Trimble Thunderbolt E

[notfaded1]

I'm going to try the autotune once this stabilizes some but I put the antenna on the roof today!  HUGE improvement.  I can actually shorten the cable inside to improve it some more as right now I have a coil of TNC cable inside that's not really needed to be hooked up but it was easiest at first.  I can shorten the run by about 20 feet by removing it.

Bill
 

[notfaded1]

Thank you D I S H network for the cabling I reused.

[ArthurDent]

You have done the first thing I recommended and that was get a stronger signal with an improved antenna location to allow making meaningful comparison test. I also found when I got my first GPS that having adapters and other connectors in the antenna line causes losses that I could easily see. I would drill a hole next to the window frame and have a single run of RG-6 foam from the arrester to the receiver or distribution amp rather than have the flat line adapter you have going through the window.  My RG-6 foam line is 53 feet long and I have several feet coiled up that I don't worry about. Knowing the exact cable length allows me to correct for cable delay in the receiver set-up if I want to.

[Mark19960]

Interesting. I never considered using my existing RG6 for a GPS antenna.
When we moved house I pulled new cable outside and one was for a timing antenna.
I left an existing RG6 run in place and added several LMR240 runs. One of those LMR240 runs was for the timing antenna.

My curiousity has been peaked. I may have to give the RG6 that's there unused a go.
It's lower loss @ 1Ghz than the LMR240 IIRC.
Never considered it. In my mind... the GPSDOs are 50 ohm, the timing antenna is 50 ohm... I need 50 ohm cable.

I like the Trimble. I have one of the older models.
Tempting

[tkamiya]

Trimble manual actually states, using 75 ohm and the resultant mismatch is not going to cause problem.  It actually recommends using RG-6 because of reduced loss.

I say the signal chart....  It's MUCH better but still not great.  With installation like that, the graph should show much better result.  Would you like to (just for testing) eliminate that flexible PCB thingie and test?  If that works, you can use a short piece of very thin coax and thread it through the gap in your window.  It can be done.  I've done it before.

[Mark19960]

It sure does.
Never paid that bit any attention.
In my case the run is not that long @ around 50 feet so I might not gain that much moving over to the RG6 but I am curious now.

[Dr. Frank]

Hello Bill,

you still have completely messy parameters, which prohibit good ADEV values!
These strange oscillations of pps and DAC value indicate that there's a big problem.

Please run the autotune function, or set EL =20..25°, AMU = 1, TC = 500 sec, Gain ~ 1..2Hz/V, like  ArthurDents receiver.
I suppose you meanwhile know, that these parameters can be stored into Thunderbolts EEPROM, so that these are available at startup.
Just a silly idea: Have you made a precise position survey (24..48h) and stored this correctly?

Frank

[notfaded1]

Hello Bill,

you still have completely messy parameters, which prohibit good ADEV values!
These strange oscillations of pps and DAC value indicate that there's a big problem.

Please run the autotune function, or set EL =20..25°, AMU = 1, TC = 500 sec, Gain ~ 1..2Hz/V, like  ArthurDents receiver.
I suppose you meanwhile know, that these parameters can be stored into Thunderbolts EEPROM, so that these are available at startup.
Just a silly idea: Have you made a precise position survey 824..48h) and stored this correctly?

Frank

That's the next step Dr. Frank... I wanted to let it run overnight and stabilize before kicking off the autotune because that's what i've read you're supposed to do?  I'll read up about the position suvey today.

Thanks,

Bill

[Theboel]

Hi notfaded1,

if You have spare time in the future please put the tv disc back in place and use it as solid ground plane You will find interesting result

[notfaded1]

I made some changes (Dr. Frank suggested making mine more like AuthurDent's) then followed the directions and the graphs looks much much better than before.  You can see the before and after on the graph here.  Now I can focus on what I can do to improve signals any further.  So far I removed the long TNC cable and replaced it with a 10 Ft. Type F cable like the other RG-6 in the run.

Bill

[notfaded1]

Props to John aka KE5FX for the Thuderbolt tuning functionality in the lady heather software... it works a treat! 

[Dr. Frank]

Hi Bill,
the signal strength looks good. I doubt that you will achieve better results here. Maybe you better chose EL >= 20°, if you're surrounded by other buildings.
Anyhow the ADEV parameters, and also the OSC value are not stable enough. I'm still wondering, why.

Which version of OCXO do you have in your TB?
The original, black labelled Trimble one, or this red labelled 'piezo' one?

The Trimble version would much better.
http://www.ke5fx.com/tbolt.htm

Frank

[ArthurDent]

Until you change the parameters you are plotting to be the same as mine you can’t make a real comparison between your Tbolt and mine. First, the view you are using is set to 1 minute/division but you can see in my graph in post #4 that I have my view set to 120 minutes/division to show 44 hours of information that is much more meaningful.  Starting off you could try 15 minutes/division until you know if your Tbolt is working properly or not.

You also should enable plotting the 10 Mhz signal that I mentioned in another thread so you can check the health of the oscillator as well. Having the resolution of the PPS set too high only gives meaningless wild swings in the graph that are hard to understand.

It is easy to make these changes in Lady Heather and you will be able to interpret the displayed data easier and compare your graph to others.   

[notfaded1]

I wanted to post an update.  It's getting better.  I'm considering replacing the DS 1620 with a 9748C2 datecode.  This would, I think, improve the temperature regulation to 0.01 C steps.  I think the current DS 1620 in it has less resolution.

Bill

[ArthurDent]

I wouldn't change the DS1620. That has basically nothing to do with how the Tbolt functions but just reports the temp on the board, not the OCXO temp and until you make the following changes, you aren't sure it's bad. Scale of all parameters have to be meaningful as I've mentioned before. Change the T sensitivity to 500mC/div, 1pps to 20 ns/div, DAC to 200 uV/div, O to 500ppt/div, and elevation mask to 20 degrees before you do anything else.

[notfaded1]

I'm working on the elevation mask @AuthurDent.  I just downloaded TimbleVTS.

Bill