Nose to Nose High Frequency RF 50GHz Oscilloscope Calibration

[AnalogDigital]

Hi all,

this is my 1st post, so I hope I do not break any rules  .
However, any of you have ever had to deal with the "nose to nose calibration" for sampling oscilloscopes? I read a lot of theory on this topic, I am able to calibrate many types of oscilloscopes and VNA, but I never got a satisfactory and reliable results with this type of calibration.

thanks for sharing your knowledge and your experiences!

PS. If need, I have free unlimited access to most articles on http://ieeexplore.ieee.org/ 

G

EDIT: edited title to be more search-friendly...

[OscarH]

Hi,
I had been running through the calibration process of oscilloscopes for years (on top of many other items like Network, sources and Spectrum analysers) but had never seen this measurment  methodology used, at least in my company.
In using the Nose-to-Nose calibration, my understanding is that main objective is to avoid using expensive test equipment and rely on a second scope reading and math to estimate characteristics.
On top, it is apparently no working so well (your experience !). I would not like to be the metrologist guy to go into the measurement uncertainties calculation with this model.
This is (may be) the reason why it is not widely used, i.e. giving too much uncertainties on the measurement itself, or simply impossible to have a real estimate of the uncertainty.
Cheers.
OH

[AnalogDigital]

Thanks for your fast and clear reply.
Well, this is my situation:

I have in my homemade lab:

• 2x Oscilloscopes HP54124T
• 1x Power Meter (can't remember model )
• 1x VNA HP8510
• many couplers, cables, connectors, attenuator and this kind of stuff.

following what is written on this, mesaure error after calibration is too high, expecially for phase error.
I can successfully calibrate the VNA ( SOLT or TRL ), both the power meter.
How would you advise me to calibrate oscilloscopes with this material?
How would you advise me to calibrate oscilloscopes in general?

Thanks again for sharing your knowledge,

G

[OscarH]

Hi,
The most important parameter for scopes in general are amplitude accuracy and time base accuracy.
Not sure what config you have for your 8510, but I you can use the 8510 source to calibrate the frequency/timebase accuracy. The source internal 10 MHz is far mor accurate than the scope, so no need to use external counter.
You also need to compare time base accuracy versus channel, in using a power splitter in reference to chan 1.
For amplitude, you need to use the source + power-meter and compare versus scope measurement.
If amplitude and time base results are ok, you can assume the scope is probably ok. If one is not ok, then is is unlikely to have a failure on the scope or an adjustment needed.

54124T is a 50 GHz scope on upper channels. Depending on source frequency you have, you may not be able to test up to this frequency (is that important to you ?).
Also, as the 8510 source cannot go below 10 MHz, you may need another source + multimeter to test scope amplitude below 10 MHz.

Just to be clear, I'm talking here about verification procedure. If there is an adjustment needed, it is unlikely you can make it (other equipment or adjustment procedure not available).
Hope this help.
Cheers.
OH

[Gallymimu]

I'd say timebase accuracy is by far the most important.  Scopes generally have pretty poor vertical accuracy and offset even when calibrated.

[AnalogDigital]

well, after a few days of testing, I was able to get a satisfactory calibration. It was a great idea to use the source of the VNA to calibrate the oscilloscope ... 

[OscarH]

I'm glad you had been able to get a satisfactory calibration.
Not sure what version of source you have with you network, but I know on some of them you can do a flatness calibration with the VNA, thus get a really flat power output to be use to test the scope.
OH.