By Daniel Sandu, YO3GJC/VE3GNO
May 24, 2010
To determine the IMDDR3 performance level of the Network Sciences filter fitted to a Yaesu FT-2000
- Essentially the same as documented in the Simple Install process from AC0C.
- From mixer via splitter (see notes) into the NS filter.
- Out of the mixer, into the prior post filter matching transformer.
- No special z match was attempted to establish a baseline of minimum performance under the simple install process
Table note: F3-, F3+ refer to the DR3 products below and above the two tones, respectivly.
Prototype SN001 improves 2 Khz spacing DR3 20-21db over the stock 3 Khz filter.
- Measurements were made following ARRL “Proceedure Manual, 2010”; Section 5.8 excluding the Gray amps which are not needed at this DUT's range of DR3 performance
- Test band is 20m
- Connecting cables by Huber Suhner
- All equipment used has current calibration certification
- From a prior modification, mixer output is split through a MC SMT combiner with one side feeding the input relays and the other feeding the DMU. This modification (post mixer splitter) is not beleived to have a significant impact in the overall DR3 performance. However, it does provide a flatter grass level due to the port-to-port combiner isolation. See my prior work on the subject for modification details.
- All testing done with the CW-U Lower/Upper F3 refers to the IMD sideband location. Lower is the spike below the two pairs in frequency. Higher is the spike above the two pairs in frequency.
- Filter: Network Sciences prototype SN001
- Data are close to being phase limited. With further tightness in the filter slope, an upper limit of perhaps 90 db may be possible on this rig before RMN becomes dominant. In a quick test of the impact of the mixer and LO quality, I temporarly bypassed the mixer with a stand alone MC external unit using the rig's LO and saw no improvement. Using an external very clean LO source (HP 8267D VSG) and the MC VALY1 mixer, the DR3 numbers were better yet. There is a path for even more improvement, but that would require work on the LO - and that was beyond the scope of the current evaluation objectives.