MMDS converters have arrived

About two weeks ago I decided that my setup for S-band reception needed to be upgraded seriously.

So I looked into possibilities for improvements. I found the filters necessary for the AR8200 receiver to work, and looked around to see if I could find some converters that were (relatively) easily modifiable.

I think I found it . MMDS converters cover the bands around 2150 MHz and around 2600 MHz, and it should be possible to modify the filters

TranSystem Inc makes some MMDS converters, I found some on ebay, TranSystem Model EIDC 3033 Down Converter, apparently with the following spec :

RF bands :
2150 - 2162 MHz and
2500 - 2682 MHz,
Intermediate frequencies :
116 - 128 MHz and
222 - 408 MHz

This is possible with a LO frequency of 2278 MHz. Since I want to use the converter for the 2200 - 2300 MHz band, some modifications are necessary :

RF filters - one pair at the input and one pair between the RF amplifier and the mixer - must be modified to cover 2200 - 2300 MHz

The Local Oscillator (LO) needs to be moved away from the wanted passband, preferably for a low side LO.

I took a look at the inner workings of the converter, and the RF frequency filters are stripline filters made of copper with air insulation (not microstrip etched on the PCB), so they are expected to be fairly high Q filters. For satellite S-band I think the best strategy will be to shorten the 2150 MHz strips (careful - we do not want to get too high in frequency), then disconnect the higher frequency filter (hmmm - that may not even be necessary).

The other modification concerns moving the LO down. The LO is a PLL with a frequency divider (256x) 2278MHz down to 8.898438 MHz. (Xtal in the reference oscillator).

It would be nice to have the oscillator running on a "rounded" frequency like 2000 or 2100 MHz, but that would require new Xtals to be made. Since I would like to be able to lock the LO frequency to a stable source that complicates things.

The other option will be to use a 8 MHz crystal oscillator, then lock that to a 10MHz TCXO or other standard. This will provide the converter with a LO frequency of 2048MHz. not exactly very "rounded", but still with a full MHZ, so the readout of the converted frequency should not be all too confusing (that remains to be seen). After all, a computer can do wonders in calculating the correct frequencies and control the receiver(s).

Time is a bit tight this week end, but I hope to be looking into it after all.