[time-nuts] Thunderbolt reception problems

Magnus Danielson magnus at rubidium.dyndns.org
Sun Nov 22 22:50:58 UTC 2009


Dear Said,

SAIDJACK at aol.com wrote:
> Hi Guys,
>  
> changing the AMU thresholds to 2.0 as suggested by Warren fixed the  
> problem, the unit is now working properly, thanks much for the hint. And no,  this 
> was not obvious to me.
>  
> It is always great to get constructive help rather than the "you should  
> know better" type of comments.

I certainly understand that one.

> In the end, it looks like this turned out to be a configuration issue  
> rather than a receiver sensitivity issue. But this also begs the question: how  
> much performance degradation will I get when using 2.0 as the AMU threshold  
> rather than the factory default. There must have been a reason why Trimble 
> set  the default so high, they mention something about multipath rejection 
> in the  manual.

I kindly disagree... sorry for the long rambling motivation.

While I have not been able to conclude exactly what the AMU measure 
really is (Amplitude Measurement Unit), but it is similar at least to 
C/N measures. When the receiver has a bad C/N measure from a certain 
source, it may be wise not to rely on it. The signal level from each sat 
can be configured to be delivered in either AMU or in dB-Hz where the 
later allows comparision among other GPSes. The best description of AMU 
I have is from the Lassen iQ manual where it states:

Note – The signal level provided in this packet is a linear measurement
of the signal strength after correlation or de-spreading. Units, either
AMU or dBHz, are controlled by Packet 0x35.

Thus, you can request the 20*log10(AMU)+Const that converts it into 
dB/sqrt(Hz) measure.

The Lassen iQ AMU limit is default at 2 and for extended sensitivity 
1.2. The Thunderbolt AMU values may not need to have the same 
relationship to dB/sqrt(Hz) as the Lassen iQ, but there may be more to 
it. The Lassen iQ is tailored towards navigation gear (hence the garage 
startup) where as the Thunderbolt is aimed at timing.

Regardless, AMU or C/N measures is an indication of correlation strength 
and thus a measure of signal quality. It should rightfully contribute to 
the TDOP measure. The AMU limit is a pre-T-RAIM cleanup protection value.

If you needed to lower your AMU limit from 4 to 2, you are feeding the 
Thunderbolt a signal level below the intended levels. A 6 dB raise and 
you would not have needed to shift the limit for the same tracking 
result. Regardless of where the limit is, the lower AMU values will 
indicate a higher noise level and thus higher uncertainty in time than 
otherwise obtained.

So, the 10 dB loss in the 1-to-4 splitter is significant here. It's just 
unfortunate that the Thunderbolt is not clearer about it.

> Why do I want to get this unit up and running: this unit has been  tested 
> to have very good phase noise performance. I'll use  anyone's low noise 
> source anyday.

Which is also why I like to see mine operational, alongside my Z3801, 
GPS-4 and others.

Cheers,
Magnus




More information about the Time-nuts_lists.febo.com mailing list