RE: MC1495's

I'm not concerned about the parts essentially going away-- I can do a
lifetime buy and support the product more or less indefinately. I
figure there's a world-wide market of at most about 200 of the things.
(I'll probably only sell about ~30 or so initially based on the response
I've seen so far.)

Even if I only get 100 parts they'll last a long time... ;-)

As for an alternative-- Harris makes some analog 4-quadrant multipliers.
They have one of those weird-ass Harris part numbers that I can never
remember though. Problem is, they're priced like the Motorola parts
used to be-- $15-20 a pop.

Consensus seemed to be that people would want a display corrector if it
was under $30. Better if under $25. I can get there with the cheap
Motorola parts, expensive Motorola parts (or Harris) and I'd have to
sell it for over $50 to break-even. I don't think many people would
bite... The design has about 50 parts on it, so it take a pretty good
sized chunk of PCB, and since I won't be making many the PCB is
relatively expensive-- thus I need cheap parts to get to the
price-point.

As for what it does...

What the multipliers are doing is basically squaring the y-input and
multiplying against a constant to scale the result. That value is then
multiplied against the x-input (which can be positive or negative) which
results in a sort of inverse linear function-- values close to "center"
on the Y axis make the X-axis stretch out, values closer to the limits
of the Y axis result in the X axis having very little expansion.

Since the "pincushion" effect on the CRT is sorta parabolic and the
correction factor is roughly linear the correction isn't perfect.
However, the majority of the distortion is in the center (larger
correction) so it ends up working rather nicely. I added a
potentiometer to "adjust" the amount of correction-- just turn on the
cross-hatch grid and turn the knob to where the side look to be more
straight than bowed.

I started playing around with the Cinematronics method of
display-correction in Electronics Workbench, but haven't got much tested
with it. (It might be cheaper if it lends itself to use with the Atari
method of vector generation...)

-Clay

> ----------
> From: Mitchell Rohde[SMTP:bovine@eecs.umich.edu]
> Reply To: vectorlist@spies.com
> Sent: Wednesday, May 13, 1998 8:13 PM
> To: 'vectorlist@spies.com'
> Cc: Clay Cowgill; Mitchell Rohde
> Subject: RE: MC1495's
>
>
> If you are making these newly designed display correctors, why use
> a part that is lookin' like it's on it's way out?
>
> What does the part do? Can we find a better choice?
>
> mitch
>
>
>
Received on Thu May 14 09:07:31 1998