Yaesu FTDX-10 1st mixer

[UR5FFR]

Hello,

Analyzing the Yaesu FTDX-10 circuit, I discovered that the first mixer has an original design. Typically, mixers use transformers to separate the signal into antiphase components. But this mixer uses a splitter. In the appendix, I have added the original mixer circuit and its simplified version, more suitable for analysis.
I have not come across a similar mixer design. What advantages does it offer compared to using a conventional transformer at the input?

PS It is interesting that the Yaesu FTDX-101 transceiver uses a mixer with the same structure and splitter at the input, but it is active and instead of digital switch FTS3257, dual-gate field-effect transistors are used.

[M0HZH]

I'm curious about this as well. FTDX-10 and 101 have exceptional dynamic range and to achieve that the 1st mixer is critical; probably this layout allows for better performance?

PA3AKE's research shows the transformer makes a big difference in mixer performance:

https://martein.home.xs4all.nl/pa3ake/hmode/hmode_intro.html

[UR5FFR]

Other non standard in FTDX-10/101 schematic is amplifier between BPF and 1st mixer. Usually this amplifier manually turn on/off. But not here. In FTDX this amplifier can't be turn off - it always working. In FTDX-10 it schematic is very simple - Norton amplifier with BFQ19 at 30mA collector current

[mag_therm]

About your equivalent circuit:
I think the ratios and mutual inductances would be needed for the T3004 to see what is going on there.
It looks like the old FM "Quadrature detector", the outputs are the voltage on input tuned circuit, +/- the centre tapped windings which are in quadrature.
Vector diagrams needed.

A question if I may..

I am testing an ft8 demodulator here. ( First RF "design" project!)
Using MC1496 Gilbert cell with Epson SG8018 carrier oscillators  on 458kHz / 452 kHz IF it seems to be working well.
I had some trouble with resonant input coupling transformer but now trying an untuned  toroid single ended to diff with windings diametrically spaced 12mm.
( Low mutual C)
Ratio 1:1 and scaling so that -10dBm IF  is about -20dBfs audio.

Next project is to see if I can use the similar topology on converters for 30 and 20 meter with dedicated XO's  just for Ft8 per band.
Maybe mast head converters with  double conversion to Low IF to computer
The proposed XO are Epson SG8018 .
The jitter is quoted in various peak to peak and rms values.
For example
At 3.3V supply, 10 MHz
Offset Frequency 12 kHz to 20 MHz : peak to peak jitter = 66.9 ps
Can be converted to dB, but I would not be wiser.

What is a figure of merit for mixer oscillator's jitter receiving ft8?

-thanks