On Sunday, 17 April 2022 07:47:40 EDT Torsten Curdt via Emc-developers wrote: > > > Maybe finding an old parallel port machine might still be the > > > easier > > > route > > > > > > :-/ > > > > Better yet, a pci-e buss and a 5i25 card. > > Unfortunately that circles back to the mesa availability problems :-(
5i25's are not available? I haven't checked recently. > Would a 7i92 also work? It looks like it would, its basicaly a 5i25 with a cat connector replacing the pci interface of the 5i25 and if the H version is available that would simplify cable making, db-25's have been problematic as the cable can work loose in the center of the cap, losing 1 or more connections. I have not had similar problems with the 26 pin header style and you can easily get both cables out to their respective bobs with only one style of common connector. That reduces you to only one db25 per cable unless you can find a bob that takes the 26 pin header. Those bob's are rare indeed. Because of the lack of termination resistors on a typical parport bob, if the cables are more than 5 feet, the data integrity would be improved by src terminating at the 7i92 by adding a nominally 120 ohm resistor in series as the ribbon cables impedance is nominally 120 to 130 ohms. The resistor absorbs a large part of the echo from the unterminated bob. Up to 5 feet, I don't worry about it. Note: I'm talking terminations not in the sense of screw terminal's but in terms of a transmission line's characteristic immpedance. VSWR IOW. However, unless your mobo has two ethernet ports, you'll have to disconnect the 7i92, plug in a cat6 to your switch/router and reconfigure the network to get online and update the pc. I consider that just as important, and explains why I went with a 7i90 on the pi, reflashed to an SPI interface. Which on the pi's, writes a 32 bit serial data packet to the 7i90 at 42 megabaud, and reads back at 25 megabaud. Apparently without interference from the net port. That could probably be done with a bash script, making it relatively painless. The huge disadvantage to that is that the 7i90 is those 50 pin scsi connectors are connected directly to the fpga, and needs 3 of the 7i42TA's to serve as fpga destroying noise absorbers, and gives you its 72 i/o's on the usual green termination facilities, while raising the cost to around $200 for all 4 pieces. The ease of wiring it up pays for the 7i42TA's IMO. At the time I built it, that was all that was available if you wanted a pi to talk to your machinery. That 7i90HD card can also be driven from a parport, and comes that way so must be reprogrammed with mesaflash to use the SPI. > Seems I either need to go with Remora or a Parallel BOB. > And I would probably have to get a PCI/PCIe Parallel Port. I don't get any hits for cnc stuff out of a DDG search for Remora. link url? > two parports lots faster than a mobo parport. > > > mobo = motherboard? Yup. > How come the 5i25 is faster? mobo parports are usually accessed thru the bios, which generally takes some number of microseconds just to wade thru the bios code and do it. The 5i25, sitting on the pci bus can update its outputs or read its inputs in sub-microsecond time frames. > cheers, > Torsten Take care and stay well. Cheers, Gene Heskett. -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author, 1940) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
