Can you tell what is the difference between Accuracy, Precision and Resolution?
Accuracy is how close a frequency or power measurement is to reality, whereas precision is how repeatable a measurement is within a certain error margin. Check here for more details.
In general terms, accuracy and precision are of the same magnitude in RF Explorer so we may specify one of them only.
Resolution is a function of the selected frequency span. The wider the span the lower the resolution.
What is the specific Frequency Accuracy in RF Explorer?
|RF Explorer Model||Frequency accuracy||Internal reference device|
|Signal Generator RFE6GEN||0.5ppm||Advanced 52MHz TCXO|
|6G Combo||0.5ppm in 6GHZ, 10ppm below 3GHz||Advanced 52MHz TCXO and 30MHz XTAL|
|6G||0.5ppm||Advanced 52MHz TCXO|
|3G Combo||10ppm||30MHz XTAL|
|ISM Combo||10ppm||26MHz and 30MHz XTAL|
Frequency accuracy of the TCXO used is +-0.5ppm, that is +-215Hz in 433Khz band or +-1.2Khz in 2.4Ghz band.
Frequency accuracy of the crystal used is +-10ppm, that is +-4.3Khz in 433Khz band or +-24Khz in 2.4Ghz band.
The aging factor of the XTAL is +-3ppm/year. The aging factor for the TCXO is 1ppm/year.
You can get detailed specifications of the crystals used below:
Unfortunately crystal tolerance is not the only factor that has influence in the initial accuracy of the unit. PCB parasitic capacitance and manufacturing tolerance plays an important role at these high frequencies. Due to this, you may expect up to +-100ppm initial value tolerance when uncalibrated for a crystal based device.
However, aging and stability is maintained within +-10ppm, in this case the crystal is the dominating factor (the parasitic capacitance is a manufacturing error and is not influenced by age or temperature).
TCXO based devices are inmune to PCB capacitance.
What about specific value for Frequency Resolution?
Regardless of accuracy, the most limiting factor is resolution. Understanding it will help to get the best value from your instrument.
There are 112 sweep steps in RF Explorer Spectrum Analyzer screen. Therefore, the maximum resolution you can get is FreqSpan/112.
In other words, for a frequency span of 10Mhz, you get a resolution of 10,000Khz/112=89.3Khz . With this span in the 433Mhz band, it is clear that 89.3Khz is a larger value than 4.3Khz and therefore the accuracy doesn't play as much as an important role as resolution.
If you now reduce the span to 200Khz, then resolution gets as fine as 200Khz/112=1.8Khz . In this case, the resolution is finer than accuracy and then accuracy is the main factor to consider.
What is the accuracy and resolution of power measurements?
Resolution is 0.5dBm for all bands.
Accuracy can be understood from three different angles:
- Absolute accuracy: this is the accuracy of dBm in absolute terms, and it has a typical value of +-5dBm when uncalibrated.
- Relative accuracy with constant RBW: also known as linearity. For a given span and Resolution BW, the relative accuracy is typically +-3dBm or better.
- Relative accuracy with variable RBW: If you measure the same signal with different RBW values (by changing the span), the relative accuracy may go between +-3dBm to +-10dBm typically.
Can you guarantee these values?
We can guarantee resolution but cannot guarantee accuracy.
To increase the accuracy and guarantee a specific range we would need to add a high quality calibration lab step in the production chain. If you are familiar with calibration techniques, you may know it will cost more than the RF Explorer itself.
Advanced users in need higher accuracy can use the frequency calibration feature available since firmware version 1.09, by following these simple instructions. A reliable RF source is the only requirement to calibrate RF Explorer anytime. Once calibrated it will typically exhibit a +-10ppm accuracy or better independently of PCB parasitic capacitance, actually +-5ppm is achievable.
We keep developing additional firmware features to offer additional calibration capabilities to reach sub-KHz accuracy and better amplitude accuracy in all bands.