PV9Y@cornella.BITNET.UUCP (01/26/87)
Received: by CORNELLA (Mailer X1.23b) id 9608; Mon, 26 Jan 87 09:09:18 EST
Date: 26 January 87 09:09 EST
From: PV9Y@CORNELLA
Subject: mouses
To: INFO-ATARI16@SCORE.STANFORD.EDU
Alright, I have a minor problem that I'm really getting tired of. My
mouse squeaks. I know that sounds sort of dumb, but it really does
whenever I move it to the right. Is there some way I can lubricate it
or generally muzzle it? I would just like to point out that it's all
Michtron's fault for programming a decent game (Major Motion) and then
using the mouse as the controller. The ST has a perfectly good joystick
port, why couldn't they have used it? If anyone out there is ever so
slightly connected with Michtron, could you please ask them to write a
patch for Major Motion so that it works with the joystick?
And one other thing. I haven't heard anything about this anywhere,
but doesn't it sound reasonable that you could replace your mouse
entirely with a trakball (Wico makes a good one that (I assume) will
connect to the ST, it connected to the Atari 2600, which had the same
joystick ports)? I don't know that much about the method of control in
the trakball, but it seems that it should be basically the same. Anyone
care to comment on this? Also, someone just mentioned getting a new
mouse system for the ST. Does anyone know of other mice that would work
with the ST?
Thanks.
Adam Engst
pv9y@cornella
engst%tcgould.tn.cornell.edu@crnlcs.bitnet (<- send mail here, please)
The opinions expressed above are representative only of my own and
those of ten thousand penguins with whom I am constant communication.
And none of us are connected in any way with Wico or Michtron.atwell@utah-cs.UUCP (01/28/87)
Keywords: I tried plugging in a standard Wico trackball to the mouse port and found that although the mouse would move, you couldn't control it at all reliably (it seemed like the mouse was "spinning its gears"). Bart
pes@bath63.UUCP (01/30/87)
You probably need to clean your mouse. I just did mine, and it makes an incredible difference. Opening it is trivial -- see the manual. In the absence of a manual, turn it over and look. The ring around the ball slides in the direction of the arrows on it. Gently, press the ring inwards and slide. Take out the rubber ball. Look at the inside, and you'll see 3 little silver rollers. Don't see them? I saw 3 black rollers when I did mine. Gently, with say a wooden match or a toothpick or a Q-tip, and maybe the tiniest bit of alchol or disk cleaner, make them into 3 shiny silver rollers. Then reassemble. It's magic. On the trackball subject, I managed (with about $2 worth of parts) to modify an 8-bit Atari trakball to look like a mouse. (I could have done it with no added bits, but I wanted to leave it the capabilities of looking like an 8-bit joystick and an 8-bit graphics tablet -- God know why, cause I don't have an 8-bit, but nevermind.) For some applications it is much easier to use than the mouse. For others it is worse. The problem is that 'holding a button and moving' using a trakball requires 2 hands, which isn't always practical. Back to the mouse, if you *do* feel the need to lube it, I'd recommend something like powdered teflon (can get it at Sears if you're in the States, sorry, I didn't note the return address).
robert@arizona.UUCP (02/03/87)
In article <747@bath63.ux63.bath.ac.uk>, pes@bath63.ux63.bath.ac.uk (Paul Smee) writes: > > On the trackball subject, I managed (with about $2 worth of parts) to modify > an 8-bit Atari trakball to look like a mouse. (I could have done it with > no added bits, but I wanted to leave it the capabilities of looking like Could you please tell us how?
pes@bath63.UUCP (02/13/87)
In article <1486@megaron.arizona.edu> robert@arizona.edu (Robert J. Drabek) writes: >In article <747@bath63.ux63.bath.ac.uk>, pes@bath63.ux63.bath.ac.uk (Paul Smee) writes: >> >> On the trackball subject, I managed (with about $2 worth of parts) to modify >> an 8-bit Atari trakball to look like a mouse. (I could have done it with >> no added bits, but I wanted to leave it the capabilities of looking like > >Could you please tell us how? Ok, here's the trakball info. The description of 'how it works when you buy it' is not rigourously correct, because I've mislaid my construction notes and feel I've been looking for them long enough. In particular, there's one chip unaccounted for. (As I recall, the chip I can't account for is only a necessary layer of buffering or inversion, not relevant to the description here.) Still, it will give the idea. Because I don't fancy trying to draw circuit diagrams on the terminal, most of this will be in prose. I will describe only one half of the ciruitry in the 'history' bit, as the X circuit and the Y circuit are identical. The entire circuit is a 'cliche' which I recognize from my long-ago technician days, so I don't feel I'm revealing any proprietary secrets. A) How it works now, first stage -- common to all modes. The motion sensor is a dual set of 'photodiode/photosensor pairs', with a single square-toothed wheel spinning in between. The spacing of the active elements and of the teeth is such that the two pulse trains generated are out of step. The pulse trains look like: xxxxx xxxxx xxxxx xxxxx (A) x x x x x x x x x xxxxx xxxxx xxxxx xxx xxxxx xxxxx xxxxx xxxxx (B) x x x x x x x x xxx xxxxx xxxxx xxxxx x If the ball is spinning left, the (A) train leads in time. If right, the (B) train leads. The frequency of the pulse train depends on how fast the ball is moving. Each pulse train is fed thru a single-input Schmitt trigger to square up the pulses and to provide the degree of fan-out needed to drive the rest of the logic. (I surmise that the occasional mouse creep you see is due to the toothed wheel stopping so that it partially blocks one of the photodiode/sensor pairs, right at the switching threshold. Then pure electronic randomness, or tiny mechanical motions -- such as by activating the disk drive sitting elsewhere on the table -- would cause a spurious pulse train in one channel.) B) How it works now, second stage -- common to both 'old' modes. The pulse trains are fed into a D-type edge-triggered latch -- one is fed in as the data input, the other as the clock input. If the clock pulse is ahead of the data pulse, the latch will latch in the 'off' mode; if the clock pulse is behind, the latch will latch in the 'on' mode. Thus the output of the D-latch gives a stable signal (until you switch directions) which indicates whether the ball is moving to the left or to the right. The D-latch also provides a 'complemented output' in reverse logical sense to the primary output. The output to the connector pins is selected (depending on how you've set the joystick/tablet switch) using a CMOS CD4019BE (allegedly the 4519 is compatible -- I mention it because you may want to buy one). This is most easily regarded as a 4-pole 2-throw switch, under remote control of the physical switch mentioned earlier. In 'joystick' mode, the primary and complemented outputs of the D-latch are switched onto the left-right pins of the connector (pins 3 and 4). In 'tablet' mode, one of these pins is made to carry the primary output from the D-latch (so giving direction) and the other is made to carry one of the pulse trains from stage 1 (giving speed). C) How to change it -- the bit you've been waiting for. PLEASE NOTE -- I am describing what I did. I am not guaranteeing that you can do it. The operations required are not complex, but are fiddly. Don't blame me if you mess it up. ALSO NOTE (and I will say it again from time to time) all of the instructions which follow assume that you have not removed the circuit card from the base -- so you are always meant to be looking at the BOTTOM (non-component side) of the board, with the BACK (side the connecting cord exits from) of the base nearest you. What you need -- A new plug and cable, because the one that comes with is one wire short (you need 8 lines). I recommend (but didn't use) getting a 'Joystick Extension Lead' from Tandy (UK) or Radio Shack (US) and cutting off one (the correct) end, so you'll have a (smaller) molded-on plug. If you try to use a standard style loose connector, you'll probably find you have to butcher it a bit to make it fit physically into the deeply recessed socket on the ST (why did you do that, Atari?) A low-leakage low-power soldering pencil -- it's all CMOS circuits. 3 resistors, 4.7K low wattage (say .1 to .25 W). (Only 1 needed if you don't want to retain tablet/joystick capability for your 8-bit :-) If you are happy to make your trakball 'mouse-only', that's it. If you want to retain the existing capabilities as well, you also need: 1 CD4019BE IC (the afore-mentioned '4-pole switch'). 1 smallest-you-can-find miniature 2-pole 2-throw (DPDT) switch. Some odd-bits of wire. Multi colours is nice so you can keep track of what you've done. The wires from inside the old cord will do fine. I recommend also a 16-pin socket for the chip, and a small piece (~5x8 cm) of veroboard or general-purpose 'experimenter's breadboard', just to keep things neat. Before you buy that: Open your trakball. How? Turn it upside down. Work with it sitting on something, because when you unscrew the cover, the cover and ball will want to fall out. There are 4 screws, one hidden under each of the rubber feet, so you have to peel the feet off. Separate the lot, turn the base back over. My circuit board says A2010 Rev A. If yours doesn't, this may not work. Pay careful attention to how you removed anything you remove (not much) so you'll be able to put it back. It is possible to un-mount the sensor mini-boards and so look at the face of the main circuit, but don't bother. It's fiddly and not necessary (probably). Note where the cord comes into the box. There are 7 clips attaching the wires to the PC board. Unclip these and (somehow) make the cord come out of the box. This is the most fiddly bit. Pay careful attention, as the switch mounts in the same plate and may well come adrift -- if you're watching, it will be easier to puzzle it out when you put it back. Butchering the wires is allowed -- butchering the box isn't. You may find you have to slide the plate containing the switch and cord hole up (it's detained by 2 little nubs on the inside of the case. If you do, the switch will definitely pop adrift. The 7 large rectangular pads to which the wires were clipped will appear again later. Holding the board with these tabs towards you, I will call them E-1, E-2, ... E-7 from left to right. E-1 is connected to the wide trace that runs all round the edge of the board (ground). (On my board they are labelled, left-right, as GRA BLUE VIOLET BRN RED ORN YELLOW but it feels safer to invent my own number system in case this is not generally true.) Set the base down with the back (ex-cord side) towards you. (On the circuit board, the 2 square corners, with the switch pads, will be towards you, and 2 cut-off corners away.) In the lower-right corner there is a switch pad (looks like a set of meshed fingers). In the trakball the two switches are equivalent. We want a *2* switch 'mouse'. While there's room to work, look at this bit -- in the following, I've done PC traces (relevant ones only) with - and |, solder pads (some irrelevant ones also for context) with o: (Approx centre o o of board here) | o o o | | o----o (link) o----------------------- | | ------------| ------------- | | ------------| ------------- | | ------------| ------------- | At the point marked '(link)' there is a wire link on the other side of the board. This is what connects the switches together, and it must be cut. Remember the solder pad on the right hand side of this link (towards the switch matrix) because you'll have to make a connection to it later. There are 4 IC's on the board. Since we're looking at the back of the board, you'll see the solder pads -- which are sets of 2 parallel rows of 7 (14-pin ICs) or 8 (16-pin) little wodges of solder. The interesting set is the set in the upper-left hand corner (this is the hex Schmitt trigger, a 40105 in mine). Remembering that we are looking at it upside-down, the pads (and IC pins) are numbered: xxxxxxxxxxxxEDGExOFxBOARDxxxxxxxx x (Bunch of irrelevant PC tracks) CUT OFF x CORNER x 8 9 10 11 12 13 14 x x 7 6 5 4 3 2 1 The buffered clock pulses that we want are on pins 10, 12, 2, and 4 (which I'll refer to now as St-10, St-12, St-2, and St-4 to save space). Now I've introduced all the parts of the board with which we're concerned. I sugggest you make a sketch, using these notes and referring to your own board sitting in front of you, as it will make referring to things easier (and also ensure you've found the right places). Or, stick little sticky labels onto the board. Winkle the new cord into the box, put back the switch and plate which have come loose. You'll need a fair bit of loose wires (at least 8 inches) from the cord inside the box, so be generous. Sorry I can't give details on how to do this, but it's definitely the least fun part. Figure out some neat way tostrain-relieve it, so you won't pull the connections loose if you yank the wire. Strip the outer cover from the at least 8" of wire and separate the separate wires. IF YOU ARE HAPPY WITH A MOUSE-ONLY TRAKBALL -- The easy way out. Make the following connections, as listed, to the pins of the new plug. If you're lucky, the inner wires you're using will be properly colour-coded. If not you'll have to trace them with a multimeter or something. Plug pin number Connects to 1 St-4 2 St-2 3 St-12 4 St-10 5 (No connection) 6 the pad connected to the 'fingers' switch 7 E-3 (+5v DC) 8 E-1 (Ground) 9 E-2 And, connect your 4.7K resistor (pull-up) between the 'pad connected to the fingers switch', and E-3. Before closing the thing up, try it. (Remember we've been looking at it wrong-way-up so far, so get on the other side of it.) Lay the ball in place, plug it in, move the ball, make sure the cursor goes the right way. Pins 1 and 2 (St-4/2) are up/down; 3 and 4 (St-10/12) are left/right. If it goes backwards (shouldn't do) reverse the appropriate pairs. If it's happy, screw the cover back on and you're off. (But DO MAKE SURE to route the wires neatly somewhere where they will not interfere with the ball, or drift over either of the 'finger switch' arrays.) IF YOU WANT A FULL-FEATURE TRAKBALL (mouse + joystick + tablet) You masochist. Ok, let me introduce you to the CD4019BE. It is a 16-pin chip. If you are doing this, I hope you know how the pins are numbered, but just in case -- orient the chip so that it is 'vertical', little legs downwards, the short edge with the U-shaped notch in it away from you. The pins are then numbered from 1 through 8 (top to bottom) on the left side, and from 9 thru 16 (BOTTOM TO TOP) on the right side. I will call them CD-1 through CD-16 in my wiring list. Be neat about the wiring. Use enough wire to avoid interfering with the ball and switch-pads. The new chip can be allowed to rest in the 'wing' of the case to the left of the circuit board (near the Schmitt Trigger). I leave the layout to you. The final member of the cast of characters is the DPDT switch. Look at it from behind, so that the 6 terminals are facing you, in 2 horizontal rows of 3 terminals. I will refer to these as: SW-1A SW-1B SW-1C SW-2A SW-2B SW-2C Figure out where you are going to put the switch (somewhere on the side that the Schmitt Trigger chip (St) is on, I'd recommend, since that's where the action is) butcher in the necessary mounting hole(s). It should be fairly out of the way so as to leave room for the chip et al. Don't mount it until it's been wired, though. Again, I strongly recommend you turn this info into a sketch. Now the wiring list. I have done this 'connection at a time', so in some cases you will find yourself connecting something to a place where you've already hung something else. Be very careful not to leave solder bridges. Particular care is needed when connecting to the St-x pads, because they are small and close together. Also, take care not to use any more heat than required in making the connections to St (and CD, if you haven't bought a socket). Connect to CD-1 St-10 CD-2 E-6 CD-3 St-12 CD-4 E-5 CD-5 St-2 CD-6 E-4 CD-7 St-4 CD-8 E-1 (ground) CD-9 SW-1C CD-10 Plug pin 1 CD-11 Plug pin 2 CD-12 Plug pin 3 CD-13 Plug pin 4 CD-14 SW-1A CD-15 E-7 CD-16 E-3 (+5vDC) Hook a 4.7K resistor between CD-9 and CD-16 (or any other handy source of the +5v). Hook a 4.7K resistor between CD-14 and CD-16 (ditto). SW-1B E-1 (or any other handy ground point) SW-2A E-2 SW-2B The pad on the 'finger switch' mentioned above, where you cut the wire link. SW-2C Plug pin 6 Hook the last 4.7K resistor between SW-2C and a convenient source of +5v, e.g. E-3. E-2 Plug pin 9 E-1 Plug pin 8 (ground) E-3 Plug pin 7 (+5vDC) That is exactly what I did to mine. Push the wires neatly out of the way, test, and put it back together again. In one position of the new switch, the trakball now looks like a mouse, and the other switch has no effect. In the other position of the new switch, the old switch selects between joystick and tablet operation. Good luck -- Paul .
pes@bath63.UUCP (02/16/87)
The unaccounted-for chip came to me in a dream. It's not relevant to the conversion of the Atari trakball to look like a mouse, but (working from memory) just to make the 'how it works when you buy it' bit complete -- The chip I couldn't remember what was is a twin mono-stable (one-shot). Its function in life is to 'kill' the output of the joystick 'D latch' if the ball stops moving in the appropriate direction -- so logically 'centering' the simulated joystick. As I recall it is triggered from one of the buffered pulse trains.