Hallo Leute!
I'm very glad see opinion developers here (sorry, LDS forum, but I will talk about hardware).
RIYA USB and bILDA have many differents in architecture.
Both type of controllers should make two process simultaneously:
- receiving data from PC;
- sending data to DACs.
RIYA use FPGA (it allows make realy multiprocessing); bILDA - Enhanced 8051 microcontroller (it must make two these process yourself)
Next, buffering of data:
- RIYA have 128K RAM on board, which allows us make two buffers for 4 000 points (16 bytes for point). So, software in our case forme full frame, and send it to our controller as once.
- bILDA can use for prebuffering only internal RAM of microcontroller AN2135SC, which size is 8K. So, it impossible have buffer for full frame 16 000 point (8 byte for point) in RAM.
Therefore, is not fuul correct afrob's replie:
...bILDA hat ein framebasiertes Treiberinterface mit doppelter Pufferung, ganz ähnlich wie bei der Riya Hardware. Schon immer gehabt. Anders funktioniert das auch gar nicht bei >=50k. Da bILDA aber das RAM des PCs benutzt können sogar wesentlich mehr Punkte zwischengespeichert werden als bei der Riya Hardware.
About scanrate:
If we declare 50 000 PPS, it means, that:
- time translation of point-to-point in one block (in our case, one block is full frame) is 20 msec
- time of translation block-to-blocK (in our case it is time between last point of previous frame and first point of next frame) is 20 msec too.
IN BOTH CASES WE GUARANTEE ERROR NOT MORE 30 NANOSECONDS!
@AFROB, can you write us these timing parameters of your controller?
And about Linux:
we begin from drivers for Windows (98, 2000, NT, XP).
One year ago some person open discussion that they will buy our PCI device if we provide support for Linux. We have done it but peoples didn't buy anything they didn't say even "Thanks".
It is our really situation with Linux now
All the best
Ihor