Intermittent operation

[snhirsch]

Ok, so this is frustrating. My newly-built Quapple worked the first time I put it in a 5150 and I was able to boot into the filer and DOS. In the midst of testing, it started having problems writing to diskette. From that point on, it refuses to boot. When I hit return in Quadlink it tries a few times to read the FILER floppy before telling me it cannot read the diskette. I reseated all the chips and swapped CPU. No joy. If I'm interpreting the writeup correctly, the seeks indicate the 6502 is executing code - is that correct? If so, can you suggest a direction for diagnosis? I have a fair amount of bench experience and own a scope and logic analyzer.

I recompiled the PLD sources for 8 of the 10 PLDs so I could target fused PALs (have a bunch of them). But, the disk controller (U40) and clock generator (U7) are GAL16V8-20 per parts list.

[schlae]

Since you had it working for a short time that makes me suspect either the drive itself or the Filer disk you're using, or perhaps the floppy data cable. You could try writing another Filer disk and going from there. If you have a logic analyzer, you could also connect a chip clip onto the disk decoding PAL and make sure you see signals that make sense.

But yes, if it tries to read from the disk then the 6502 is running code. That gets bootstrapped through the keyboard port (!) from the PC software side.

[snhirsch]

Ok, good to know that something is running on the Quapple. As a corrollary to my original question, do you have a guess as to why CUPL 4 won't compile the U40.PLD source? I've attached the .LST and .DOC output files. The latter complains of too many product terms.
U40.zip
This certainly isn't a showstopper, since I was able to program a GAL16V8 with your original JED, but my curiousity is getting the better of me.

I've already swapped the cabling and floppy drive and have re-written the media several times from a real Apple 2. But it's possible the diskette is bad, so I'll revisit it.

[snhirsch]

I'm starting to suspect an issue with the host machine. When I hit to load FILER it accesses the A: drive briefly, then moves on to B:. Need to dig out another floppy controller. Working with Olde Hardware is always somewhat of an adventure. But I did have another brief window of proper operation - so there's that.

[snhirsch]

Well, not sure what was going on. After numerous substitute cables and drives, I removed and shuffled boards in my 5150 to place the Quapple closest to the power supply. Everything then started working and I cannot make it fail. So, I'll call this a success. Thanks for your amazing reverse engineering work!