[rescue] Sick Dreamcast hacker

[Dave McGuire]

On June 18, Joshua D Boyd wrote:
> I don't get it either.  Nice Crystal Semiconductor AD and DA chips are
> pretty cheap.  If someone had the skill and wanted to take the time, I
> bet they could build a 4 chanel card and sell it in small quantities
> for under $100 (but then, I don't know what additional expense being
> PCI incurs over being an ISA card or USB device) and they would make a
> profit.  Better yet, make it four chanel in, and 6 chanel out.  Then
> you could also sell it to the PC for home theatre crowd.  

  CS even makes a single chip 6-channel DAC.  Indeed, the bus
interfacing would be most of the pain of designing such a board, as
very highly-integrated chips (like those from CS) are readily
available.

> >   Well right now it's a plain old 50% duty cycle square wave, and I need
> > to reduce the "on" time from 50% to (not by) 12-15%.  I'm using a
> > 74HC123 retriggerable one-shot now (running about 28%) but I seem to
> > be at the chip's pulse width limit. :-(
> 
> Ahh, so you are trying to reduce the width of peaks of the wave
> without changing the frequency.  I get it.

  As it turns out, I'm no where near that chip's limit...after an
Express Clue Delivery(tm) from Scott Newell (thanks Scott!) I grabbed
the SGS-Thompson 74HC123 datasheet, which reads a bit differently from
the TI datasheet.  The 74HC123 uses a resistor (Rx) and a capacitor
(Cx) for its timing.  As it turns out, my Cx was *way* too high.  Time
to spend a little quality time with my HP41.

            -Dave

-- 
Dave McGuire                       "*BURRRP!*  Mmm...tastes like cannoli,
St. Petersburg, FL                    smells like garlic!" -Sridhar