From: galt@dsd.es.com (Greg Alt - Perp)
Subject: Re: Power Glove
Date: Tue, 18 Jun 91 15:38:59 GMT
Organization: Evans & Sutherland Computer Corp., Salt Lake City, UT


In article <1991Jun17.010236.4172@milton.u.washington.edu>, yonder@netcom.com (Christopher Russell) writes:
> Well, I finally got my PowerGlove hooked up to my computer (Atari ST).
> My plans are to make a controller in the future.  Currently I have 
> just used the BYTE article and have it hooked up in standard "Low Res"
> mode.  The glove seems to be very sensitive to the frequency of the 
> "RESET" and "CLOCK" (BYTE article terminology).  It seems to work 
> very intermittently (or not at all) as I add or delete instructions 
> from my program inbetween the reading of data.  I'm not sure exactly
> how it should "solid" the glove should work, because I do not have a
> Nintendo.  When I am using it there always seems to be at least some 
> "noise", i.e. the LED's and the data read from the glove is "mostly"
> correct.  Uhhg, anyways, I could use any tips or ideas (schematics).
> Thanks to everyone who has already supplied me with info.  Looks like
> it is time to track down an O'scope....

I had the same problem, and was thinking it was my cheap power supply...
I will try to fine tune the code and see if I can get it to work better.
When used with nintendo, the data is very accurate, (the LED's don't blink,
and you can play games without much difficulty).  
 
I accidentally came across extension cords, and they also just happened to
have Glove Ball for almost half price, so I bought both.  The game is very
interesting, kind of a 3D breakout...  I was amazed at how accurate and 
jitter-free the positioning of the glove was.  After playing for a while,
I noticed a few things...
 
1) My knuckles are rubbed raw!  Something needs to be done to make the
   glove less painful.  I will try a few ideas and report back...
 
2) I have more insight about how the hi-res mode works.
   The nintendo CPU is capable of discovering if you have a power glove,
   so I would bet the CPU sends some code to the glove to which the glove
   responds in a way that the normal control wouldn't. (any other ideas 
   how it can tell if you have a glove or not?)
   Also, here is how the glove goes into hi-res mode:
 
Set-up the nintendo system plug in the glove.  Insert the cartridge.
Turn the nintendo on.  The glove beeps, and at about the same time a title
screen is put up.  There is about a 5 second pause, then the glove beeps
again.  At this moment, the game acts as if it is ready (music starts 
playing, etc...).
So, it seems that some time around the first beep, the CPU decides you 
have a glove.  It then loads the data to the glove, and the glove then
beeps again.  The glove then sends out the hi-res data packets.
 

I am thinking about trying to get my PC to put the glove into hi-res mode.
To do this, I need to do 2 things:  read data from the nintendo, send data
to the glove.
 
The connector only has five wires that are used (+5V,ground,data,clock,reset)
It would seem safe to assume data goes both ways on the data line.  What I
would like to know is how could I connect the nintendo to my PC to read the
data off the line?  Somehow I need to have the PC catch the clock and reset
and use them to synchronize the data the same way the glove would.  Then,
to send data to the glove, how could I do this using a the parallel port
interface described in BYTE?  Could I just connect the glove's data line to
both in and out on the parallel port?  This seems dangerous...  would I use
diodes to prevent current from flowing in the out and out the in?
     Greg