NES Power Glove Repair Possible?

[HelloNESMAN]

Hello All,

I just bought this NES power glove with a missing DB9 cord. Does anyone know the pin out from the cable to the board so I can attempt to atatch a new one? My soldering skills are just fine so it won't be an issue, I just cant find any information as to what connects where.

Also, one other slight problem, I don't have a Male DB9 cord, just a female one from a Genesis. I also have a male DB9 port off of a dead Atari. In theory I should be able to de-solder the connector from the small Power Glove box, and put the other one in without issues correct?

Last thing, before I even start any of this. My sensor bar seems to be working but I'm really not sure. If no game is inserted, while in port 1, lights stay on. I put in SMB3 and the light flash on then off. Is this normal?

Thanks to all!

[Bavi_H]

Does anyone know the pin out from the cable to the board so I can attempt to atatch a new one?

I do not have a Power Glove nor an NES, and my hardware electronics knowledge is basic and hands-on experience is virtually none. But I have researched how the Power Glove's DE9 cable connects to its main board by combining information from the following sources:

• The Power Glove source code. It has comments about what the various signals on the Power Glove CPU are used for.
• A data sheet for the Power Glove CPU. It helps match up the CPU signal names with the CPU pin numbers.
• Descriptions of a 4021 shift register. This helps explain what the pins on the shift regsiter are for.
• Various pictures of the Power Glove circuit board that I found online. I was able to follow the traces between the pads on the circuit board where the wires for the DE9 cable connect back to various pins on the CPU or the shift register.
• A webpage that described some of the signals that were observed on a Power Glove DE9 pinout.

(See the links below.)

The part that I'm still unsure of is correct match between the DE9 pins and circuit board pads for the three microphone signals.

Here is what I know so far:

Code: Select all

DE9 pin  board pad  description        signal direction
-------  ---------  -----------------  -------------------------------------------------------
     1   ?          microphone signal  (sensor bar to junction box then) junction box to glove
     2   3 and 9    ground
     3   ?          microphone signal  (sensor bar to junction box then) junction box to glove
     4   ?          microphone signal  (sensor bar to junction box then) junction box to glove
     5   2          +5 volts
     6   3 and 9    ground
     7   7          latch              (NES to junction box then) junction box to glove
     8   6          clock              (NES to junction box then) junction box to glove
     9   1          data               glove to junction box (then junction box to NES)

The three ? above should match up to one of these, but I don't know exactly how they match up:

board pad  description                   signal direction
---------  ----------------------------  -------------------------------------------------------
       4   middle microphone signal      (sensor bar to junction box then) junction box to glove
       5   horizontal microphone signal  (sensor bar to junction box then) junction box to glove
       8   vertical microphone signal    (sensor bar to junction box then) junction box to glove

DE9 pins:

de9-pins.png

board pads:

board-pads.png

Links

• Power Glove Source Code (alt)
• Notes about 4021 shift register from the NesDev wiki.
• A web page that describes some of the signals that were observed on a Power Glove DE9 pinout: Hacking the Nintendo Powerglove.
• power-glove-pins-notes.zip (14,488,290 bytes). A zip file containing an ods spreadsheet and images showing how I followed the traces from the DE9 connection pads to the chip pins. Also includes the CPU datasheet PDF. I originally made and posted this zip file in the following Discord discussion.
• Discussions on NesDev Discord server about the Power Glove DE9 pinout. See the messages in the #research channel from 2024-07-06 to 2024-07-07 starting with this: So I am doing research on PAX Power glove. (To join the NesDev Discord server, you can go to the main nesdev.org page and click the Discord link in the top menu.)

[Bavi_H]

I don't have a Male DB9 cord, just a female one from a Genesis. I also have a male DB9 port off of a dead Atari. In theory I should be able to de-solder the connector from the small Power Glove box, and put the other one in without issues correct?

I imagine it would work. But if you or anyone else needs to diagnose or repair it in the future, the DE9 genders and pin numbers being reversed from any Power Glove pinout descriptions you find online will probably cause confusion.

My sensor bar seems to be working but I'm really not sure. If no game is inserted, while in port 1, lights stay on. I put in SMB3 and the light flash on then off. Is this normal?

I suspect this is normal. (But be aware I have little hardware experience, so this is all thoretical understanding.)

The lights in the sensor bar are basically snooping on the data signal from the glove back to the NES. When the glove is predenting to be a NES controller the lights help show what controller buttons are being sent to the NES.

When there is no game inserted (or if a game is inserted, but it isn't actively asking for data from the controller port), I imagine the 8-bit buffer in the LED controller board is not getting updated, and its initial default state might cause all the lights to be on (?).

When there is no glove connected to the junction box, and a game is actively asking for data from the controller port, I imagine the data signal that is supposed to be coming from the glove might but that is actually not currently connected might be misinterpreted by the LED controller board as random data (?) or always "button not pressed" (?).

Some hardware information

An NES controller communicates with the NES using a serial signal. The controller sends the 8 data bits for the buttons to the NES one after the other on a data line. A latch line from the NES to the controller tells the controller to start or restart the data sequence. A clock line from the NES to the controller tells the controller when to switch to the next data bit in the sequence. The main chip in the controller that is responsible for sending the serial data is a parallel to serial shift register.

The main board of the Power Glove also has a parallel to serial shift register like this. The Power Glove CPU connects 8 of its data lines to the shift register to control the simulated controller buttons that the NES will see (or in some cases, to send other data bytes to the NES).

In the Power Glove sensor bar, there is a sub board near the LEDs that snoops on the data signal going from the glove to the NES and uses a serial to parallel shift register to get the data back out from the data line to control the six LEDs.

[HelloNESMAN]

Hi bavi, thank you that picture of the board is exactly what I needed. I saw some videos on youtube of people opening the glove, but the board is covered in wires without showing what goes where, so that is a MAJOR help!

[Bavi_H]

Hi bavi, thank you that picture of the board is exactly what I needed. I saw some videos on youtube of people opening the glove, but the board is covered in wires without showing what goes where, so that is a MAJOR help!

Just to make sure you noticed:

The picture in my post of the board pads does not show what goes where, it just shows the same numbers that are on the board, since some of them are hard to see. The numbers on the board do not correspond to the DE9 pin numbers. See the text table above the pictures for the correspondence between the DE9 pin numbers and the board pad numbers that I know of. There are three connections that I don't know the correct correspondence for.

[HelloNESMAN]

Hi bavi, thank you that picture of the board is exactly what I needed. I saw some videos on youtube of people opening the glove, but the board is covered in wires without showing what goes where, so that is a MAJOR help!

Just to make sure you noticed:

The picture in my post of the board pads does not show what goes where, it just shows the same numbers that are on the board, since some of them are hard to see. The numbers on the board do not correspond to the DE9 pin numbers. See the text table above the pictures for the correspondence between the DE9 pin numbers and the board pad numbers that I know of. There are three connections that I don't know the correct correspondence for.

Hello Bavi,

Yes i see what your saying now. I'm guessing the only way to know what those last 3 pins correspond to is to have someone use a multimeter in continuity mode to see what goes where? I've tried researching myself but haven't found the answer to that. Is there any website/active forum that might be worth checking on?

[Bavi_H]

Is there any website/active forum that might be worth checking on?

Sorry, I don't know of any.

I'm guessing the only way to know what those last 3 pins correspond to is to have someone use a multimeter in continuity mode to see what goes where?

If we find someone with an intact glove willing to try that, that sounds like the easiest way. But maybe there are other ways to find out. Here are some possible ideas. (My electronics understanding might be lacking, so maybe some of the hardware ideas won't work as well as I thought.)


Idea 1. Probe unknown pins while glove in use?

Maybe you could somehow wire up all the other pins, but leave some way to test the unknown three pins. (I don't understand the hardware well enough to know if you'd need to have some sort of "in line" connection to snoop on the signal?) Then maybe you could wrap heavy towels around two of the sensor bar's microphone sensors, leaving only one microphone sensor exposed at a time. You might be able to power up the glove with any game and press the START button on the glove just to make sure it is sending the ultrasonic pings out to the microphone sensors. Then with only one microphone sensor exposed, probe the response pins until you can tell which pin corresponds to the exposed microphone, then repeat for the other two microphone sensors (?).


Idea 2. Try all combinations?

There's 6 possible permutations for those three unknown microphone signal pins. If you have the Super Glove Ball game, maybe you might be willing to test the different possible connections for those microphone signals until you find the one that causes the physical glove motions to correctly control the on-screen glove motions (left-right, up-down, in-out). Or if you don't have the Super Glove Ball game, the default start up mode of the glove (Program 1) will light up the directional lights on the sensor bar to indicate the relative up-down-left-right directions from the center point.


Idea 3. Connect Arduino or similar to DE9 board pads?

In the Discord conversations I linked to, a user had connected the DE9 pins of their glove to an Arduino. They wrote a program for the Arduino that effectively re-created the polling signals the NES sends to the glove and captured the return signals the glove sends back to the NES. (They were exploring the possibility of replacing the junction box and sensor bar with some kind of wireless solution, and were investigating the signals available on the glove's DE9 pins.)

With that kind of set up, I thought of a possible test: (direct link to Discord messages)

I have an idea to enter a custom program on the glove that will just return information about which speaker signals are getting to which microphones. Then if you could write a test program to set one of the three unknown pins, then read data bytes from the clock/latch/data pins, we may be able to tell which pin corresponds with which microphone.

To enter the custom program, press the following buttons on the glove:

Prog A 0 Enter
0 0 Left
1 0 8
0 0 0 2
0 0 0 1
Enter Enter

If I have planned it correctly, this custom program is supposed to put the glove into "glove" mode, and tells the glove to just return one byte of data, the information byte about which speaker signals got to which microphones.

According to my research, when one of the unknown pins is set, then the response from the glove will be one of the following cycles:

Code: Select all

  binary    hex  decimal  meaning
  01011111   5F     95    response data is ready
  00001001   09      9    upper-left mic received both speakers

  01011111   5F     95    response data is ready
  00010010   12     18    upper-right mic received both speakers

  01011111   5F     95    response data is ready
  00100100   24     36    lower-right mic received both speakers

However, the other user never responded to that suggestion, so I don't know if they tried it or if it even works like I thought.

[HelloNESMAN]

Is there any website/active forum that might be worth checking on?

Sorry, I don't know of any.

I'm guessing the only way to know what those last 3 pins correspond to is to have someone use a multimeter in continuity mode to see what goes where?

If we find someone with an intact glove willing to try that, that sounds like the easiest way. But maybe there are other ways to find out. Here are some possible ideas. (My electronics understanding might be lacking, so maybe some of the hardware ideas won't work as well as I thought.)


Idea 1. Probe unknown pins while glove in use?

Maybe you could somehow wire up all the other pins, but leave some way to test the unknown three pins. (I don't understand the hardware well enough to know if you'd need to have some sort of "in line" connection to snoop on the signal?) Then maybe you could wrap heavy towels around two of the sensor bar's microphone sensors, leaving only one microphone sensor exposed at a time. You might be able to power up the glove with any game and press the START button on the glove just to make sure it is sending the ultrasonic pings out to the microphone sensors. Then with only one microphone sensor exposed, probe the response pins until you can tell which pin corresponds to the exposed microphone, then repeat for the other two microphone sensors (?).


Idea 2. Try all combinations?

There's 6 possible permutations for those three unknown microphone signal pins. If you have the Super Glove Ball game, maybe you might be willing to test the different possible connections for those microphone signals until you find the one that causes the physical glove motions to correctly control the on-screen glove motions (left-right, up-down, in-out). Or if you don't have the Super Glove Ball game, the default start up mode of the glove (Program 1) will light up the directional lights on the sensor bar to indicate the relative up-down-left-right directions from the center point.


Idea 3. Connect Arduino or similar to DE9 board pads?

In the Discord conversations I linked to, a user had connected the DE9 pins of their glove to an Arduino. They wrote a program for the Arduino that effectively re-created the polling signals the NES sends to the glove and captured the return signals the glove sends back to the NES. (They were exploring the possibility of replacing the junction box and sensor bar with some kind of wireless solution, and were investigating the signals available on the glove's DE9 pins.)

With that kind of set up, I thought of a possible test: (direct link to Discord messages)

I have an idea to enter a custom program on the glove that will just return information about which speaker signals are getting to which microphones. Then if you could write a test program to set one of the three unknown pins, then read data bytes from the clock/latch/data pins, we may be able to tell which pin corresponds with which microphone.

To enter the custom program, press the following buttons on the glove:

Prog A 0 Enter
0 0 Left
1 0 8
0 0 0 2
0 0 0 1
Enter Enter

If I have planned it correctly, this custom program is supposed to put the glove into "glove" mode, and tells the glove to just return one byte of data, the information byte about which speaker signals got to which microphones.

According to my research, when one of the unknown pins is set, then the response from the glove will be one of the following cycles:

Code: Select all

  binary    hex  decimal  meaning
  01011111   5F     95    response data is ready
  00001001   09      9    upper-left mic received both speakers

  01011111   5F     95    response data is ready
  00010010   12     18    upper-right mic received both speakers

  01011111   5F     95    response data is ready
  00100100   24     36    lower-right mic received both speakers

However, the other user never responded to that suggestion, so I don't know if they tried it or if it even works like I thought.

Hi Bavi, thanks I think idea two is the easiest for me, I do own super glove ball so its worth a shot. There shouldn't be any problem/damage if I wire the three microphone pins incorrectly while trying the connections correct?

[Bavi_H]

Idea 2. Try all combinations?

There's 6 possible permutations for those three unknown microphone signal pins. If you have the Super Glove Ball game, maybe you might be willing to test the different possible connections for those microphone signals until you find the one that causes the physical glove motions to correctly control the on-screen glove motions (left-right, up-down, in-out). Or if you don't have the Super Glove Ball game, the default start up mode of the glove (Program 1) will light up the directional lights on the sensor bar to indicate the relative up-down-left-right directions from the center point.

I think idea two is the easiest for me, I do own super glove ball so its worth a shot. There shouldn't be any problem/damage if I wire the three microphone pins incorrectly while trying the connections correct?

I do not have any hands-on electronics experience, but I suspect it's no problem. I think it would cause a similar result as rotating or flipping the sensor bar of a fully intact Power Glove system.