Gottlieb Caveman: Stops accepting inputs after game start

GB Ward

New member
Joined
Mar 10, 2016
Messages
134
Reaction score
8
Location
Arkansas
Gottlieb Caveman: Stops accepting inputs after game start

Hey guys. I got a Gottlieb Caveman in a couple weeks ago. It's been a neat one, but I've come to an issue that is out of my knowledge.

The video explains most of it, but here's the repair history:
1. Won't power on > bad fuses and shorted components in the PSU.
2. Won't boot > bad connectors, and no 5V to 6502, then reset low from corrosion.
3. Won't accept inputs > corrosion on input resistors / capacitors.
4. Inputs not making it to U4 > Bad z13, z14, and z12
5. RIOT U4 not communicating with data bus > Bad z10 not toggling ^BAB9, but oddly enough driving BAB9. Guess it was a shorted input on z10.
7. Game now passes test modes. I have tested ROMs, RAMs, and RIOTs and are known good.

What it's doing is not accepting inputs, additional players, or scoring after start is pressed, and video module will never leave attract. Video module will not go into test mode. I don't think I'm getting proper game sounds either.

Also, game will not accept inputs unless DIP 6,7,8 are ON. I don't know if this is weird or normal. I don't have the manual, but I do have the DIP page, and it's not helpful. I have tested all the dip switches and they are working properly, and their diodes are good.

My theory is that A1 and the video module are not communicating properly, but I have little knowledge about this.

Video:
https://youtu.be/-S-oIWXyKfI

Ah yes and I have checked my tilt switches. They are open and the slam switch is closed and they all work. Tilt does not work after the game mode starts, but the slam switch does.

Thanks in advance!
-GB
 
Last edited:
This is known to be a bear of a machine which more often than not is trashed (literally) because of its complexity. Normally, if you don't get anything on the video the machine is usually parted out or destroyed as replacement boards are stupidly expensive. Todd Tuckey of TNT had at least a half a dozen parts machines for sale at Pinfest a few years ago. I think the ones he didn't sell he threw off his roof ;)

The good news for your machine is that it looks like the video cage is working which means your problem is in the pinball side. Your problem seems to lie somewhere in the switch matrix. The only part of the matrix you did not mention is Z11. After you start a game is U4 still strobing port b?

I'm not sure how you tested the RIOT - I use Marco Albus' Sys80 test ROM. In addition to testing the RIOT memory it also verifies the RIOT timer interrupts as well as verify the functioning of the ports for both inputs and outputs as well as insuring all DIP switches can be read by the system. If you'd like a copy PM me - he's given me permission to redistribute his code.

Of course all this assumes you have mitigated all corrosion on the board. The fact that the game is not accepting inputs unless DIPs 7-9 are on further indicates a switch matrix problem since the only things they do are set how the coin chutes are pulsed. Different returns/different buffers points to a RIOT issue from where I sit.

Good luck!

viperrwk
 
I tested the RIOT chips by several methods: 1) initially swapping them out with the display section that appears to work and observing no changes in either system. 2) Probing all PA and PB lines with Oscope to verify they are driving high and low. 3) Making a memory map from the schematic and testing RAM with the Fluke, (0x0080 * 3 = 0x0000-0x0180). 4) Forcing PB and PA lines high and low with the fluke and testing their output with oscilloscope. 5) Testing Dip switches with the fluke, ie, READ @ memory address of DIP, LOOP, and turn switches on and off to make sure their hex value appears on Fluke, and that all off reads 00 and all on reads FF.

I pulled Z11 and it passed the vector test ok, and the outputs are strong on the oscilloscope.

U4 is still strobing on all lines after the game is started, and the switches are returning strobes when enabled. However, when all the switches are open I am not getting anything on the return lines. According to this: http://www.flipprojets.fr/pdf/Video_Caveman.pdf The video system is connected to R0 - R7, so maybe I should expect some data from them.

This page also suggests that the video module's only input is from the lamp matrix on A3J2 and A3J3. I have not troubleshooted the lamp matrix yet.

So here's the new plan:
1. Run back through the A1 switching logic with the oscope to make sure it's working.
2. I will check U4 U5 U7 U8 U9, and U6 U17 U11 and U18 of the video module
3. Check lamp diodes and lamp matrix.

Looks like I need to buy the schematics.
Thanks for the help Viper.
-GB
 
Last edited:
Page 3 contains my answer:

http://pinwiki.com/wiki/images/2/2e/Ontarget_octnovdec82.pdf

Someone in the history of the machine plugged A23J2 in upside down. Fortunately, "This plug can be reversed without damage to the system." This didn't even occur to me, as all the other connectors are keyed, and it has obviously been plugged in this way for years.

Caveman is fully functioning now.
Thanks!
 
Back
Top Bottom