(Topic ID: 352558)

Odd switch matrix behavior - Fish Tales

By Jahkub

59 days ago



Topic Stats

  • 10 posts
  • 4 Pinsiders participating
  • Latest reply 59 days ago by Jahkub
  • No one calls this topic a favorite

You

Linked Games

Topic Gallery

View topic image gallery

switch_matrix_theory_of_operation.jpg
#1 59 days ago

Hi, I'm looking at a friend's Fish Tales and what I once thought was a simple diode issue, is actually on the mpu and has me a bit confused.
Rather than just buying and shot gunning the matrix chips, I thought I'd see if anyone has encountered this before. I'll get over there with a logic probe soon anyway.

At this point, with all the switch connectors unplugged from the mpu, this is the behavior.

If any switch in column 1 is closed, AND 41, or 51, or 61, are also closed, a switch will register on the same row as the closed column 1 switch, on column 4, 5, or 6.

Examples
if 14 is closed, 41 will trigger 44 and 41
if 16 is closed, 51 will trigger 56 and 51
if 18 is closed, 61 will trigger 68 and 61
if 15, 16, and 17 are closed, 51 will trigger 55, 56, 57, and 51
ect.

This only happens with 41, 51, 61 and only in combination with any switch on column 1.

I just thought I'd see if this is a failure any of you have seen before. Thanks!

#2 59 days ago
Quoted from Jahkub:

At this point, with all the switch connectors unplugged from the mpu, this is the behavior.
If any switch in column 1 is closed, AND 41, or 51, or 61, are also closed, a switch will register on the same row as the closed column 1 switch, on column 4, 5, or 6.

https://www.pinwiki.com/wiki/index.php/Williams_WPC#Switch_Problems

I am struggling to understand how you are executing the tests to produce the results you are reporting. If you unplug all the switch connectors from the CPU board, to close a column 1 switch you would need to ground the pin on the header. How can you close another switch when you've already grounded that pin on the header? Please help me understand.

The key thing to think of here is that you can only interpret results if the tests are meaningful. A less polite way of expressing this is "garbage in, garbage out". I'm not saying your test is "garbage", I'm saying that you have to have a meaningful test to draw a meaningful conclusion.

#3 59 days ago

Ok,

Go into switch edges test.

On the bottom of the CPU, disconnect the plugs at the bottom of the CPU board. You should have nothing connected to J205, J206, J207, J208, J209, and J212.

Put an alligator clip on J207, pin 6.

Touch the other end of the alligator clip to J209, each pin in sequence, and you should get switch 61, 62, 63, 64, 65, 66, 67, 68.

Move the alligator clip to J207 pin 5

Touch the other end of the alligator clip to J209, each pin in sequence, and you should get switch 51, 52, 53, 54, 55, 56, 57, 58.

You can do the same thing down J207, pin 4, pin 3, pin 2, pin 1.

I think that's what you are doing, right?

The purpose of this test is to make sure that the cpu can register all the switch rows, and all the switch columns.

However, if you were to connect two different connections DIRECTLY, without the switch diodes being in the circuit, normal behavior would be for it to register switches along the lines that you've directly connected inappropriately.

The test on the bottom of the CPU board for the switch rows and columns is not useful in combinations of connections... only test one connection at a time.

Multiple connections require the diodes on the switches to prevent the circuit from triggering switch registration inappropriately.

By the way, you've revealed why all the switches in pinballs have diodes attached.

#4 59 days ago
Quoted from PinRetail:

On the bottom of the CPU, disconnect the plugs at the bottom of the CPU board. You should have nothing connected to J205, J206, J207, J208, J209, and J212.

If you remove J205 then you cannot enter the diagnostic tests. J205 is only related to the direct input switches and not the switch matrix switches.

I know you specify the order here as:

  1. Enter switch edge test.
  2. Unplug connectors.

The operation must be done in that order if you remove J205. You cannot remove J205 and then enter switch edge test.

Just being nitpicky and detail oriented to prevent a post saying "I removed all the connectors and now I can't get into tests". Also note that it is safe to remove the connectors with the power on since the switch matrix runs on 12V. You can certainly unplug everything except J205 and power on but doing so will potentially result in a "check fuse" message since SW24 is not detected.

#5 59 days ago
Quoted from DumbAss:

If you remove J205 then you cannot enter the diagnostic tests. J205 is only related to the direct input switches and not the switch matrix switches.
I know you specify the order here as:

Enter switch edge test.
Unplug connectors.

The operation must be done in that order if you remove J205. You cannot remove J205 and then enter switch edge test.
Just being nitpicky and detail oriented to prevent a post saying "I removed all the connectors and now I can't get into tests". Also note that it is safe to remove the connectors with the power on since the switch matrix runs on 12V. You can certainly unplug everything except J205 and power on but doing so will potentially result in a "check fuse" message since SW24 is not detected.

Appreciate this!

I didn't think deeply about this, and I was concerned about getting everything out of the switch matrix entirely except for the jumper wire alligator clip.

This is a really good example of people who do some things, but don't fully understand the circuits involved.

I always learn things when you step in.

Appreciate the extra help!

#6 59 days ago

I did the test by entering switch test, removing the switch connectors, then I connected the column 1 pin to the row 6 pin (sw16) on the mpu with one jumper wire, then used a separate jumper wire from column 4 to row 1 (41) I did this to other column 1 switches with 41, 51, and 61 as the secondary jumper and it behaved as explained above.

Am I being dumb about this? I could be, it's been a long week already. The game displayed the exact same symptom before I tried isolating the issue to the mpu.

#7 59 days ago
Quoted from Jahkub:

I did the test by entering switch test, removing the switch connectors, then I connected the column 1 pin to the row 6 pin (sw16) on the mpu with one jumper wire, then used a separate jumper wire from column 4 to row 1 (41) I did this to other column 1 switches with 41, 51, and 61 as the secondary jumper and it behaved as explained above.
Am I being dumb about this? I could be, it's been a long week already. The game displayed the exact same symptom before I tried isolating the issue to the mpu.

A couple of points.

  • Writing "removing the switch connectors" is not specific enough. You need to specify which connectors you removed. When I read what you wrote, I assume you only removed J206/J207/J208/J209 and did not remove J212. It might seem nitpicky but it's important because J212 is part of the switch matrix.
  • The test you are doing is not standard. Just connect a single wire from one row to one column. It should register only one switch. You only need to do 16 combinations. One column to all rows and expect a single switch to register. One row to all columns and expect a single switch to register.

The CPU (software) only sees rows and columns. It works in a very simple way. It grounds column 1 and reads the state of the 8x rows. It then grounds column 2 and reads the state of the 8x rows. Do this for all 8 columns and repeat. It really is as simple as that.

switch_matrix_theory_of_operation.jpgswitch_matrix_theory_of_operation.jpg

#8 59 days ago

Sorry I didn't list the history, but I did the single jumper row to column test first, with all 4 switch connectors removed from the MPU.
All single switches register fine doing that jumper test, and sw24 is closed when it should be.

It had me convinced that it was a playfield issue until I seemingly ran into a dead end there and tried to reproduce the issue on the MPU with the two jumpers.

I could be wrong, but using two jumpers to test the odd switch combination behavior seemed logical as column 1 is separated from column 4 by using two separate wires going to two different rows. Right?

#9 59 days ago
Quoted from Jahkub:

Sorry I didn't list the history, but I did the single jumper row to column test first, with all 4 switch connectors removed from the MPU.
All single switches register fine doing that jumper test, and sw24 is closed as it should be.
It had me convinced that it was a playfield issue until I seemingly ran into a dead end there and tried to reproduce the issue on the MPU with the two jumpers.
I could be wrong, but using two jumpers to test the odd switch combination behavior seemed logical as column 1 is seperated from column 4 by using two seperate wires going to two different rows. Right?

Nope, if you plan on testing 2 columns at the same time , you MUST use a diode at the end of your jumpers. You can only use a jumper when testing a single row/column combination.

#10 59 days ago
Quoted from Roamin:

Nope, if you plan on testing 2 columns at the same time , you MUST use a diode at the end of your jumpers. You can only use a jumper when testing a single row/column combination.

Thank you, I needed to hear that. I should start looking at the playfield again then. Bad experiment lead to a bad assumption.

Promoted items from Pinside Marketplace and Pinside Shops!
$ 9.95
Eproms
Pinballrom
 
$ 49.00
Playfield - Toys/Add-ons
Pixels Arcade Games
 
$ 9.99
Eproms
Matt's Basement Arcade
 
$ 220.00
Playfields
Pinball Haus
 
From: $ 27.00
Boards
KAHR.US Circuits
 
$ 49.99
Playfield - Toys/Add-ons
Lighted Pinball Mods
 
5,999
Machine - For Sale
Tacoma, WA
$ 30.00
Playfield - Other
YouBentMyWookie
 
$ 8.50
Hardware
NO GOUGE PINBALL™
 
$ 169.00
12,500 (OBO)
$ 49.99
Playfield - Toys/Add-ons
Lighted Pinball Mods
 
6,400 (OBO)
Machine - For Sale
Pleasant Hill, CA
$ 79.99
Cabinet - Armor And Blades
PinGraffix Pinside Shop
 
$ 59.99
Playfield - Toys/Add-ons
Lighted Pinball Mods
 
$ 5.00
Cabinet - Other
Creekside Hideaway, LLC
 
$ 399.00
Cabinet - Decals
Mircoplayfields
 
$ 49.99
Playfield - Toys/Add-ons
Lighted Pinball Mods
 
$ 9.95
Eproms
Pinballrom
 
Wanted
Machine - Wanted
Watertown, MA
$ 1,059.00
Flipper Parts
Mircoplayfields
 
$ 5.00
Cabinet - Other
Creekside Hideaway, LLC
 
$ 119.99
Displays
FlyLand Designs
 
$ 60.00
Cabinet - Decals
Pinball Haus
 
$ 84.99
Displays
FlyLand Designs
 
$ 14.95
Playfield - Toys/Add-ons
ULEKstore
 
$ 24.95
Lighting - Led
Mitchell Lighting
 

Reply

Wanna join the discussion? Please sign in to reply to this topic.

Hey there! Welcome to Pinside!

Donate to Pinside

Great to see you're enjoying Pinside! Did you know Pinside is able to run without any 3rd-party banners or ads, thanks to the support from our visitors? Please consider a donation to Pinside and get anext to your username to show for it! Or better yet, subscribe to Pinside+!


This page was printed from https://pinside.com/pinball/forum/topic/odd-switch-matrix-behavior-fish-tales- and we tried optimising it for printing. Some page elements may have been deliberately hidden.

Scan the QR code on the left to jump to the URL this document was printed from.