Answering the questions from your last post:
- To do a quick test at +15dBm we chose PATABLE values to give roughly equal power-increases at each PATABLE step. SmartRF studio wouldn't produce any PATABLE values (other than at PATABLE0)... so we had a guess here assuming they should be indicative.
[PATABLE = 0x6F, 0x00, 0x6E, 0x30, 0x13, 0x0F, 0x6D, 0x27]
We believe the increased noise to be a problem with the CC1101/CC1190 when performing the EN300220-1 transient power test (for the reasons below).
The most recent revisions of EN300220-1 (2.3.1 and 2.4.1) both specify a different measurement technique from previous revisions. We have studied (in detail) AN050/swra146b and AN094/swra356. In particular:
- AN094 suggests the CC1101/CC1190 EM 868MHz reference design is compliant with EN 300 220-1 (v2.3.1)
- AN094 does not have any mention of Transient Power test compliance for the CC1101/CC1190 868MHz reference design (EM board).
- AN050 is for the CC1101 (transceiver alone); we note this uses the *old* Transient Power measurement method [based on readings at 2, 4 & 10x the channel-spacing]
In the new ETSI specification, the Transient Power is measured at 100kHz offset from the Modulation BW edges (and at 120kHz steps up to 2MHz). Also, quasi-peak detection with a 120KHz RBW is specified for these measurements. AN050 (section 4.1.3) suggests "...an RBW of 10 kHz, is a conservative choice in terms of meeting the filter characteristics defined in EN 300 220-1, annex B [1]..." but, there is a vast difference (upwards of 20dB) in the measured transient power level between these two RBW settings (as seen on our Anritsu MT8222A spectrum analyzer).
We believe we're probably missing some key point here :- but wonder if you might be able to explain where our interpretation might be incorrect?
Thanks very much in advance for your help once again.