Re: [avr-chat] FTDI232 chips & Missing data

[Graham Davies]

"Erik Christiansen wrote:

> The experiment seems to leave the bridge or wince box as the weak link.

Not necessarily.  AFAIK the nature of USB is to gather octets up in buffers
and then burst them over the cable so they show up in a bunch at the other
end.  The nature of start/stop serial transmission is that the octets show
up well spaced out.  This could make a difference to a poorly implemented
receiver.

> We've had similar serial data loss problems when (other team members ;)
> have used Hyperterm on the PC end. Changing to Tera Term eliminated the
> data loss.

There you go.  Hyperterm and poorly implemented are the same thing.  No end
of serial data problems clear up if you throw out Hyperterm and generally
you can trace them back to data overwhelming a poor receiver.

Graham.