(Topic ID: 224245)

Gottleib System 80 (HH) Only displaying ODD NUMBERS!

By sneakerpin

3 years ago



Topic Stats

  • 10 posts
  • 2 Pinsiders participating
  • Latest reply 3 years ago by sneakerpin
  • Topic is favorited by 1 Pinsider

You

Linked Games

Topic Gallery

View topic image gallery

20180830_150437 (resized).jpg
20180830_150354 (resized).jpg
20180830_150336 (resized).jpg
20180830_145452 (resized).jpg
20180830_091441 (resized).jpg
20180830_092050 (resized).jpg
Capture (resized).PNG
20180828_233426 (resized).jpg

#1 3 years ago

Hello,

I have a strange problem that came out of nowhere. It appears my MPU logic can no longer display even numbers!

There are no EVEN numbers! Not in the score, High Score, or even in the bookkeeping/credit/match display!

My Haunted House boots up and displays all "1"'s in all five 6-digit displays instead of zeroes. Credit will only show odd number, so at zero credits, it's '1' credit.

I can coin in, (credit display shows 1 even on zero credits. Credit display never reaches zero, but game will not start unless I add a credit if previously zero). If I add 1 credit, credit display remains at 1. If I add 2 credits, display reads "3" if I add 3 credits, it remains at 3, If I add 4, it jumps to 5 and so on.

Game scoring works exactly like the credit display, never an even number. All sound, playfield elements work, lamp matrix, etc. Only even numbers are not working. Scoring accumulates, but even numbers are incremented by one to odd...

Even during the display test routine, my CPU just repeats 1, 3, 5, 7, 9 twice each! During bookkeeping menu, never an even number although if there is an even numbered test, as in Step 16, Lamp Driver Test, the test functions fine, but display reads "17".

I've done the ground mods a long time ago, and the only thing I have done since was to replace player 4 display with a fresh one.

I measured correct voltages on the PSU Test Points.

I have not re-pinned my cables, and I know what you are thinking. It's time. This seems like a logic issue more than a connector issue, however, and I'd rather find the source on the MPU before re-pinning hundreds of wires that work.

Anyone ever experience this?

Thanks!

20180828_233426 (resized).jpg

#2 3 years ago

possibly an MPU issue. The CPU sends out the digits as BCD (4 digit binary) through an inverter (Z18) to three three BCD decoders (Z19,21,23) which are what output the actual segments to the displays. If the lowest binary bit was locked on it would force all numbers to display as odd, so I'd suspect the Z18 inverter or the U5 RIOT chip driving it of having a bad output, or maybe a bad trace/connection between them

Capture (resized).PNG
#3 3 years ago

Thank You Zacaj!

I was looking at my schematics, and funny, am looking at the datasheet for a 6532 RIOT when I saw your reply.

I wasn't sure if it was a segment or strobe issue, or the RIOT itself.

I'm re-learning what strobing is right now.

This is a great start, and I bet you are right. I haven't a clue how this just happened all of a sudden.

One other thing to note, my game won't boot correctly without the batteries backing up the RAM... Weird. Maybe I have a bad 5V trace somewhere too.

Thanks again!

Tim

#4 3 years ago

So, am I guessing right, that on power up, in attract mode, or player1 up before plunging, I should measure a low on my logic probe on U5 pin 21 and also Z16 Pin 1?

Can I measure using the same +5 and GND, or do I have to reconnect my probe to each chip? It's been a long while.

Thanks.

#5 3 years ago

I'm not sure which way it would read, but I'd expect that if this was the issue, especially if you get multiple digits displaying, that you'd see pulsing on the D2,3,4 lines, but D1 (U5-24, Z18-8,9) would stay high or low (Z17-8 and 9 should be the opposite of each other)

5V+Ground should be the same anywhere.

#6 3 years ago

Thanks. I'll report my results when I get a chance to test later today hopefully.

Tim

#7 3 years ago

Getting home, I did a bit more digging, and dug as deep as I could into the schematic, and truth tables on the IC datasheets.

I deciphered the signals as follows, and I could be wrong in places, so if anyone wants to correct me, I'd learn more:

The 6502 issues the address and data signals of scoring, and lamps and solenoids, etc. for the entire system. Each RIOT chip is like a Peripheral Interface Adapter chip, or "PIA" as known in Williams System 11 games.

The specific 6532 (RIOT), in the scoring case, is U5 (6532-2) and depending on which clock cycles of the 6502, is engaged with refreshing ALL of the display information and this happens repeatedly as long as the game is on.

U5 has two bi-directional 8-bit peripheral busses to drive all five 6-segment displays and the one 4-segment display in Haunted House. Bus B Handles the bit pattern to determine segment pattern of each digit (0-9), while Bus A handles the strobe, or "frozen/latched" segment pattern of each digit (position 0-5).

For SEGMENTS:

Working a bit backwards, the 7448 IC's have a 4-bit input to drive all of the segments to create 0-9

Prior to the 7448, there is a 74175 D-Flip Flop, it's job is to "latch" the 4-bit BCD so that other clock cycles of the system don't clear the segment pattern going to the 7448. The output of the flip flop is inverted back to what it was supposed to be (Q!), because the input of the flip flop is also inverted.

The 7404 inverter preceding the flip flop is used as a buffer?

The 7404 receives it's 4-bit BCD from the RIOT, and coupled with the RIOT's A bus doing the digit strobing, we have our display logic in a nutshell (I hope I got this right)

So, I'm really down to the 7404 (Z16), pin 9 (input) and 8 (output). If the output remains HIGH, I will not see even digits displayed because the truth table of the 7448 tells me so. Since the 74175 acts only to preserve BCD data between refresh cycles, and all score displays act the same, the problem might just end up with a simple inverter chip. Oh man.

BTW, while the B bus of U5 appears strictly for driving segments... the A bus has other data lines that appear to control "Segment h" (The extra "1" on each 6 segment display) and also the notorious SLAM SWITCH, as well as some "enable input" lines going to the switch matrix.

@zacaj, you've done well. I haven't troubleshot this deep since the early '90s. Man, this was fun looking at datasheets and schematics!

I hope to find my set of "chip clips" buried in the garage, as I am really nervous about testing the board by leaning over the cabinet and having that damn back box door in the way... I suppose if I had a real pin repair work space, I'd have a spare PSU on the bench...

I'll try to do this tomorrow.

Thanks again, @zacaj. I hope this thread helps another baffled 7-segment display pinhead!

Tim

#8 3 years ago
Quoted from sneakerpin:

Getting home, I did a bit more digging, and dug as deep as I could into the schematic, and truth tables on the IC datasheets.
I deciphered the signals as follows, and I could be wrong in places, so if anyone wants to correct me, I'd learn more:
The 6502 issues the address and data signals of scoring, and lamps and solenoids, etc. for the entire system. Each RIOT chip is like a Peripheral Interface Adapter chip, or "PIA" as known in Williams System 11 games.
The specific 6532 (RIOT), in the scoring case, is U5 (6532-2) and depending on which clock cycles of the 6502, is engaged with refreshing ALL of the display information and this happens repeatedly as long as the game is on.
U5 has two bi-directional 8-bit peripheral busses to drive all five 6-segment displays and the one 4-segment display in Haunted House. Bus B Handles the bit pattern to determine segment pattern of each digit (0-9), while Bus A handles the strobe, or "frozen/latched" segment pattern of each digit (position 0-5).
For SEGMENTS:
Working a bit backwards, the 7448 IC's have a 4-bit input to drive all of the segments to create 0-9
Prior to the 7448, there is a 74175 D-Flip Flop, it's job is to "latch" the 4-bit BCD so that other clock cycles of the system don't clear the segment pattern going to the 7448. The output of the flip flop is inverted back to what it was supposed to be (Q!), because the input of the flip flop is also inverted.
The 7404 inverter preceding the flip flop is used as a buffer?
The 7404 receives it's 4-bit BCD from the RIOT, and coupled with the RIOT's A bus doing the digit strobing, we have our display logic in a nutshell (I hope I got this right)
So, I'm really down to the 7404 (Z16), pin 9 (input) and 8 (output). If the output remains HIGH, I will not see even digits displayed because the truth table of the 7448 tells me so. Since the 74175 acts only to preserve BCD data between refresh cycles, and all score displays act the same, the problem might just end up with a simple inverter chip. Oh man.
BTW, while the B bus of U5 appears strictly for driving segments... the A bus has other data lines that appear to control "Segment h" (The extra "1" on each 6 segment display) and also the notorious SLAM SWITCH, as well as some "enable input" lines going to the switch matrix.
@zacaj, you've done well. I haven't troubleshot this deep since the early '90s. Man, this was fun looking at datasheets and schematics!
I hope to find my set of "chip clips" buried in the garage, as I am really nervous about testing the board by leaning over the cabinet and having that damn back box door in the way... I suppose if I had a real pin repair work space, I'd have a spare PSU on the bench...
I'll try to do this tomorrow.
Thanks again, @zacaj. I hope this thread helps another baffled 7-segment display pinhead!
Tim

That all sounds right to me! I assume the inverter is just so all those other chips aren't drawing directly off the riot...

#9 3 years ago

UPDATE 1:

Removed the backbox swing door, not that difficult to do. Allowed me to get up close to the inverter IC in question (U16 74LS04).

Input Pins 1, 3, 5, 9 all pulsing LOW
Output Pins 2, 4, 6 all pulsing HI/LO
Output Pin 8 is LOW, no pulsing

Pin 8 is what I suspected, lo and behold, I actually have a new 74LS04 and DIP socket that I bought for another project that I gave up on... Hurrah!

I pulled the MPU board off and will clean the finger contacts as well.

I'll update when I get the work done.

20180830_091441 (resized).jpg20180830_092050 (resized).jpg
#10 3 years ago

UPDATE 2:

De-soldered the Inverter IC and soldered in a DIP socket. Put new IC in socket, reconnected everything and viola!

I have odd and even numbers again!

Thinking back, maybe I left the power on when I (dis)connected that player 4 display connector.

I hope this thread helps another Pin-sider.

Cheers,

Tim

20180830_145452 (resized).jpg20180830_150336 (resized).jpg20180830_150354 (resized).jpg20180830_150437 (resized).jpg

Hey there! Got a moment?

Great to see you're enjoying Pinside! Did you know Pinside is able to run thanks to donations from our visitors? Please donate to Pinside, support the site and get anext to your username to show for it! Donate to Pinside