stephan@cup.portal.com (Stephen Derek Schaem) (07/11/89)
I need information about the lucas and the supra 4x4 scsi controller. don't have a 68881 for the lucas board (working at 25mhz like a charm, 74F74) so I don't the little hack of connecting pin 7 and 15of U7. The problem is that the supra controller is not 'working' like it wasn't even there... It it cause by the U7 'trick' or simply or having the lucas board? Also, How will Fransis react with lucas at 25mhz? I want to keep it at 25mhz since I really saw a difference between 20 and 25mhz, and everything that work at 16mhz work at 25.
douglee@becker.UUCP (Doug Lee) (07/13/89)
In article <20321@cup.portal.com> stephan@cup.portal.com (Stephen Derek Schaem) writes: > > I need information about the lucas and the supra 4x4 scsi > controller. > don't have a 68881 for the lucas board (working at 25mhz like a > charm, 74F74) so I don't the little hack of connecting pin 7 and > 15of U7. > > The problem is that the supra controller is not 'working' > like it wasn't even there... It it cause by the U7 'trick' or > simply or having the lucas board? > I had a Supra controller working with the Lucas board, but it would not work reliably at 20 Mhz. It would work at 16 though. You may have to slow things down. You can also try different chips in the U9 position. On my particular machine and many others, I have had best luck with a National 74LS74AN. Note that the brand is critical too. You should try LS, F, ALS and plain 7474 parts from DIFFERENT manufacturers. I have found that some 1000 motherboard just don't work well no matter what is done. > Also, How will Fransis react with lucas at 25mhz? I want to > keep it at 25mhz since I really saw a difference between 20 and > 25mhz, and everything that work at 16mhz work at 25. You will have a problem running the Francis board this fast unless you get REAL fast ram chips. Even then it might not work. The problem is that above 16 Mhz or so, you have to insert extra wait states (wasted clock cycles) because the ram cannot keep up with the CPU. I don't yet have my own Francis, but according to Brad Fowles, the designer, the system is faster running at 16 Mhz with 1 wait state ram than at 20 with 2 wait states. You MAY get it to work at 20 with one wait state as I think some people have had limited success at this. <<<Doug>>> douglee@becker
timg@ziebmef.uucp (Tim Grantham) (07/22/89)
In article <643@becker.UUCP> douglee@becker.UUCP (Doug Lee) writes: >You will have a problem running the Francis board this fast unless you get >REAL fast ram chips. Even then it might not work. The problem is that above >16 Mhz or so, you have to insert extra wait states (wasted clock cycles) >because the ram cannot keep up with the CPU. I don't yet have my own Francis, >but according to Brad Fowles, the designer, the system is faster running at >16 Mhz with 1 wait state ram than at 20 with 2 wait states. You MAY get it to >work at 20 with one wait state as I think some people have had limited success >at this. > ><<<Doug>>> >douglee@becker I have been running my Frances/Lucas board at 20 MHz with one wait state for a while now. The crucial factors in getting this to work for me was not the U9 chip -- they were one, trying different clocks until I found one that worked, and two, always letting the computer warm up for at least half an hour before attempting to do anything with it. Actually, I simply don't turn it off now. The U9 chip I used is a TI SN7474, the clock is a MTRON make. Mind you, I don't have a bus terminator and I don't have anything on the expansion bus. Tim. P.S. I'm using Siemens 70 nsec. memory chips. -- timg@ziebmef.UUCP {uunet!mnetor!lsuc,utgpu}!ncrcan!ziebmef!timg ------------------------------------------------------------------------------- How many Amiga users does it take to change a light bulb?