn about the fact that you now accept lower power signals, 
And we fully agree here

>and is not just a magical twist of knobs that makes everything  good again.
We agree here, no magic, just need to change the default AMU, which is all 'Said" was asking for.

>  There is a benefit in lowering that limit given the  situation, and I happily agree with that.
And we agree again. 

So we do not actually disagree on much, just like to argue the Finer points.
If you would like to discuss a fix that is better than the AMU band-aid I'm now using, 
glad to talk about how I plan to modify their control loop to work better.

Have fun
ws

***************

----- Original Message ----- 
From: "Magnus Danielson" <magnus at rubidium.dyndns.org>
To: "WarrenS" <warrensjmail-one at yahoo.com>; "Discussion of precise time and frequency measurement" <time-nuts at febo.com>
Sent: Monday, November 23, 2009 4:06 PM
Subject: Re: [time-nuts] Thunderbolt reception problems


> WarrenS wrote:
>> Some of the disagreement has to do with the fact that Two similar topics, each with a different answer are being mixed together here.
>> 
>> Magnus's point:  
>> 1) How to make the Tbolt the best that it can be? 
>> Answer: Start with a good strong signal and a quiet environment.
> 
> Actually, that was not my point. My point was that the signal levels is 
> lower than what is the normally recommended level and worse performance 
> may be expected otherwise.
> 
>> Said's situation: 
>> 2) How to make the Tbolt work the best that it can with a less than optimized existing setup.
>> Answer: Lower the AMU to 1, rise the elevation to 15or20, increase the TC setting to 500 sec.
>> (It will work better than when the factory defaults are use with the #1 case above)
> 
> I rather viewed it as, when things isn't as optimal... lower the level 
> to get more sats to play with and thus a more stable situation.
> 
> These things isn't really in conflict... one is being aware that you 
> left stable ground and the other is how to best handle that situation.
> 
>> Interesting enough, I have both cases with optimized setting running on my bench now and although the #1 is generally about 25% to 50% quieter, 
>> It is not always so.  About 25% of the time the #2 case is as quiet or quieter.  So the less than perfect #2 case is not really a big deal to most. 
>> There are much more important things that can be done if one likes to 'tweak & fiddle'.
> 
> Lowering the AMU limit would hopefully get sufficient sats in place, but 
> it can be used to find a balance so that the effective constallation 
> doesn't change, so weak potential dropouts can be cleared off while many 
> reliable (altought maybe just a thad weak) remain in the solution. The 
> AMU limit is a two-edged sword... at least.
> 
> I was trying to warn about the fact that you now accept lower power 
> signals, and is not just a magical twist of knobs that makes everything 
> good again. There is a benefit in lowering that limit given the 
> situation, and I happilly agree with that.
> 
>> concerning:
>>> that you may not get the performance of the spec-sheet. 
>> Not a problem, cause they seem careful not to include any specs concerning this except for the 1e-12 per day average.
> 
> TvBs measurements could be another source... it's just so you know that 
> beyond that point your milage may vary...
> 
> Cheers,
> Magnus



More information about the time-nuts mailing list