Hello Martijn, Friday, May 11, 2001, 11:32:41 PM, you wrote: MvB> 2) Re-reading old discussions, I stumbled across a comment by William MvB> Levak, claiming that untalk is used as a substitute for EOI on the MvB> serial bus. At the controller side. MvB> Somehow, I don't really understand this comment, unless MvB> my ideas about untalk and EOI are off. However, I browsed through MvB> the sources of Ruud's PCDISK, and they seem to correspond to my ideas. MvB> How can Untalk be a substitute for EOI? Consider a device transmitting MvB> data through the bus. If the controller wants the data stream to stop, MvB> it transmits UNTALK; if the transmitting device is out of data, it MvB> sends EOI. To me, these things are something entirely different, or MvB> am i wrong ? Why? It is clear for me. MvB> Can a controller 'end' a transmission by asserting EOI MvB> as well, parallel to the talker? What you mean: IEC or IEEE? On the IEC EOI transmits only devices TO controller. Controller simply sends UNTALK or UNLISTEN when neccessary. MvB> Related to this: What is the use of UNLISTEN? Isn't that exactly MvB> the same as doing an EOI? UNTalk and UNListen commands on the serial bus are for stopping TALKER and LISTENER respectively, so they can change the direction of transfer after that command (well after LISTEN and UNLISTEN). MvB> 3) Consider a serial talker, to an IEEE-listener. The talker happily MvB> starts transmitting data, which I gateway to the IEEE bus. If, at a MvB> given time, the controller had enough, and asserts ATN, this will MvB> probably be in the middle of a byte being transfer on IEC. How should MvB> I handle this? In a 'best case scenario' :), this will result in the MvB> IEC talker 'talking' one byte too much. Don't forget that IECOUT it is alyways buffering a byte forward. Probably diagrams can help you find the right answer for any question. Nick did an IEC oscilloscope prg, so I hope he analyzed the IEC timings quit many time. -- Best regards, Frank mailto:frank@kontros.uzhgorod.ua - This message was sent through the cbm-hackers mailing list. To unsubscribe: echo unsubscribe | mail cbm-hackers-request@dot.tml.hut.fi.
Archive generated by hypermail 2.1.1.