@Jpspock Please copy-paste what you are referring to here (the text with that measurement and others). When talking about master and word clock specs, there are 3 main measurement points for phase noise quoted plus many other considerations to assess overall performance. The measurements are taken at 1Hz, 10Hz and 100Hz (and sometimes 1000MHz is quoted) with the units being "
dbc/Hz", not "db". Phase noise (?(f)) is typically expressed in units of dBc/Hz, and it
represents the noise power relative to the carrier contained in a 1 Hz bandwidth centered at certain offsets from the carrier....
For Cybershaft as an example (taken from their website):
"
It is not possible to evaluate the product by measuring only the output characteristics of the built-in OCXO unit. Our calibration measurements do not directly measure the output of the OCXO unit, but rather from the output terminal of the product in order to evaluate all performance including the buffer circuit (so measurement of the actual overall component, not just an OCXO assembly):
<Grade selection value> PN = Phase noise ADEV = Allan variance
| product grade | PN offset 1Hz | PN offset 10Hz | ADEV TAU=1sec |
|---|
| OP21A-D | < -121dBc/Hz | < -140dBc/Hz | < 1.6E-13 (0.00016ppb/s) |
*We have changed the previously published reference values for each grade to the screening values we actually use. All products already sold are shipped with the OP selection values in this table.
-The specifications of OP21E/-D are positioned as intermediate characteristics between OP21A and OP20A.
?As shown in [1] and [2], either PN or ADEV satisfies the specifications of OP21A.
?The overall characteristics are almost the same as [1] and [2]. [1] or [2] cannot be specified.
The above values are not general catalog values (typical) but guaranteed values measured individually, so the quality of the product is constant. There are no so-called "mistakes".
*For the value of phase noise offset 1Hz, an appropriate value from offset 1.000Hz to 1.099Hz is adopted to avoid averaging errors. Phase noise measurement is extremely sensitive and subject to measurement equipment and measurement errors. The phase noise values shown are not absolute values and include measurement errors. The phase noise values listed are values measured under the same conditions at our company, and are indicators of product grading.
The short-term stability value of Allan dispersion is not clearly defined by international standards, so there may be large errors between calibration laboratories. The above selection values are the values determined by our calibration equipment."
For the above, the numbers given are "standard values" (rounded) which are target certification values published for oscillators or this type. The test report on my 2017 OP21A OCXO Type2 model gives these exact numbers in the unit certification report (one compiled for each unit at time of final test before shipping):
< -122.1 dBc/Hz @ 1Hz offset actual (standard value is < -121 dBc/Hz)
< -142.1 dBc/Hz @ 10Hz offset actual (standard value is < -140 dBc/Hz)
< -147.4 dBc/Hz @ 100 Hz offset actual (no standard value noted)
Frequency value/stability must also be measured against a standard value of 9.999,999,995,00 MHz ? 10.000,000,005,00 MHz
with these devices over a prolonged period; to certify, there are several reference devices compared against including a Cesium frequency standard ("the standard" in the scientific world for clock accuracy and long-term stability).
Allan Deviation must also be taken into account and has standard values behavior at 0.1s, 1s, 10s intervals. Output level (from the BNC or other outputs) and Harmonic Distortion must also be measured and evaluated.
Allan Deviation, Phase Noise at the main measurement points and many others are used to calculate the overall performance of the master clock.
Looking at one number, no matter how impressive or a gap between numbers regardless of what it is does not allow the one device to be claimed as being superior (or not) to another.
I've included a Cybershaft mention and example text as I own their clocks over many years....this is not meant to spark any debates, only to help answer your question. Footnote: I have also owned Esoteric G-0s, followed by G-0Rb, then G-03X over many earlier years.
It would help quite a bit to see the rest of the text you are referencing and determine the overall context.
