Trimble thunderbolt or HP Z3805A

[AlfBaz]

Let me start by saying that I am a complete noob at this time nuttery stuff, so be patient with me

I have the opportunity to purchase a GPSDO. Unfortunately there are quite a number of options.
I have considered some of the BG7TBL units but SHMBO has given me permission to splurge, so now I'm looking at either a Trimble thunderbolt or Z3805A

So far, from what I've been able to find, the major difference between the two is configurability. Lady heather and the thunderbolt provide oodles of adjustable paramaters where as the Z3805A is more a set and forget unit.
Also the Thunderbolt seems to have a better GPS front end so it may be able to find and track more satellites. Is this a factor worth considering?

The OCXO in the Z3805A looks to be the better of the two but I am not certain what, if any, are the overall effects on the final reference output. I've also read that disciplining good OCXO's can actually degrade their performance

Whilst long term stability is important to me, phase noise and jitter are equally important to me and to that end I'll end up using the GPSDO to cal a decent OCXO somewhere down the track

If given a choice between the two units what would some of the more time nutty members here choose?

[Deathwish]

Im not a time nut, but I think you answered it for yourself, thunderbolt.

[TSL]

This is a question that has been asked many times before on various forums and often on the TimeNuts list.

The common answer is archived here...

http://ko4bb.com/dokuwiki/doku.php?id=precision_timing:comparing_thunderbolt_and_z3805a_gps_stabilized_crystal_oscillators

[TSL]

Whilst long term stability is important to me, phase noise and jitter are equally important to me and to that end I'll end up using the GPSDO to cal a decent OCXO somewhere down the track

I should ask - If you're only going to cal an OCXO why is phase noise and jitter important to you ? neither will really affect the ability to cal an OCXO.

FYI... jitter is the integral of phase noise and there's plenty of calcs on the net for you to covert between the two. Both are short term quantities, one in the time domain the other in the frequency domain.

Phase noise becomes important for local oscillators used in microwave radios since it affects the MDS. Stability at those frequencies becomes the other important quantity.

Knowing what you want to do beyond cal of an OCXO will help to point you in the right direction.

All this being said, both the Z and the Tbolt are exceptional units with exceptional specification and it would be hard for most people to get themselves into a situation where neither was adequate !

[AlfBaz]

This is a question that has been asked many times before on various forums and often on the TimeNuts list.

The common answer is archived here...

http://ko4bb.com/dokuwiki/doku.php?id=precision_timing:comparing_thunderbolt_and_z3805a_gps_stabilized_crystal_oscillators

I have read that one. Assuming it's the same Bob, he goes a littler further the 3rd post I found here, putting the thunderbolt slightly in front
http://comments.gmane.org/gmane.comp.time.nuts/4403

I should ask - If you're only going to cal an OCXO why is phase noise and jitter important to you ? neither will really affect the ability to cal an OCXO.

Sorry, I didn't make myself clear. I am lead believe that a good OCXO has less phase noise and jitter than a GPSDO. If this is true then I can use the GPSDO as an accurate time base and the OCXO if I need low jitter and phase noise

FYI... jitter is the integral of phase noise and there's plenty of calcs on the net for you to covert between the two. Both are short term quantities, one in the time domain the other in the frequency domain.

Phase noise becomes important for local oscillators used in microwave radios since it affects the MDS. Stability at those frequencies becomes the other important quantity.

Yes, I still need to read up on phase noise specs, at a glance they don't seem intuitive, to me at least

Knowing what you want to do beyond cal of an OCXO will help to point you in the right direction.

All this being said, both the Z and the Tbolt are exceptional units with exceptional specification and it would be hard for most people to get themselves into a situation where neither was adequate !

The "what do I want to do with it" question is inevitable however it is difficult to tell at this point as I dabble more in the high speed digital and RF world. It would be nice if a few months down the track I don't say "damn I should of gotten the other one". This is why I ask what would you get if you had the choice

[DimitriP]

The TB will be more fun to play with.
Granted you can only change the antenna, software parameters and the OCXO but that should provide many days (weeks/months/years) of fun .
Chances are if you buy or build one GPSDO, you'll probably end up buying /building one more.

http://www.ke5fx.com/gpscomp.htm
http://www.ke5fx.com/tbolt.htm
 

[TSL]

I have read that one. Assuming it's the same Bob, he goes a littler further the 3rd post I found here, putting the thunderbolt slightly in front
http://comments.gmane.org/gmane.comp.time.nuts/4403

Yes - same Bob - the Thunderbolt is slightly in front because it has many settings that can be tuned to provide truly exceptional performance.

Sorry, I didn't make myself clear. I am lead believe that a good OCXO has less phase noise and jitter than a GPSDO. If this is true then I can use the GPSDO as an accurate time base and the OCXO if I need low jitter and phase noise

A GPS  driving an OCXO maks it an GPSDO - the phase noise is then directly related to the quality of the OCXO being driven. This only relates to the 10Mhz output. The PPS out is often a direct function of the GPS engine although some GPSDO divide the 10MHz down to provide the PPS in an effort to minimize the inherent 20ns jitter in any GPS system.

The "what do I want to do with it" question is inevitable however it is difficult to tell at this point as I dabble more in the high speed digital and RF world. It would be nice if a few months down the track I don't say "damn I should of gotten the other one". This is why I ask what would you get if you had the choice

Fair enough

Given digital, you might want low jitter PPS to form the basis of digital clocking, alternatively you might divide or multiply a low phase noise 10MHz output into the frequency you want.

Given RF, you most certainly want low phase noise especially if you are going to use the 10MHz signal to lock a Spectrum Analyzer since that will probably affect the performance of the SA. This depends on the specs of the analyzer though.

Likewise if your dabbling in RF into the microwave spectrum both phase noise and stability become the issue.

The OCXO in the Z you've pointed at is the venerable HP 10811. This is an exceptional low phase noise oscillator and the HP 10811 can be retrofitted to a thunderbolt.

More info and test results here...
http://www.ke5fx.com/tbolt.htm

For exceptionally low phase noise and jitter its hard to go past the venerable Rubidium Oscillator PRS-10. This Rb has the ability to be guided by a PPS output of a timing GPSDO. This gives you the best possible outcome for all three quantities of accuracy, stability, and phase noise.

The PRS-10's come out of Symetricomm primary standards which couple GPS's with RB. The only thing better than these are Cesium or Hydrogen Maser standards.

Time Nuttery is a dangerous path to walk down as you start to become obsessed about pico seconds of difference or phase noise above -100dBc

All in all its a fun thing to play with

 

[AlfBaz]

Thanks very much for that post TSL
Lots of new info in there for me to digest

[bingo600]

Hi AB

I'd go for the Tbolt.

Tbolt:
http://www.prc68.com/I/ThunderBolt.shtml
Smaller , emits less heat , super tweakable w. Heather , No hanging bridges (by design)


PRS10: (But they wear out)  , i have one
http://www.ke5fx.com/rb.htm
http://www.ebay.com/itm/321861322781


Z3805:
I don't have a Z3805 , but a Z3801 , and it's quite ok.
But it uses 48v and emits a lot of heat.

Note that "some" says that the Z3805's sold by "yixunhk" , have "unoriginal" HP enclosures.
Indicating that the inside might NOT be corresponding to the "Newer enclosure".


Watch out for timenutting, the next you want is a HP-5370B or a SR620.
And then a PRS10 to verify your tbolt , and .........

/Bingo

[bingo600]

Re: Tbolts

If you get a late Rev E model (Not to be confused with the RED Tbolt E) , you might want to change the temperature sensor chip.
I did replace it on my latest two, ith a couple of Rev C chips.
Some time-nuts says the new "E" chip doesn't have much influence on the stability though.

I had a thunderbolt that displayed a temperature reading of -54.99°C, whch
I was pretty sure wasn't correct. ;-)   Apparently the reference in th DS1620
chip either shorted or opened so this "D" revision chip was outputting its
lowest possible reading. I replaced it with a "C" revision chip that I foud
(date code 9710C2) and it works great.

So If you have an "E" revision DS1620 chip that displays the temperature i steps
and you want to replace it, either a "C" or a "D" revision chip appears towork
as a replacement. Trimble apparently started using the "E" revision chip aound
mid 2004.   

This one i saw a few years back , he might still have Rev C's.

Alternately, I have a large quantity of DS1620S rev C chips.  $1.99 ea +
shipping. Yes, I ship worldwide.

http://standingwave.org/webdav/maximdsst-1264178984-1808.jpg

http://standingwave.org/

It's en easy mod

/Bingo

[DimitriP]

Watch out for timenutting, the next you want is a HP-5370B or a SR620.

Mmmmmm..........

[AlfBaz]

Hey Bingo, thanks for the tips regarding the tbolt. I never knew about the temperature sensor issue. I think I'll go ahead and order the thunderbolt

[Xtremexp]

Hi,

I am debating myself which tbolt I should buy, blue tbolt E, red tbolt E or the tbolt without labels. by looking date code blue tbolt E is manufactured on 2014, is this good enough for experimenting with GPSDO?

Thanks!!