[GPS_Standard] Requirements: a low jitter Sine to Square-wave circuit

[Zack Widup]

Hi Dave,

On Tue, 26 Dec 2006, David Garnier wrote:

> Hello,
> 
> I have not thought this question out completely but maybe someone
> more knowledgeable on PLL's or interfacing sine-wave oscillators to
> TTL interfaces could answer these basic questions.
> 
> I have a Isotemp OCXO 134-10, 10 Mhz sine wave oven oscillator
> http://www.isotemp.com/docs/OCXO134-10Spec.pdf
> 
> 1) I need a Sine wave to TTL (circuit or IC) to interface this oscillator
> to an existing high quality RF signal generator. 
> 
> _Question._ It seems to me that this application would require a "low 
> jitter"
> sine to square-wave circuit. But, wouldn't the PLL in this RF signal 
> generator
> filter out this added jitter?
> 

The chip Bert uses is an LTC1485.  This is a Schmitt-trigger device, so I 
imagine the jitter is fairly low.

In another application, I was looking for a low phase-noise oscillator to 
drive an Analog Devices AD9835 DDS chip.  Analog has an application note, 
AN419, with just such a circuit.  It uses the MC10ELT21 chip as a low 
phase-noise sine-to-TTL converter. Digi-Key stocks these. These chips are 
good to at least 150 MHz!

> 2) _Inside the loop of a disciplined GPS oscillator_, is it less 
> critical to have
> a "low jitter" sine to square-wave converter?  ((Does the excess jitter 
> in this
> converter "come out in the wash" being within the disciplined GPS 
> oscillator?))
> 

It would depend on how narow the loop filter is.  The narrower the 
better, but the longer it takes to lock up.

In Bert's circuit I don't think it's critical but I haven't done any tests 
yet.

> 3) Measuring pico-second jitter isn't trivial.  Any ideas on how to do 
> this easily?
> 

Jitter translates into phase noise in the frequency domain.  The OCXO spec 
sheet you listed gives the phase noise for three frequencies, the close-in 
and the farther-out phase noise.  It looks respectable to me.

You will need a spectrum analyzer with the capability of looking at least 
145-150 dB below the carrier.

> 4) Maybe these questions cannot be initially answered but need elaborate
> "jitter measuring equipment" to make these determinations?
> 

The spectrum analyzer I mentioned above should do the trick.

> 5) There's some lab equipment available that maybe could help answer this
> question - but this is vacation time!!!
> 
> Thanks in advance for any thoughts you have on the subject.
> 
> Regards & 73's
> 
> Dave Garnier - wb9own

What are you going to use the circuit for?  I'm going to use it to either 
phase-lock or directly synthesize LO's for microwave transverters for 24, 
47, 76 GHz and up.  You lose out on phase noise each time you multiply the 
LO frequency, so when you get up around 76 GHz it becomes a factor to 
consider.  I'm trying to avoid spending as much as I'd spend for a new HF 
rig for a rubidium standard, as WA1ZMS did. We'll see how well it works. 
It should do OK at least on 24 and 47 GHz.

73, Zack W9SZ