douglee@becker.UUCP (Doug Lee) (07/01/89)
Here's a little something I discovered recently, while debugging a Lucas 68020/881 board. You can now run it WITHOUT the 68881 installed. Just connect a jumper between DSACK1 and CS lines on the 68881. DSACK1 is pin K4 and CS is pin J3 on the 881. You can also jumper pins 7 and 15 of PAL U7 as an alternative. THis works 100% O.K. with the system and setcpu finding the 68020 and not the 68881. Of course you can'd run code with inline 'F' instructions, but you CAN have the 68020 without buying the 881. I will be building a Francis board soon, so will find out if this works O.K. with it too. I don't see why noth though. <<<Doug>>> douglee@becker
chymes@ics.uci.edu (Charles Hymes) (07/03/89)
In article <616@becker.UUCP> douglee@becker.UUCP (Doug Lee) writes: >Here's a little something I discovered recently, while debugging a Lucas >68020/881 board. You can now run it WITHOUT the 68881 installed. Just connect >a jumper between DSACK1 and CS lines on the 68881. DSACK1 is pin K4 and CS is >pin J3 on the 881. You can also jumper pins 7 and 15 of PAL U7 as an >alternative. THis works 100% O.K. with the system and setcpu finding the 68020 >and not the 68881. Of course you can'd run code with inline 'F' instructions, >but you CAN have the 68020 without buying the 881. I will be building a Francis board soon, so will find out if this works O.K. with it too. I don't see why noth though. <<<Doug>>> > >douglee@becker Have all the problems with geting LUCAS to work with Microbotics Starboard been fixed? I would like to build/buy one, but I have a Starboard with 2 Meg and a SCSI module installed. If they have been, who should be contacted for the latest plans & parts ? Also, is this a resonable progect for a novice? I have never built anything with lsi chips on it, though I am fairly good with a soldering iron. Assuming I have nothing else but a soldering iron, and only a rudimentry knowledge of electronics, what would I have to read+learn+buy to put a LUCAS board together and get it to run with my Amiga+Starboard ? Thanks in advance, Charles Hymes
daveh@cbmvax.UUCP (Dave Haynie) (07/06/89)
in article <616@becker.UUCP>, douglee@becker.UUCP (Doug Lee) says: > Keywords: lucas > Here's a little something I discovered recently, while debugging a Lucas > 68020/881 board. You can now run it WITHOUT the 68881 installed. Just connect > a jumper between DSACK1 and CS lines on the 68881. DSACK1 is pin K4 and CS is > pin J3 on the 881. You can also jumper pins 7 and 15 of PAL U7 as an > alternative. THis works 100% O.K. with the system and setcpu finding the 68020 > and not the 68881. Of course you can'd run code with inline 'F' instructions, > but you CAN have the 68020 without buying the 881. I will be building a Francis board soon, so will find out if this works O.K. with it too. I don't see why noth though. <<<Doug>>> You really want the 68881 CS line jumpered to BERR*, not either DSACK. Though that will only work if LUCAS makes sure that it's BERR* line is buffered from the motherboard BERR*, something like this: LUCAS MOM /| L_BERR* _______/ |_____________ M_BERR* \ | \| 7407 The reason for this is that the 68020 doesn't normally generate an F-Line exception for a non-existant FPU, but it will generate such an exception if BERR* is asserted in response to the FPU space address. If you just assert DSACK* as above, it'll look to the system as if an FPU is installed but messing up bad. So in the BERR* case, the FPU instruction will generate a proper F-line exception, while in the DSACK* case you'll probably get some kind of coprocessor protocol violation exception. SetCPU returns the right answer in both cases anyway, since it uses the OS's opinion of whether there's an FPU or not. The OS tests for the FPU simply by trapping any exceptions and trying a simple FPU operation. If the test succeeds, the OS sets the 68881 bit, if it fails for any reason, it clears the bit. So while most things will work OK in both cases (since F-line emulators aren't popular right now on the Amiga), the BERR* method is the technically correct method. > douglee@becker -- Dave Haynie Commodore-Amiga (Systems Engineering) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: D-DAVE H BIX: hazy Be careful what you wish for -- you just might get it