Re: SEGA to Amplifone converter

Clay Cowgill wrote:
>> I found David Shuman's info on building a converter to run
>> SEGA vector games on Atari XY monitors, including a caveat
>> that it can't seem to keep up when there are a lot of vectors
>> to draw.
>>
>
> That more or less mirrors my experience. It was fine for testing and
> tinkering, but I probably wouldn't want to use it as a final solution.
>
>
>> David postulates that it's a bandwidth problem with the amplifiers.
>> Could the cause be the slew rate of the WG6100? Since the
>> Amplifone is faster, perhaps the converter would work 100%
>> when using an Amplifone?
>> Anyone have any experience running this setup?
>>
>
> I really don't remember what I did when I used my Wg6100 on the
> SegaMultigame prototyping. I probably had something very similar to David's
> circuit if it wasn't the actual one he posted. (The timing is quite
> suspicious since that's about when I was working on the Sega Multigame, so I
> probably saw his post and tried it.)
>
> As I mentioned before in another thread, probably 98% of things looked fine,
> but that 2% resulted in visible tearing. I mostly noticed it when there was
> a large delta between points (not necessarily just when there was "lots" to
> draw). I'd wager a guess that that was the fallout from GO8 being able to
> reposition the beam faster than the WG6100 could. (In the Atari AVG's you
> generally don't ever jump the beam more than half a screen width at a time,
> so the SegaXY vector generator doing that probably exceeds the WG6100's
> design spec.)
>
> At the time I remember thinking that if you could essentially add a hardware
> wait state in the G-80 system vector generator when there's a long beam jump
> that it'd probably work 100%, but at the expense of a slight framerate hit
> possibly. If I ever explored that further I don't remember what became of
> it now. ;-)
>
> Zonn will probably know off the top of his head, but I don't think it would
> be any sort of bandwidth problem. I suspect it's just slew rate limited
> with the WG6100's lower deflection voltages. It would be interesting to
> find out if the WG6400 could keep up though.
>

Off the top of my head... Clay is correct. :-)

The issue is vector drawing speed of the WG6101. The speed a vector can
be drawn is dependent upon the voltage available and the inductance of
the yoke. The bandwidth of the drive electronics, on both monitors, is
way beyond being an issue.

The WG6101 has a higher inductance yoke than the G08, and uses a +/-25V
supply, it ends up being one of the slower vector monitors.

The G08 has a fairly low inductance yoke and a high voltage supply of
+/-50V (and also runs very hot, and originally shipped with underrated
resistors that could catch fire). The G08 is pushing some design limits,
especially the safe operation area of it transistors and the ability to
keep them cool, hence the "wind tunnel heatsink". The G08's inductor is
small enough that even running it at a lower voltage, it would still be
fast enough to run any vector game built. Someone posted at one time
(John Robertson maybe?) that they always set the voltage jumpers on the
main transformer to the 130V settings (or whatever the highest setting
is without jumping up to 200V range), that's really a good idea since
since it's operating so close to it's limits, a 10% reduction in voltage
would be really helpful.

The thing about the Sega games is that the vector generator is not that
fast (about the same as Asteriods), however it jumps very quickly
between vectors, and that's the problem. Even though the WG6101 could
easily handle the vector drawing speed, the Sega games don't wait long
enough, when jumping between vectors, for the WG6101 to make it to the
start of the next vector.

The frame rate for the Sega games is 40Hz, and the games don't use most
of this time, they quickly draw the frame and then wait a the next 40Hz
tick. If a wait state could be added between vector jumps it should work
just fine and not have any affect on the game's frame rate, since there
is a lot of idle time at end of a frame draw. This is why the ZVG can
run Sega games at a full 40Hz on a WG6101, it just adds wait states
between vector jumps, with plenty of time left at the end of the current
frame and the start of the next.

-Zonn
---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Wed Jul 15 21:46:20 2009