65 Mhz jitter measurement

[Antonio Pasini]

I'd like to measure jitter (cycle by cycle, rms, pk-pk) on a LVCMOS
oscillator, 65 Mhz.

I have is a Tek 3054B...

I didn't found a clear explanation on tek website of the options I could
purchase, if any...

What could I use ?

Low cost is a concern...

 

[Bill Janssen]

I'd like to measure jitter (cycle by cycle, rms, pk-pk) on a LVCMOS
oscillator, 65 Mhz.

I have is a Tek 3054B...

I didn't found a clear explanation on tek website of the options I could
purchase, if any...

What could I use ?

Low cost is a concern...

One way is to compare a delayed signal and the non delayed signal in a
mixer. The output of the mixer
is the jitter signal.

Use a splitter to get two signals with the same phase. Use a section of
able to delay one
signal 45 degrees. Compare those signals using a double balanced mixer.
The output signal
the can be displayed on a scope or AC meter etc.

Good luck.

Bill K7NOM

 

[JS]

I'd like to measure jitter (cycle by cycle, rms, pk-pk) on a LVCMOS
oscillator, 65 Mhz.

I have is a Tek 3054B...

I didn't found a clear explanation on tek website of the options I could
purchase, if any...

What could I use ?

Low cost is a concern...

The 3054B has a very short memory and no statistics on the measurements
it takes. It has a pretty high trigger jitter - around 300 ps or greater
which is why the cleanest signal gets pretty wide and thick when you
watch it with persistence turned on.

Perhaps you can rent or borrow a scope already programmed to take
jitter measurements?

Some of the concerns and recent developments related to jitter
measurements are posted at http://www.qsl.net/wd1v in the Scope FAQ
section.

Best regards,

John

 

[Antonio Pasini]

I'd like to measure jitter (cycle by cycle, rms, pk-pk) on a LVCMOS

The 3054B has a very short memory and no statistics on the measurements
it takes. It has a pretty high trigger jitter - around 300 ps or greater
which is why the cleanest signal gets pretty wide and thick when you
watch it with persistence turned on.

Perhaps you can rent or borrow a scope already programmed to take
jitter measurements?

Agree, will be the best solution. I'm already looking around...
The http://www.qsl.net/wd1v link is very interesting, also.

Thanks!

 

[ddwyer]

Agree, will be the best solution. I'm already looking around...
The http://www.qsl.net/wd1v link is very interesting, also.

Thanks!

For the purposes of comparison I would suggest a frequency domain
measurements i.e. a synthesised receiver or spectrum analyzer or perhaps
just a modulation meter.
Look at the residual fm.
conversion from frequency to time domain to be made employing Alan
Variance.
See mr Allen's excellent web page