-Yosemite Sam,_Roger Rabbit_) (02/08/89)
After tinkering for a while with my coco III,I belive I've found the root cause of a lot of the timing problems in it. I tried most of the "sparklie"fixes to no avail.The scs gating fix might work for peripherals,but it treats the symptom and not the disease. Aparrently,RS wanted to save space and money by eliminating a NOR gate chip in the address decoder circuit.The address decoder(74ls138) controls cts,scs,and some of the memory decoding.On the original coco,a NOR gate was used to gate e clock and the c select line together.This was used for one enable(the other was tied active all the time).On the coco III, both enables are tied active all the time.This means the select lines aren't allowed to settle before the output line is selected! I added a NOR gate to my coco III to duplicate the original coco's gating for the 74ls138.My performance peripherals dual mode controller now functions perfectly with my unmodified coco-xt and I have yet to see a single "sparklie".I'm not sure if this modification will work without the "A" GIME,but it seems to have done the trick for me.If anyone is interested,email me a message and I'll mail you more specific directions.If anyone uses this mod,I'd be interested to hear the results. vern
knudsen@ihlpl.ATT.COM (Knudsen) (02/09/89)
In article <8902080504.AA07308@decwrl.dec.com>, burke_vern@dneast.dec.com (Mah biscuits 're burnin'!-Yosemite Sam,_Roger Rabbit_) writes: > I added a NOR gate to my coco III to duplicate the original coco's > gating for the 74ls138.My performance peripherals dual mode Thanks for posting this. I'd been wondering about this timing nit ever since I got the Tech Manual for my Coco 3. I've been afraid to tinker with it, thinking that the extra gating might mess something else up. You have proven it won't. Question: Do you really need a separate NOR gate? The '138 has three enables -- one positive, two negative -- that are internally ANDed. One of these is used for the pulse itself. Now if one of the negative enables is left over, shouldn't you be able to lift its hard-wired ground and feed the enabling clock into it? I'll have to check the schematic tonite. Meanwhile, could you post more details? Thanks, mike k -- Mike Knudsen Bell Labs(AT&T) att!ihlpl!knudsen "Five hundred twelve K bytes of RAM ... out of control ... "
dnelson@umbio.MIAMI.EDU (Dru Nelson) (02/12/89)
> > > vern I tried to mail you but it bounced. I am a little out of it so please bear with me. Does this fix the problem with both the Burke and Burke and the dual mode controller? Also, whats a sparklie? How did you find this problem (oscilloscope, schematic, logic analyzer, noticed missing chip?) If this fixes the Burke and Burke or makes the coco III fully compatible with the older ones, what is the fix? -- Dru Nelson UUCP: ....!uunet!gould!umbio!dnelson Miami, Florida MCI: dnelson Internet: dnelson%umbio@umigw.miami.edu