Received: from watserv1.uwaterloo.ca (watserv1.waterloo.edu) by karazm.math.UH.EDU with SMTP id AA28677 (5.65c/IDA-1.4.4 for ); Fri, 18 Oct 1991 13:15:15 -0500 Received: by watserv1.uwaterloo.ca id ; Fri, 18 Oct 91 14:10:33 -0400 Date: Fri, 18 Oct 91 14:10:33 -0400 From: Dave Stampe-Psy+Eng Message-Id: <9110181810.AA24742@watserv1.uwaterloo.ca> To: gbnewby@alexia.lis.uiuc.edu, jdb9608@cs.rit.edu Subject: Re: sega glasses available? Cc: glove-list@karazm.math.uh.edu > From glove-list-request@karazm.math.UH.EDU Fri Oct 18 13:25:24 1991 > From: Greg Newby > To: jdb9608@cs.rit.edu > Subject: Re: sega glasses available? > Cc: glove-list@karazm.math.uh.edu > > For the circuit, has anyone tested it out? It appears that > you hook in the Sega glasses, and then send a signal to switch > the LCDs (from left-on, right-off to left-off, right-on). > > The signal is sent via pin 4, which is not the regular pin > for reading or writing data. > > So, the question is, what's involved in switching the signal? > If it's a call to termio() or somesuch, what sort of speed > limitations are we talking about? I didn't check the > RS232 standard just now, but I wonder about the +/- 10V switching > signal -- isn't that rather 0-12V? (effectively, probably > +/-3 to 8-12V). > > Finally, does anyone have the tech specs on the glasses (or > an estimate, as I suspect Sega doesn't include such things > with the glasses)? Switching speed, %dark, etc... > > With the brain trust we have on this list, I'm not too worried > that someone will figure out how to hook in the glasses. I'm > just wondering in advance if performance (especially switching > speed) will be up to snuff. > (For example, to give some numbers, we need a minimum > of about 10 frames per second to perceive continuity > between frames of a moving scene. That means the glasses > need to switch 20 times per second, right? A left-right > sequence for each frame. > If there's more than a few u-second lag for the > glasses to switch, we'll have to compensate in the > display program. Being that we're usually using all > the computing power we have just to keep the graphic > images coming fast enough, incorporating some sort > of delay for the glasses would be undesirable.) > > Oh, BTW: I've done a bit of experimenting (back at Syracuse) with > some Tektronics shutter glasses -- a very similar beast, tho > presumably higher performance. I found that the flicker was > perceivable at 30Hz (60 switches / second), and intolerable > (read, "doesn't work") at about 7-15 Hz. I didn't experiment > with the full range....this was because the software-driven > switching circuit never operated more quickly than about 15Hz. > > Ok, I'm rambling now: the alternative, to get 30Hz, is to > NOT switch the glasses via software. Just let them operate > at the frequency of the power supply (60Hz in the US, which, in > the circuit we used @ Syracuse, was run through a rectifier). > Then, no coordinating between your graphics program and the > glasses are necessary, PROVIDED that your program is able to > operate at "top-speed," switching the image on the screen > as often as possible (as in, the speed of the power supply). > I don't know if this technique works for all platforms, > but it worked on the Amiga and Iris we used at Syracuse. I even > synched the glasses off of a Sony VCR/editor to look at > the display on the Iris. > > 'Nuff said. > -- Greg > gbnewby@alexia.lis.uiuc.edu > Here's the EXACT driving specs on the Sega, Haitex, and X-Spec glasses. These all use the same type of LCD pi-cell. They are normally fairly dark (30% transmissive) and drop to 1% transmissive (very dark) when sent a 24V. 2 KHz signal (square wave). Do not accept drivers that use lower frequencies or voltages, as these will not become fully opaque! I spent several month designing drivers for these things about a year ago. We used both computer and video (alternating-field) signals to drive them. I have the following designs: - A 5V supply design with a flyback power supply - A 12V supply design using push-pull drive - A (theoretical) design using the RS232 port of a computer and 4 components. This last design has not been built or tested yet, as I temporarily do not have access to glasses. However, if you want to try it, go ahead. Just tell me it it works. Basically, you hook the outside (common) of the glasses connector (use a Walkman-type stero phone jack) to the ground pin of the RS232 port. Connect 2 of 2.2K resistors to the TX pin. Connect a 1N914,1N4005, etc. diode's cathode to the DTR and another to the RTS pin. Connect one anode of diode and one of the resistors to each of the pins on the glasses jack. THAT'S IT. The software is a bit complex. Basically, you have to set up an interrupt to keep the TX buffer of the serial port stuffed with $55's. Set the baud rate to 4800 baud, and voila! a 24V 2200 Hz output. Raise and lower the DTR and RTS pins (via the UART in the PC) to turn the pulses to each eye on and off. This should work on any PC that supplies at least +/-10V (the RS232 standard is +/- 12V). Interrupt loading is not terribly high: about 5-10%. No guarantees that this will work, but it would be pretty easy to try. -Dave Stampe