PDA

View Full Version : Why has nobody tried to remake the first videogame?



per
February 22nd, 2010, 01:41 AM
It's generally considered that the "CRT Amusement Device" from 1947 is the very first videogame. However, only a prototype were made, and I really doubt that prototype still exist.

However, the patent is widely availble, and it includes full schematics(!!!). The only problem is that it's not labeled with resistor-values, however voltage levels are present. Based on information present in the document, it should be possible to calculate the values. It should therefore be possible to make a working repilka, and it shouldn't be too hard either.

The device is actually very simple. It displays a dot (represents a missile), moving in a certain path a certain time based uppon some variable resistors. After the certain time, it will quicly defocus to visualize an explosion. The graphics is present as a picture on blank plastic-overlays, and to play the game you have to guess the settings to make the missile to explode rigth on a target on the overlay.

Dwight Elvey
February 22nd, 2010, 05:17 AM
Hi
With that date, it would need to be tubes.
I would guess, it was intended to drive a XY
scope or such. To bad, scopes don't have an
external focus control.
I've not looked at the patent ( number would
be nice ). Is it for a vector or rastor scan
CRT?
I knew an idiot that had the prototype for
Pong and scrapped it. He was a scrap dealer
though. He also scrapped a Systron Donner
analog computer before I could get to it
( it had tubes in it you know ).
Dwight

per
February 22nd, 2010, 05:52 AM
Here it is: http://www.google.com/patents/about?id=n-NZAAAAEBAJ&dq=2455992

It's basically just a sawtooth-wave generator, two voltage regulators that uses the sawtooth-wave to generate voltage levels for the X and Y coordinates of a dot, and a focusing controll which will suddenly turn the dot out of focus on the very end of each sawtooth-wave cycle (for the "explosion" effect).

It turns out that the dot will move in a strigth line if a normal CRT is used (the angle of the line is variable, and that's actually the point of the game). The reason it moves in a curved manner in the actual description, is that the vertical deflection plates have a non-conductive base that is coated with a high-resistive (~30KOhm) material. One of the sides is connected to +400v (problably the rigth side), while the opposit side (problably the left side) is connected to the Y voltage output. This causes a varying vertical deflection field for the different horizontal positions. The horizontal deflectors are regular.

The tubes used are four of the 6V6, and four of the 6Q5. The main voltage used for the sawthooth-wave is +250v, the CRT is sourced by -1100v, and the deflection plates use +400v as source. There is one relay, and a ton of resistors/v-resistors.

Dwight Elvey
February 22nd, 2010, 07:55 AM
Hi
Making a CRT with the resistive plate would be tricky
but just creating a parabolic curve is not that hard.
You just stack 2 integrators.
For the defocus, one could add a chaotic high speed
oscillator. That would give the impression of defocused.
These are all things that I could implement on my
analog computer and a regular XY scope.
I've been busy playing with it lately. I've been mostly
making rose engine plots with it.
Dwight

nige the hippy
February 22nd, 2010, 10:02 AM
it's electrostatic deflection, so it can be done outside the tube. A couple of resistors glued to the outside of the tube should do it!

Chuck(G)
February 22nd, 2010, 10:20 AM
it's electrostatic deflection, so it can be done outside the tube. A couple of resistors glued to the outside of the tube should do it!

:huh: ??????

It should be possible to find all of the parts needed to create ths today. It might even be less work than trying to program a simulation of it.

per
February 22nd, 2010, 10:21 AM
it's electrostatic deflection, so it can be done outside the tube. A couple of resistors glued to the outside of the tube should do it!

That's indeed true!
But most CRT's used in osclioscopes are shielded on the inside, even the really old ones... And it should still be plates since they have a greater surface area.

per
February 22nd, 2010, 12:34 PM
I have noted down the controls for those not too interested in reading through the document:
1 Speed control
4 Voltage control (scale calibration)
5 Enable/Disable sawtooth wave

11 Init. X component of missile motion vector control
12 Init. Y component of missile motion vector control
14 (11/12) Init. angle of missile control
26 X offset of display (X position calibration)
26' Y offset of display (Y position calibration)

42 Primary Fire-power control (time taken before missile explodes)
46 Secondary Fire-power control (time taken before missile explodes)

54 focus control
55 defocus control (durning explosion)
62 Brigthness control

It actually sounds like a rather fun game, as you can aim the missile rather freely. I guess it all depends on how detailed you make the overlay, say, it's way easier to hit a big building that covers most of the display than a small/distant tree covering only a small fraction of the display.

MikeS
February 22nd, 2010, 12:55 PM
it's electrostatic deflection, so it can be done outside the tube. A couple of resistors glued to the outside of the tube should do it!Huh? Tell me more; how does that work?

per
February 22nd, 2010, 01:12 PM
Huh? Tell me more; how does that work?

It's a charge of (kinetic) energy stored in the form of an electromagnetic field between two opposedly charged plates, usually. A big capacitor in other words. The electron-ray will pass between the plates, and it's path will be bendt by the field. How much depends on the charge stored in the field (voltage difference).

Usually, plates are used so it will basically just work as a normal capacitor with constant field on all locations between the plates. However, the problem here is that resistors are used, so the strength of the field varies based on where the ray passes between the plates.

But the main thing is that electromagnetic fields aren't stopped by glass, so the plates can be placed on the outside of the tube. However my point is that many tubes contains an internal metal shield that will pick up the energy stored in fields that may interact from the outside, thus making it impossible to do this trick.

Dwight Elvey
February 22nd, 2010, 01:53 PM
It's a charge of (kinetic) energy stored in the form of an electromagnetic field between two opposedly charged plates, usually. A big capacitor in other words. The electron-ray will pass between the plates, and it's path will be bendt by the field. How much depends on the charge stored in the field (voltage difference).

Usually, plates are used so it will basically just work as a normal capacitor with constant field on all locations between the plates. However, the problem here is that resistors are used, so the strength of the field varies based on where the ray passes between the plates.

But the main thing is that electromagnetic fields aren't stopped by glass, so the plates can be placed on the outside of the tube. However my point is that many tubes contains an internal metal shield that will pick up the energy stored in fields that may interact from the outside, thus making it impossible to do this trick.

Hi
As I stated before, one could make a parabolic path with two integrators. It doesn't have to be
fancy, the ramp generator circuit ( what has been called a saw tooth but actually a one shot ramp )
doubled up would be enough.
Another option would be to use a coil. Just increase the current in the coil based on the
other axis voltage ( not sure if that alone would make it parabolic ).
Dwight

Chuck(G)
February 22nd, 2010, 02:07 PM
It's a charge of (kinetic) energy stored in the form of an electromagnetic field between two opposedly charged plates, usually. A big capacitor in other words. The electron-ray will pass between the plates, and it's path will be bendt by the field. How much depends on the charge stored in the field (voltage difference).

I believe the field is electrostatic, not electromagnetic in the case of electrostatic deflection CRTs. While it might be possible to create a high-intensity electrostatic field and have it penetrate the glass, I don't know how well-focused it could be made, nor how any internal coatings on the CRT or glass envelope makeup might interfere.

per
February 22nd, 2010, 02:19 PM
I believe the field is electrostatic, not electromagnetic in the case of electrostatic deflection CRTs. While it might be possible to create a high-intensity electrostatic field and have it penetrate the glass, I don't know how well-focused it could be made, nor how any internal coatings on the CRT or glass envelope makeup might interfere.

sorry, my bad... I did actually believe electrostatic and electromagnetic was the same thing, I now clearly see it's not. The book I got explaining it is in danish...

Dwight Elvey
February 22nd, 2010, 04:12 PM
I believe the field is electrostatic, not electromagnetic in the case of electrostatic deflection CRTs. While it might be possible to create a high-intensity electrostatic field and have it penetrate the glass, I don't know how well-focused it could be made, nor how any internal coatings on the CRT or glass envelope makeup might interfere.

Hi
Magnetic deflection works fine in a picture tube on a TV. I don't think it would be
any different for a scope CRT. The focus is related to the time of flight of the
electron beam and the focusing electrode. The magnetic field of a deflection coil
doesn't change these so long as it is used to replace a deflection only compared
to an electrostatic field.
It does have to be oriented perpendicular to the electrostatic but that is just a placement
issue. The winging would have to be such that it create a gradient similar to that created
by the resistive divider in the CRT used for the original.
Still, it would be easier to create the effect of the parabolic effect from outside and
stay with a simple electrostatic deflection of a standard scope CRT. It is not a complicated
circuit. It just takes more tubes.
Dwight

Chuck(G)
February 22nd, 2010, 04:20 PM
Well, in a way that was my point, Dwight. If you're going external, you may as well stick with magnetics. But I don't believe it's hard to find (small) electrostatic deflection CRTs even today:

http://www.sphere.bc.ca/test/crts.html

MikeS
February 22nd, 2010, 05:06 PM
I get all that, even Dwight's hybrid electrostatic/electromagnetic idea, but a couple of resistors glued to the outside for deflection? That I'd have to see...

In my younger days I'd take an old b&w TV, disconnect the yoke windings and play with feeding various signals into them. If you hooked them up to a stereo amp you'd get neat Lissajous patterns that could keep you fascinated for hours, especially if it was the right kind of music and you were drinking/smoking the right stuff; made a few for my friends and the best one was an old Zenith with a round picture tube.

Of course this was the sixties and we were easily amused; we all had colour organs too, of course...

Hmm, may have to revisit this with some old colour TVs; enough sitting at the curb these days.

Chuck(G)
February 22nd, 2010, 05:19 PM
In the 60's, you'd see ads in the back of Popular Electronics and similar magazines advertising "How to turn your TV into an oscilloscope". Well, technically yes, but your sweep rate would be fixed and the vertical input would be somewhat frequency-limited.

Yeah, I didn't get the bit about the resistors either. AFAIK, most (carbon or film) resistors create an insignificant external field.

Dwight Elvey
February 22nd, 2010, 05:41 PM
Well, in a way that was my point, Dwight. If you're going external, you may as well stick with magnetics. But I don't believe it's hard to find (small) electrostatic deflection CRTs even today:

http://www.sphere.bc.ca/test/crts.html

Hi
It isn't a problem of getting an electrostatic CRT. It is a problem of the way they create the
parabolic curve of the beam. There is a resistive surface on the deflection plates
that have a gradient across them. This increases the deflection in a nonlinear way
as the beam is deflected more by the other axis.
Generating the parabolic curve is the key to making it seem like a real
life shot.
Puting an external resistor doesn't do anything since electrostatic fields
would be blocked by the shield. I wouldn't even want to try to wind
such a coil. Making the path parabolic with a couple more external intergrators
is the way to go. One more dual triode tube would be enough. We don't
have to be super accurate. You'd need a relay or multipole switch in place
of the ramp start switch to put initial conditions on the caps for the intergrators
but that is relatively easy.
One could use a TV tube and yoke because we are not talking about
really fast signals here. The vertical coil would work well enough but
the horizontal coil might not have enough inductance to run really slow.
Still, one would need to generate the needed parabolic curve in the one
axis and that is where the intergrators come in.
Dwight

Chuck(G)
February 22nd, 2010, 07:00 PM
So simply using a capacitor as an integrator driven from a sawtooth generator won't do it? Follow up with a stage or two of amplification and you should have it close enough for a game...

But let's think 1940's. You want something to be reliable. So how about a cam follower on a disc with a parabolic curve cut into it? The follower is attached to a potentiometer. Vary the speed of the motor and you vary the speed of the ball. If you didn't trust mechanics and potentiometers, you could use a variable capacitor or you could use a glass disc with a film overlay whose transparency varies according to the parabola you desire. Put a lamp on one side and a phototube on the other. Or cut a curve into the periphery of a steel disc and spin it to induce a parabolic waveform in a pickup coil.

Dwight Elvey
February 22nd, 2010, 08:02 PM
Hi Chuck
It would take two intergrators cascaded to make it parabolic, not
just one. Your idea of a pot or optical method has merit.
The optical disk is the best but you need two inputs. One
could be rotation and the other radial distance. One represents
time while the other would be the horizantal distance traveled.
The design already has the amplifier.
Dwight

per
February 23rd, 2010, 04:52 PM
I just buildt a small simulator of the device in Javascript. It's not the best thing I've done, but it's useable and should give an idea how it works.

i have tried to keep as much as the thechnology from the schematics in the simulation, and stuff like the sawtooth wave is being included (it's even displayed!). Anyways, it's playable if you make an overlay.

BTW, you will need to tell your web-browser to allow local files to use javascript in order to run it. I know javascript isn't the best thing to use, but it's quite fast to program unless you plan to make something complex and of high stability.