Paul__B:
blengon:
I think I will try with this two: http://tinyurl.com/lh5eeyq and http://tinyurl.com/l3ky8ufSo you have checked that the connections match? I have not looked into that so far - I have both Pro Minis and Nanos but have had some trouble with the Nanos so have avoided them for the present.
I have checked the description, and it looks like it can work, I have ordered it, as soon it is here, I will try to connect them and write here what I get 
Paul__B:
blengon:
But with the keyboard encoder: why it is difficult to interface, what could be the problem here? To my mind, it is maybe easier to interface the keyboard encoder then to solve the USB problems by hand ... but I could be wrongThe problem is that keyboard encoders are multiplexed - they activate one set of lines one after another and read another set of lines to see which crosspoints come up. To interface with such an arrangement you would have to determine which multiplex was active at any instant and provide a matching signal to correspond to the "virtual" keypress. That turns out to be extremely difficult to do. And a PC keyboard is at least a 10 by 12 grid - 22 lines - which explains why the chip you originally cited has so many legs.
I have understood it too and decided against the encoder, the direct combination of Pro Micro and USB Host Mini is also much more compact as to have 3 or more chips!