Sorry just joined this forum so catching up on all the interesting queries. I originally developed this at GSK on EVOs with the Trinian system and rolled this out to EVOs and Fluents with integrated Lunatic.
It is a bit trickly as the liquid class needs to be optimised in a very non-natural way. Manually they use 2ul but the minimum we managed while maintaining accuracy of results was with 3ul. You get worst CV the high volume you use as it floods the channels.
So they need to dispense very slowly to create a 3ul droplet of sample on the end of the tip. This should be done just within the small well. You then have to introduce a side well touch to allow surface tension to pull the droplet of the tip, before retracting.
If you try dispensing too low volume it end up forcing sample past the priming channel and you get a lot of fails. So the side well touch is necessary to let the surface tension naturally fill the priming channel.
Took me a long time to work this out but was rarely getting a well fail so this was rolled out across a number for EVO and Fluent systems at GSK and Thermo integrated systems. It is not a trade secret just painful know how.
Hope this helps, I think I have a liquid class tucked away but might take me a while to dig out?