Am 25.02.19 um 23:07 schrieb andy pugh:
> Is this definitely not a type that can be supported by the existing 14
> patterns?

yes, it is so....
(or I am to blind, to see it)


> If so, what is it?

I'll try to explain it:

the signal at the output-pin is a square - in every type.
because a usual output pin is stupid....

but raspberry or andruino may be smarter.

instead of using a low-high-low square. it  is able to use
a rising or falling edge and "translate" it into a low-high-low
square (for the TA6600 as an example).
so after setting the pin from 0 to 1,  stepgen.c must not wait
"timer1"-time and lower the signal again after his time.
the state of the pin does not matter - only the changing of its
level causes the raspi/arduino-board to generate its own
low-high-low signal.
if the signal is serialized, like in an eth-telegram, with the same
amount of data per time, the output-frequency could be doubled.
because the driver must not send:
low-high-(wait-time)-low - for one step, it only has to change the
logic-leve once.


> I am a bit unclear what you mean by "Raspberry / Arduino driver" given
> the extreme differences between the two.


yes, you are right. but he hardware on the receiving side of the
raw-eth-driver does not matter. it only must be able to create a
0___1___0 signal from an rising or falling edge of the stepgen
output.
I am also thinking about using such a kind of driver on other
hardware with eth-interfaces.


I hope, this makes it a little bit more clear.....


regards

wicki





_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to