|
|
The N64 interface project (still alive) |
 |
|
|
|
|
The N64 interface project (still alive) |
|
Ready made interfaces (and
more) now available !!!
click here for
more info
| Introduction | ||||||||||||||||||||||||||||||||
| Stuff about the N64 controller | ||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||
| Contributions & further links | ||||||||||||||||||||||||||||||||
| Disclaimer |
News:
---------12.04.1999: Started selling ready made N64 interfaces.
The N64 pad is a great controller for all kind of games. It is very handy, has plentiful pads and buttons and in addition it provides an analog joystick (in my opinion it's one of the best).
In november 1998 I searched through the www for informations about the N64 controller, because I liked to use it with my PC.
After some time of searching, I encountered that it'll be not easy to interface the controller to the PC because it uses a fast and bidirectional data exchange.
At the end of november I got my own N64 pad (birthday) and started the research.To establish a data communication with the controller, I started with the design of a hardware "request sequence" generator using standard cmos integrated circuits. But soon I had a little cmos graveyard which still doesn't work properly. So I dropped this "project" in it's priority.
In february this year I had the idea to abuse an eprom as a part of the poll sequence generator, with it I was able to build my first working N64 controller interface (which still works reliable with the SNES or arcade driver within SNESKey). But it's still a complicated design (you can see it in the "history" section).
To shrink the circuit together I tried to find an approach using PICs (little microcontrollers from microchip).
In the meantime Earle F.P.III (the guy who makes this tremendous: dpadpro driver) mentioned that there is someone in Great Britain who tries to solve everything with software. So I came into contact with Simon Nield. He had problems processing the incomming data through the printerport, and I had problems with the amount of hardware I used for the request sequence.
So we put our ideas together and developed the now available interface.
We use the output of the printerport to generate the commands for the N64 controller and a small circuit to store the incomming data (so it could be read out at a slower speed through the printerport).
So you can find at this page the circuit of the interface and informations about how to build it, two test programs ("C" and assembler) and the links to Benji Yorks (SNESKey) and Earle F.P.III's (dpadpro) site.
Benji and Earle now have the info about the circuit and the driver developement is in full progress, most likely they release the drivers within the next days.
Stuff about the N64 controller
The data exchange between the controller and the host:
The N64 controller has only 3 cables to connect to the N64 console. A red one for the power supply (Vcc +3,6V), a black one for ground and a white one for the data exchange.
The data exchange through the "data wire" is bidirectional and could be driven by an open collector device (as the N64 controller itself does).
In its idle state the data line is high ( 3,6 V).
Each bit that is transmitted takes 4 µs of time. When a bit is send to, or from the controller, it starts with pulling data to low (0 V). If the bit has the value "0", the data line remains 3 µs low and then rises to high for 1 µs. If the bit that should be send has the value "1", the data line remains only 1 µs low and rises high for 3 µs.
The N64 controller needs to be polled through the host.
So a "request sequence" must be send to the controller.For the simple data polling (buttons and joystick), sending a 9 bit value : "000000011" to the controller is necessary.
If someone has more informations about the commands that could be send to the controller, please inform Simon or me about it !!!
Immediately**) after the request sequence, the N64 controller starts to transmit its data.
It starts with the value for the "A" button, proceeds with the "B" button and transmits at last information the y-axis value for the analog stick:**) unfortunately this "immediately" is not constant and causes a jitter to the incomming data.
bit data for:
1 A 2 B 3 Z 4 START 5 UP 6 DOWN 7 LEFT 8 RIGHT 9 unused 10 unuesd 11 L 12 R 13 C-UP 14 C-DOWN 15 C-LEFT 16 C-RIGHT 17-24 X-AXIS 25-32 Y-AXIS 33 stopbit For each button a "1" represents a pressed button, and a "0" means: button released. Now for the values of the analog stick:
The data seems to be appear in a signed format. For the x-axis for example it looks like the follwing:
position: bit 17 bit 18 bit 19 bit 20 bit 21 bit 22 bit 23 bit 24 left edge 0 1 1 1 1 1 1 1 ...... 0 x x x x x x x 0 0 0 0 0 0 1 1 0 0 0 0 0 0 1 0 left 0 0 0 0 0 0 0 1 middle 0 0 0 0 0 0 0 0 right 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 0 0 ...... 1 x x x x x x x right edge 1 0 0 0 0 0 0 0 It seems that I can't reach the highest and the lowest values for the x- and y- axis with my N64 controller.
The same format is valid for the y-axis (obviously), where the corresponding data is transmitted from bit 23 to 32.
The stopbit is always "1".
If you count together the times for the polling sequence (9*4 µs) and the returning data (33*4 µs) you get 164 µs at all for the bidirectional data transmission sequence. If you now immediately send the next request sequence to the controller you will receive.... nothing. The controller needs approximately 200 further µs to be ready for the next poll cycle.
How it works:
All the data lines of the printerport supply the N64 controller with power through diodes. The two cmos 4006 are supplied directly through D7.
To send data to the N64 controller, D0 needs to be pulled low as required (see the chapter above for the timing of the interface) for the data transmission.The two cmos 4006 form a 33 bit shift register.
The incomming data reaches the shift register through two different tracks. The first path leads through the 4.7 k resistor, where the falling edge of the incomming signal is used to clock the shift register. At it's second path, the signal passes the R/C (10k /100pF) link, which integrates the signal. So in the moment of the falling edge (clock), there is a "high" voltage at the input of the shift register if the bit send was "1" and a "low" voltage if the corresponding bit has been a "0".
At this picture you can see the voltages at the data input of the shift register (upper beam) and the clock input (lower beam):
10 µs / div
2 V / div
If the data is stored into the shift register, it could be dumped to the PC through the output of the second 4006.
For this, the PC needs to pull D1 to low 33 times (the PC generates the 33 clock pulses).By the way, my N64 controller runs with this circuit at the weak printerport of my spare PC with only 2,6 V.
Connect 2 N64 controller interfaces to a single printerport:
Now you may ask: Can I attach two N64 controllers to my computer ?
And the answer is: YES ! You can.You can do this in two ways:
1. If your computer has two printerports, just attach the second controller with it's interface to the other printerport and install another instance of the dpadpro driver (at joystick ID#2).2. If you have only one printerport available, you need the following Y-adaptor and a new release of Earles dpadpro 4.9 (build 4.9.0.276 or higher).
But remember: At the moment, UltraHLP supports only 2 players with 1 joystick and 1 keyboard.
But today I just send two interfaces to the denuclearized crew, and I'm sure that they'll provide full dual-player mode with two joysticks in one of the next releases.
As you can see from the schematic, you will need an additional power source because a single printerport is not able to supply two N64 controllers with power.
To "pick up" the required +5V from your computer you have several possibilities. A good idea is to use pin 1 at the gameport, or pin 4 at one of the ps/2 ports. If you have an older computer with DIN5 connector for the keyboard, you can use pin 5 of this connector.
If you like, you could even use a 4.5V battery as power supply or a stabilized external power supply.... or you can get the +5V from inside of your PC (from an unused power connector inside)..... or.... or.....
Connecting more than 2 controllers/interfaces to a single printerport.
With dpadpro 5.0 you can use up to 4 n64 controller interfaces at a single printerport.
The schematics for the required "double-Y-adapter" is similar to the scheme above, but obviously you need two more female DB25 connectors (and some wire).Here's how to build the 4x adapter:
signal name pin at printerport pin at female db25 #1 pin at female db25 #2 pin at female db25 #3 pin at female db25 #4 shift-clock 3 3 3 3 3 data-out pad#1 2 2 - - - data-out pad#2 4 - 2 - - data-out pad#3 5 - - 2 - data-out pad#4 6 - - - 2 data-in pad#1 10 10 - - - data-in pad#2 12 - 10 - - data-in pad#3 13 - - 10 - data-in pad#4 15 - - - 10 GND (ground) 18 to 25 18 to 25 18 to 25 18 to 25 18 to 25 power (+ 4,3V) *) 9 9 9 9 *) You can get the required 4.3V through a single diode from your gameport, keyboard-connector or ps2/plug as described above.
Hey, we found a cool page where someone describes in detail how to build the Interface. Unfortunately the site is in spanish, but it has plenty of excellent pictures which could be understood without the knowledge of this language.
I will put the pictures here at my website, with english text in the next days. In the meantime take a look at dev's website at emulatronia.
Unfortunately the N64 controller uses no standard connector. Nintendo uses a proprietary 3 pin plug for it
.
Now there are different ways to "connect" the controller to the circuit.An elegant way is to hack a N64 extension cable, but these parts are quite expensive (12 - 15 $ at my site).
The simplest way (not recommended !) would be just to cut off the connector and solder the wires directly to the interface, but then you can't use the controller any more for your N64 console.
So I used another way: I just inserted a pair of DIN 5 connectors (male and female) at the end of the N64 cable, so I can still use it for an N64 if I plug both connectors together, or I can plug it into the interface circuit where I assembled a DIN 5 pin female connector for pcb assembly (like the standard keyboard connector at motherboards). In addition this solution has the advantage that you can use a cheap keyboard extension cable with the N64 controller.
Here you can see the female DIN 5 pin connector (for print assembly) at the side of the interface:
At the end of the N64 controller cable there is a filter assembled (the cable just forms a little coil through a ferrite core) to minimize radio interferences. If you want to insert a male/female pair of connectors, it's a good idea to move the filter a few inches towards the controller, so that there'll be enough space to mount the plugs.
It's a little tricky to open the snap-in mechanism, but two little screwdrivers and a pocket knife will do the job.
Here you can see the standard circuit assembled on an experimental board :
component side
soldering side
At the left you can see the DIN 5 pin connector which leads to the N64 controller. In the middle there are the two cmos circuits, at right you can see the diodes and the cables leading to the SUB-D 25 pin connector (male).
If I'll find the time I make a small pcb. If it's ready, you'll find it here.
Here you can see now Simon's design:
If you have a closer look at it, you'll see that Simon only uses 4 diodes for the "power supply" of the N64 controller. Furthermore he has not fitted the transistor. You can do this if your printerport is sensitive enough (just connect pin 10 of the printerport to pin 9 of IC-2.
You can see only a single cmos 4006 at the picture, your eyes don't fool you. Simon mounted both circuits like a little "sandwich" (piggy-back design) while he spread away all the pins (except pin 14, 7 and 3 which needs to be connected together).
Design studies for a "ready made interface"
Many of you have been asking me if I could provide some ready made interfaces.
So I'm now working on it.Here you can see the first possible solution of such an interface:
As you can see, the circuit fit's into the hood of an "normal" DB25 connector. It uses a very small double sided pcb (which is going to be optimized a little bit).
The cmos circuits are assembled "smd - like". You can't see the second one, because it is assembled at the bottom of the board.
As connection to the N64 controller I used a standard mini DIN connector (like the one for ps/2 mices and keyboads). But it's also possible to attach a hacked N64 extension cable to it.
Here you can see the upper copper layer.
And now a "look through" to the bottom layer.The design still needs a little bit of optimization. Especially the pcb could look a little bit more professional. I made the pcb yesterday night in my kitchen
. I had no possibility to make real junctions at the board, so I needed to solder the parts at both sides (that's why I lost approximately 1mm in height for the diodes..... and now, after closing the hood, they look a little bit "crunched").
So if you would like that I go into that direction, please mail me ! Tell me what you think, if you like it or not.
And tell me how much it's worth for you (be honest please) if I'm going to sell it once.
I'm especially interested if you would prefer to pay 5$ more if I use an attached N64 cable instead of the DIN ones
(I already received an email from a norwegian guy, who would like to give me 50$ for such an ready made interface, but I expect that it'll be much cheaper).I don't know when it'll be ready, but my actual pcb source needs at least 14 days.
If you have further informations about where I can get those N64 extension cables for a good price, or if you have someone who could manufacture double sided pcbs for cheap (maybe yourself), please mail me too.
| Standard circuit: | |
| 2 * cmos 4006 | IC1, IC2 |
| 10 * standard diode (like 1N4148 or similar) | D1 - D10 |
| 1 * resistor 4.7 k, 1/4 watt or less (yellow, purple, red) | R1 |
| 1 * resistor 100 k, 1/4 watt or less (brown, black, yellow) | R2 |
| 1 * resistor 10 k, 1/4 watt or less (brown, black, orange) | R3 |
| 1 * resistor 2.2 k, 1/4 watt or less (red, red, red) | R4 |
| 1 * capacitor
100 pF (instead of 220 pF as
previously recommended), ceramic |
C1 |
| 1 * transistor standard pnp-type, like BC559 or 2N3906 | T1 |
| 1 * experimental board (single pads) | |
| 1 * DIN 5 pin connector female for print assembly | |
| 1 * connector DB25 male | |
| For the N64 controller cable: | |
| 1 * DIN 5 pin connector male | |
| 1 * DIN 5 pin connector female | |
| or: | |
| 1 * DIN 5 pin connector male | |
| 1 * N64 extension cable |
At the moment there are two test programs available:
N64pad.zip from Simon Nield.
Press "1" and "2" to change the printerport. Press +/- to change the timing delay.
To change the command being sent use the '[' and ']' keys (change-inventory item).
Commands known so far;
00 = get status
01 = read button values
02 = read from mempack
03 = write to mempack
ff = reset controller
04 = read eeprom
05 = write eeprom
N64_0278.com, N64_0378.com and N64_03BC.com for the different printerport adresses (click here for the assembler source). It's not as cool as the one from Simon, but it's short and reliable.
Best would be if you run it in a fullscreen !!! DOS box (so you could close it through windows).
Just launch the program in a fullscreen DOS box. Each incomming bit is indicated with a "1" if set and with a "0" if not.
You can see the stopbit (33th bit) which is always set. Furthermore you can detect the C-UP and C-LEFT buttons pressed at the N64 controller.
To close the program, you need to close the DOS box manually (the program itself runs in an endless loop).As already mentioned in the introduction, Earle F.P. III and Benji York are currently updating their programs/drivers for the interface.
Finally I was able to figure out how the rumble-pak is going to be accessed.
I checked the data line while the N64 generates an "rumbling" event with my tektronix scope. The sequence to switch the rumble motor on and off was easy to find (because it was a repetitive event which could be displayed by the scope).
Then I made a little test program that reproduces this sequences. The controller (with the rumble-pak attached) responded at the data line exactly as he should (and as he did at the original N64), but that %&$% rumble-pak refused to rumble.
So there must have been a short "initialisation" sequence, which enables the rumble-pak functionality at the controller.
To "catch" this short sequence, a friend helped me out with an HP Logic analyzer to observe the data communication between the N64 and the controller, but the memory depth of the Logic analyzer was to small to catch everything of interest. After hours of successless measuring, we encountered that we would never be able to catch the sequence with the available modules and memory.
So I borroughed a transient recorder (able to store 2 channels with a duration of 65 ms (1µs per sample)), and I was lucky to catch the missing initialisation sequence for the rumble-pak.
Now you can download my small assembler test utility with rumble-pak support.
To send the "initialisation" sequence just press the <START> button (the ? changes to an "i" then).
To switch the motor on for a short time, just press the <A> Button, to switch it off, simply press <B>. That's all.
Remember to run the testprogram in a full-screen DOS box.
Test program for LPT1: rumbl378.com LPT2: rumbl3bc.com LPT3: rumbl278.com
Because daily we have hundreds of emails begging for the direct-x drivers (dpadpro), Earle was so kind to "pre-release" version 4.9 of his incredibly cool program !!!Be thankful now, and do not disturb him with your bugreports, wishes and further ideas (especially don't ask for rumble pack support right now) until the end of march !!!
He is just in preparation for his final exams. And we all wish him the best for it. So cross your fingers now and support him to succeed by beeing patient and giving him the time for his preparations.Oops, nearly forgot the link
ps: Before installing dpadpro4.9 remove previous versions of dpadpro from your computer. And in addition, to ensure that the driver works properly (and the axis and buttons are displayed correct), it's a good idea to restart your computer after installation.
Great !!! There is a new driver available for our interface.Actually it's only a beta-release (0.9) which only supports 1 controller at LPT#1 (I/O 0378h).
It has been written for Windows 98, but works good for Win95b too (at least here at my site).You can download Jemy's DirectN64 pad driver here: Jemy's DirectN64 driver.
Stay tuned at this site !
New (18th of april'00):
There's now beta-release 0.95 available, which provides a generic rumble/force-feedback support.
Due to the numerous emails I get each day since we released the circuit, I just added this s
