> Good idea on the single unit thing.  I guess I can program the chip to
> demand the bus to itself ?-)

Only a fraction of the bandwidth.  That is what isochronous transfer mode
is.  You ask for a specific bandwidth and the controller either grants it
or tells you to stick it.  I would use bulk mode and just make sure
nothing else is active on the USB at the time.  That also alleviates the
need to phase lock your sample rate to the USB isochronous rate.

> As for the transfer, I think some chip with a buildt in fifo will do the
> trick.
> All three chips I'm looking at have this.

Yeah, how big?

> I mentioned oscilloscope as it will be possible to use the device as a
PC
> oscilloscope by incorperating triggering in the software instead of on
> chip/device.  It would make electronics cheaper and simpler and allow
for me
> playing around with the transfer.

In theory that should work as long as there is enough buffering everywhere
so that the USB and the CPU can catch up as bandwidth and cycles are
available.  I don't think this would be acceptable as a commercial product
because it requires too much of both the USB and the CPU, but for a
personal project it sounds interesting.  For a more robust approach, I
would put a DSP on the device to perform the trigger detection, then
upload only triggered traces.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body