forum.pistar.uk

KG6RUT wrote: ↑Thu Jan 11, 2024 3:44 pm that hat looks like its using an on-board antenna, as I dont see any antenna mounts.....

Yep, on board antennas marked TX and RX.

I’ve seen M0FXB talked about it in his YouTube channel. I’ve just realized, he never used it in duplex mode.

Maybe this is a big design flaw and no one ever made it work in duplex.

You need to work out the link budgets for Tx and Rx, assume 0dB isolation between the antennas and use the ADF7021 phase noise plots to estimate the energy from the Tx at the Rx frequency. You also need to consider the blocking performance of the Rx against the Tx signal at +13dBm or similar.

Full duplex designs only work when the unwanted RF energy that the Rx sees is low enough to allow the SNR of the wanted signal to be adequate.

G8SEZ wrote: ↑Thu Jan 11, 2024 8:19 pm You need to work out the link budgets for Tx and Rx, assume 0dB isolation between the antennas and use the ADF7021 phase noise plots to estimate the energy from the Tx at the Rx frequency. You also need to consider the blocking performance of the Rx against the Tx signal at +13dBm or similar.

Full duplex designs only work when the unwanted RF energy that the Rx sees is low enough to allow the SNR of the wanted signal to be adequate.

Wait a second, I thought this is a Half Duplex Hat. Is it a full duplex hat? If it is, it doesn't make sense on the design perspective in the first place.

I’ve tried 30MHz split today, a little bit better, but still unsatisfied. Might have to try cross band.

By the way, is it possible to make it simplex but have the ability to use both time slots?

G8SEZ wrote: ↑Thu Jan 11, 2024 8:19 pm You need to work out the link budgets for Tx and Rx, assume 0dB isolation between the antennas and use the ADF7021 phase noise plots to estimate the energy from the Tx at the Rx frequency. You also need to consider the blocking performance of the Rx against the Tx signal at +13dBm or similar.

Full duplex designs only work when the unwanted RF energy that the Rx sees is low enough to allow the SNR of the wanted signal to be adequate.

Do you have a link to a page about this?

KG6RUT wrote: ↑Thu Jan 11, 2024 3:44 pm that hat looks like its using an on-board antenna, as I dont see any antenna mounts.....

The board has solder pads for SMA connectors, they are adjacent to the on-board antennae.

Not sure that the configuration of the board, as sold, is the most usable / user friendly. If you must run that as a duplex configuration, I'd try turning down the transmit power and using a very very wide split, FWIW, but not sure that the improvement will be tremendous.

KB1IEE wrote: ↑Fri Jan 12, 2024 9:02 pm

KG6RUT wrote: ↑Thu Jan 11, 2024 3:44 pm that hat looks like its using an on-board antenna, as I dont see any antenna mounts.....

The board has solder pads for SMA connectors, they are adjacent to the on-board antennae.

Not sure that the configuration of the board, as sold, is the most usable / user friendly. If you must run that as a duplex configuration, I'd try turning down the transmit power and using a very very wide split, FWIW, but not sure that the improvement will be tremendous.

I've tried lower TX power to 10 (1mW if I'm not mistaken) and using 30MHz split, a little better, still not usable.

ka1cm wrote: ↑Fri Jan 12, 2024 6:25 pm

G8SEZ wrote: ↑Thu Jan 11, 2024 8:19 pm You need to work out the link budgets for Tx and Rx, assume 0dB isolation between the antennas and use the ADF7021 phase noise plots to estimate the energy from the Tx at the Rx frequency. You also need to consider the blocking performance of the Rx against the Tx signal at +13dBm or similar.

Full duplex designs only work when the unwanted RF energy that the Rx sees is low enough to allow the SNR of the wanted signal to be adequate.

Do you have a link to a page about this?

Well, not that I can find immediately, but you should be able to find something about reciprocal mixing (where a strong received signal mixes with the LO noise sidebands and appears in the IF/baseband) and then apply the explanation to Tx noise sidebands entering a receiver at a given offset.

I see you mentioned that 30MHz offset helps, but your fundamental limitation is that the 2 surface mounted PCB antennas share a ground plane which removes most or all of the potential isolation. In a normal repeater with a single antenna it requires something like a 6-cavity diplexer to isolate Tx and Rx, and sometimes you need an additional isolator on the Tx to prevent intermodulation from strong off-air signals mixing in the PA.