Problems with Williams 49-ways, not working in all directions.

[quarterback]

I've got a couple williams 49 ways and a solderless (IDE connected) GPwiz49 that I just got.  I've hooked up one of the joys but it's not working in all directions.

Specifically, it won't go "up".  When I test it in Windows' "game options" calibration screen, it will only move around in the bottom half of the square.  When I push the joystick all the way 'up' the calibration cursor will jump from the center of the calibration screen down to the bottom.

I've checked all my connections and they're in the right places and have continuity from one end to the other, but I still can't get it to work.

So I hooked up my other 49-way and it will go up and down, but will only work for the RIGHT half, not allowing any movement onto the left side of the calibration square.  Same kind of deal, when I push it to the LEFT, nothing happens until I get all the way left and then the cursor instantly jumps from the middle all the way over to the RIGHT side..

Does anybody know why this might be happening?  Is this a symptom of something?  Did I get two 'bum' 49 ways?   

Thanks

[quarterback]

Dooood, I'm feeling very blue right now.    

I slept on the whole thing hoping that things would be better this morning.  So I got up and just tried a different compter, different OS and different USB cables with the joys, but with the same results.   The spiders look great and the joys look like they're in good condition.  You know, they're old and used and there's wear and rust on the joystick shafts and whatnot, but I'm starting to worry that there's something wrong with the sensors.

Anybody recognize this kind of 49-way behavior?
Thanks

[KenToad]

Oh, I thought you had gotten a new 49-way and that it might have been defective.  Sorry to hear that your old sticks aren't working.  Wish I could help. 

Cheers,
KenToad

[Flinkly]

well, short of looking at my old 49-ways, it might be the way it's all screwed together.  if you take the thing apart, you can stick it back together in all sorts of ways, but only one is the right way.  i know it seems obvious, but maybe you didn't notice.  after that, i'd check the lines on the pcb, mine are all bubbly and i wouldn't be suprised if there was a scratch accross a few.  lastly, i would consider the optics kerput, but i would b suprised if you have two bum joysticks.

also, i've never tested mine, so this is all based on knowledge of the workings and not experience with these ancient joysticks.

[quarterback]

Oh, I thought you had gotten a new 49-way and that it might have been defective.  Sorry to hear that your old sticks aren't working.  Wish I could help. 

Thanks for the consolation anyway.   I've been giving this some more thought....

Okay, so when I'm in 49-way and I push the joystick slowly to the RIGHT, I see the cursor go from: center, to right-center, to right-right-center, to full-right like this.


|      +      |

|        +    |

|          +  |

|            +|

That tells me that all three optical sensors are working, right?  They're going from: all blocked, to 2 blocked, to 1 blocked to none blocked.

Okay, but when I push to the LEFT, the results are:

|      +      |

|      +      |

|      +      |

|            +|

The push to the left is using the same optical sensors that a push to the right would use.  But, for some reason, when I push to the left, it's not reading (or registering) the fact that the sensors are being unblocked.    It's not the placement of the sensors.  I can see them being "unblocked" as I move the joystick.  So why is it that it's not registering on the calibration software?

The GPWiz49 is the link between the joy and the PC, but I don't know why it wouldn't be working correctly......

[u_rebelscum]

... but I'm starting to worry that there's something wrong with the sensors.

Anybody recognize this kind of 49-way behavior?
Thanks

I'd guess nothing's wrong with the sensors.  It's the "direction" pin that you're having problems with.

The problem one of the following: it (pin "D") is not connected well, is short circuited, or the circuit on the PCB isn't working. 

The seven levels the stick outputs per axis are:

1 111
1 011
1 001
0 000 <- center
0 100
0 110
0 111

Notice how the bolded number, that's the direction pin, the other three are the three sensors.  The circuit in the joystick figures out what to send down the dirction pin: it does not have a sensor.  See how the two extremes differ only by the direction pin.

Frist thing to try is re-connecting the X axis D pin.

[quarterback]

well, short of looking at my old 49-ways, it might be the way it's all screwed together.  if you take the thing apart, you can stick it back together in all sorts of ways, but only one is the right way.  i know it seems obvious, but maybe you didn't notice.  after that, i'd check the lines on the pcb, mine are all bubbly and i wouldn't be suprised if there was a scratch accross a few.  lastly, i would consider the optics kerput, but i would b suprised if you have two bum joysticks.

Yeah.  I had taken one of them apart, but not the other, I've double checked them and they seem to be all hooked up correctly.   

I was afraid the optics (or electronics) might be screwed, but as I played with them some more, I came to the conclusion that it's not the optics because they're working in 3/4 of the directions (up, down and right, but not left.  And left, right and down, but not up)   

Hmmmm.... still pondering this one....

[quarterback]

Frist thing to try is re-connecting the Y axis D pin.

Thanks.  I'll check that out.

Frist thing to try is re-connecting the X axis D pin.

And that one too

[quarterback]

Well, the connection between the  x axis D and the the GPWiz49 is fine.

I know there's a way to hook these up directly (using analog mame and whatnot) without the encoder.  And I've got a multimeter here.  How can I 'test' the x axis d pin but remove the GPWiz49 from the equation?

[u_rebelscum]

Heh, I read your first post about the up and down problems, so put "Y axis" and posted, then read the posts that came after I logged and thought I might have misread the axis, so "corrected" to the X.  Now I reread the whole thing, and see you have the same problem on different axes on different sticks.  That's what I get for taking so long to type a reply.     (Of course, my excuse is work keeps interupting me. )  (Another post while I'm typing.)

Well, the connection between the x axis D and the the GPWiz49 is fine.

I know there's a way to hook these up directly (using analog mame and whatnot) without the encoder. And I've got a multimeter here. How can I 'test' the x axis d pin but remove the GPWiz49 from the equation?

You could either connect all 8 pins to keywiz/ipac etc, or through 8 leds (with the right resistors), and look at whats going on.  I'd go with the leds, personally, but I already have them.  If leds are strong enough, you could go from the stick to led to the gwiz49.  Check what the stick's outputing while centered; make sure both axes output the same while centered (and when gwiz49 plug-in and computer bootup).

I really hope it's not the joystick's PCB where the problems are.  That would take tracing the PCB and making sure there aren't any breaks or shorts.  Good luck!

[quarterback]

I know there's a way to hook these up directly (using analog mame and whatnot) without the encoder. And I've got a multimeter here. How can I 'test' the x axis d pin but remove the GPWiz49 from the equation?

You could either connect all 8 pins to keywiz/ipac etc, or through 8 leds (with the right resistors), and look at whats going on.  I'd go with the leds, personally, but I already have them.   

Yeah, no LEDs here.   I was hoping for a "look for continuity between the ground and the XD when joystick is pushed" or "look for 5v between something and something else" or something along those lines.

Since each of the sensors is a NC 'switch', I know I get continuity when they're all centered and I can test each sensor to see if it 'opens' when I push the joystick.   I'm just not sure what the XD and XY should be 'reading' when I go one direction or the other.

Is there just a way for me to test to see if the "direction pin" is activating or not?
Thanks man
qb

[fl0yd]

At one point I was trying to identify other possible wiring scenarios to make it possible to have the stick mounted with the PCB in other directions.

[quarterback]

At one point I was trying to identify other possible wiring scenarios to make it possible to have the stick mounted with the PCB in other directions.  I never found a working wiring out of the possible permutations, but I did see similar results to what you are reporting.

I know you've mentioned that your wiring is correct, but this is identical to some of the wiring possibilities that I tried.

Thanks floyd.  I'm going to pull a different harness and start from scratch, just to eliminate that possibility.  The thing that befuddles me (in multiple ways) is that niether joy works, but with slightly different problems..... curious.   

I'll start over and try again. But dude, I just wanna play Sinistar!!!!!

[quarterback]

Rewired, new harness, I went back to the GGG directions and started over but the results are exactly the same.   Here are some pics in case somebody sees something that I'm doing totally wrong here.  Human Error is always a possibility here

[RandyT]

QB, you should drop me an email.  There's a possibility that one of the inputs got toasted during the wiring process (please tell me you aren't wiring it with the power on.)

You should be able to test the GP-Wiz49 by grounding combinations of inputs to simulate the joystick behavior.  But I'll have to go through them with you (after I educate myself with your situation).

On the other hand, I have a 49-way here that I was careless with and it does exactly what you have been describing.  I'm guessing a dead opto-sensor on mine.  Hopefully you don't have that problem.

Let me know,

RandyT

*edit*

I just noticed that both of your sticks are Williams types.  If you hook both of them up EXACTLY the same way and get different results, this is pointing pretty hard in the direction of bad components on both of the sticks.

[quarterback]

QB, you should drop me an email.  There's a possibility that one of the inputs got toasted during the wiring process (please tell me you aren't wiring it with the power on.)

All wiring was done with the USB unplugged, so 'no' on that front. 

You should be able to test the GP-Wiz49 by grounding combinations of inputs to simulate the joystick behavior.  But I'll have to go through them with you (after I educate myself with your situation).

Cool.  I'm on my way out (have to go to work today ) but I'll drop you an email and/or feel free to gimme' instructions on testing each input.

I just noticed that both of your sticks are Williams types.  If you hook both of them up EXACTLY the same way and get different results, this is pointing pretty hard in the direction of bad components on both of the sticks.

Yeah, exactly.  I thought that was odd as well.  It makes me wonder what's up.  Although, realistically, I should pull the other one out of my CP again to make sure I didn't add to the confusion by putting it back together 'wrong' as I mounted it.   Just to make sure that it's being hooked up in the exact same way.  I have been using the same harness on both but maybe I flipped something or something when I took it apart and put it back together.... it was 2am... then 3am... then 4am as I was first doing this

It's always possible that I crossed a wire at some point or something, so any tips you can give me on testing the GPwiz49 would be great.   PM me whatever thoughts you might have and I'll drop you a line when I get back to my computer.

Thanks

[quarterback]

At one point I was trying to identify other possible wiring scenarios to make it possible to have the stick mounted with the PCB in other directions.  I never found a working wiring out of the possible permutations, but I did see similar results to what you are reporting.

Hey there fl0yd, I've pulled my other 49way out of my CP (meaning I took it apart and put it back together again) and it seems like there's only one way to put it back together.  I mean, I guess I could spin the top metal plate, but as far as the PCB and the moving parts, they only seem to fit one direction.  Am I missing something?

Thanks

[quarterback]

Looks like I'm screwed

WIth Randy's help, I've multi-meter tested them both and it's clear that the Y-axis "direction" is kaput on one of the joys and the X-Axis "direction" is kaput on the other.   

I swear that the post in B/S/T (where I got these) said that they were tested and working great, but now I can't even find the original "for sale" post.   I'll have to go back through my 153 PMs and find the date and go from there.

Anyway, the end result is I have one that works one way and one that works the other.  If anybody has any ssuggestions for making one working 49-way out of two half-working ones, lemme' know.   

[RandyT]

As Urebel wrote, it sounds like the optical components are ok.

The direction output appears to be generated by a quad, 2 input NAND chip.  The logic works like this:

input A1 HIGH and input A2 HIGH = Output A LOW

Else

Output A = HIGH

So if the problem is that the direction output never transitions, but the sensors seem to be working in the other direction,  then either the chip is toast or a part leading to one of the inputs is.

I haven't taken mine apart fully to trace the path to the inputs, but if it looks like a simple split from the optical components, the chip is probably the problem.

The good news is that the chip isn't really rare nor expensive.  The bad news is that the PCBs on these old sticks look like their  traces are made from poorly adhered aluminum foil.  Replacing the soldered-in chip could be rough to do without causing more problems, and it might not even be the culprit.

FWIW,

RandyT

[quarterback]

How about this.   What if I hook up a leaf switch that allows (or cuts) the 5v to the XD or YD connections on the GPWiz?

If pins 8 & 6 on the GPwiz are basically detecting if the 49way is sending   5v or no-5v, I could attach a leaf switch with the leaf going across the edge of the optical sensors.  When I move the joystick, the sensor-blockers will shift, pushing the leaf switch to make contact.  The appropriate info would be sent to the GPwiz and all would be good in the universe.

What do you think?  Possible?

[RandyT]

How about this.   What if I hook up a leaf switch that allows (or cuts) the 5v to the XD or YD connections on the GPWiz?

If pins 8 & 6 on the GPwiz are basically detecting if the 49way is sending   5v or no-5v, I could attach a leaf switch with the leaf going across the edge of the optical sensors.  When I move the joystick, the sensor-blockers will shift, pushing the leaf switch to make contact.  The appropriate info would be sent to the GPwiz and all would be good in the universe.

What do you think?  Possible?

Interesting idea, but I don't think it will work very well.  First of all, you can't just cut the 5v, it would need to be brought low through a resistor.  The mechanical challenge of mounting and getting everything to actuate reliably and at the right point would also be just shy of impossible.

If the chip is the problem, replacing it would probably be the better option.

RandyT

[quarterback]

Interesting idea, but I don't think it will work very well.  First of all, you can't just cut the 5v, it would need to be brought low through a resistor.

Yeah, I originally thought it was 5v or no-volts, but now I realize the direction 'info' is either 5V or ~233mVolts.  Not as simple as I might have hoped

If the chip is the problem, replacing it would probably be the better option.

Yeah, the problem is that I'm sukekekekekeee at soldering.  And even if I didn't burn up the board while trying to solder in a new chip, it could end up being a problem with the traces somewhere along the way.

Unless anybody here knows of someone who can fix/refurb these, they may be off to ebay.

Thanks again for everybody's help.  Any suggestions are welcome.

[Major Rock Hardy]

Well, I don't know how stuck you are on DIY, but you might find a local electronics repair shop that would swap out the chip for you pretty reasonably.  Their solder skills should be very good (they do it for a living) . 

Most shops would have a minimum charge but I'd bet that they'd knock out both joys as one charge.  And if the parts are as common as Randy suggested then they might even have them on hand.  They may cut you a really good deal since you aren't paying them to troubleshoot it, but just to swap out a couple parts (assuming you've decided the parts are the culprit). 

Or you just might wander in to one of those places and give them the sob story and if they're video game fans (good chance) they might do it for free!  Here's to wishful thinking - never hurts to ask.

[quarterback]

Well, I don't know how stuck you are on DIY, but you might find a local electronics repair shop that would swap out the chip for you pretty reasonably.  Their solder skills should be very good (they do it for a living) . 

It's true, but unfortunately I'm not 100% sure it's the chips themselves.  One of the joys chips is outputting 5v no matter whether I go right or left, but this could be because some other link in the chain is 'broken'.  The chip may be fine, but the path of 'info' "into" or "out of" the chip might be the problem, and there are a number of other components in that path. 

If I was 100% sure that the chip replacement would help, then I'd replace it in a second, but with all the other variables involved, it's just not clear that that's the solution.   If I wasn't afraid of blowing up my GPWiz, I'd probably try a couple other things as well.

And if I can figure out exaclty how to drop the 4.95V to 233mV, I still may try and rig up my leaf-switch 'add on'.  Since there are plastic parts that slide in and out in the 4 directions, I really don't think it'd be that tough to attach a switch that manually told the GPWiz that the joy was headed in the opposite direction.

I just hate to give up on these though.  I'd much rather have a Williams 49-way than a new Happ 49 way but there don't seem to be many of these available out there.

[Flinkly]

they're out there, just not as much as everything else.  i've got three and now your making me worry about them.

i've seen them for sale for 35 bucks, and they go up on ebay every once in a while, so don't give up on a dream.  if it costs less than 15 bucks, i'd try the chip.

[Major Rock Hardy]

It's true, but unfortunately I'm not 100% sure it's the chips themselves. One of the joys chips is outputting 5v no matter whether I go right or left, but this could be because some other link in the chain is 'broken'. The chip may be fine, but the path of 'info' "into" or "out of" the chip might be the problem, and there are a number of other components in that path.

I'm looking at my Midway 49 way stick right now and its PCB looks like a 2-layer, which should be relatively easy to check the continuity of - using a multimeter set on ohms.  I would imagine the Williams wouldn't have a much more complicated PCB since the technology is basically the same (from what I understand).  That would tell you if any of the PCB traces are just plain broken.  If you find any, you can repair the connectivity with a piece of wire soldered between the components that should be connected - but aren't.

If you don't find any broken traces you can also look at the inputs of the chip itself.  Since one of your sticks has a working X axis you can use that as a basis for determining whether the X axis inputs to the chip on the stick w/ the broken X axis are good or not.

If you get to this point, and you can't find anything different on the input side of the chip (what is the part number on this chip on the Williams stick by the way?), then it is probably the chip.

Rock

[quarterback]

what is the part number on this chip on the Williams stick by the way?

The chip has this printed on top:

Code: [Select]

Toshiba 8840HB
TC4011BP  Japan

I'm looking at my Midway 49 way stick right now and its PCB looks like a 2-layer, which should be relatively easy to check the continuity of - using a multimeter set on ohms.  I would imagine the Williams wouldn't have a much more complicated PCB since the technology is basically the same (from what I understand).

There's an exploded view of the joystick on page 21 of the sinistar manual

http://www.arcadedocs.com/vidmanuals/S/Sinistar.pdf

And here's a photo of one of my joy's PCB

[Major Rock Hardy]

your chip is indeed a quad, 2-input nand part:

http://pdf.alldatasheet.co.kr/datasheet-pdf/view/31627/TOSHIBA/TC4011BP/datasheet.pdf

so pin 14 is VDD, pin 7 is VSS (ground), the rest of the logic should go like this:

pin 11 = !(pin 12 & pin 13)
pin 10 = !(pin 8 & pin 9)
pin 3 = !(pin 2 & pin 1)
pin 4 = !(pin 5 & pin 6)

hope this helps,
Rock

[madmagician]

not sure if you've noticed this but it looks to me like C2 is missing (upper right corner)

[RandyT]

not sure if you've noticed this but it looks to me like C2 is missing (upper right corner)

That's intentional.  It's missing on mine too

RandyT

[quarterback]

your chip is indeed a quad, 2-input nand part:

Yeah, I found the part on mouser for $.48/chip and I've been doing my best to educate myself on what all this means.   So I now get (conceptually) what a NAND gate is, but I'm still not clear on the best way to test the chip before I do something drastic like pulling it off the board.

I can't meter anything directly off the chip while the joystick is assembled, but I can power up the chip by connecting the PCB to my GPwiz49.  So if somebody could give me tip on how to mose easily test each of the four NAND input/outputs on the chip, I'd appreciate it.

Thanks

[SirPeale]

I've got a 49 way that does the same thing.  At first I thought maybe it was one of the optics, but replacing that didn't help.  Now I'm curious what part got toasted, and if it's replacable.

[quarterback]

I've got a 49 way that does the same thing.  At first I thought maybe it was one of the optics, but replacing that didn't help.  Now I'm curious what part got toasted, and if it's replacable.

After having this problem last year, I immediately started looking for a replacement 49-way that worked.  In that pursuit, I talked to a number of people selling them and found 1 or 2 people who reported the same thing with their Arch Rival's 49-ways.  So it's not a completely uncommon thing.   

It seems likely that it's the nand chip as people in this thread have suggested.  My traces all look clean and since we can rule out the optical sensors, there's not much left on the board.   It just seems odd (to me) that so many nand chips just 'go bad'.   But, since these are used joysticks, it might be something as simple as a spilled-Coke "short" across something that burns out part of the chip.  I don't know.

Since I have two which are both bad, but bad on different axis', I was going to test the chips by comparing the two with my multi-meter and see if I could really confirm that it was the chip... but I'm not sure how to do that, or if it's even possible.

A replacement nand chip is cheap and, if you wanna' try it, I'm sure there are those who would be interested to know if that is the answer.   I currently have two working ones that I bought off of eBay but I have no idea if they're going to 'go bad' in the future.  I didn't talk to anybody who had picked up working joysticks that then failed while in their possesion, so nobody was able to determine exactly what might have gone wrong.