You're not the only one with a headache Terry.

To summarize the current state of affairs.

- Machine powers on, tests ok and enters audit mode 100% of the time. (haven't bothered installing batteries yet)
- Machine will sit in attract mode without issues for as long as its let to sit.
- All illumination works, both the general and logic powered lights fire or stay lit as needed by the game.
- Segmented displays all work reliably and are in excellent condition.
- Self tests for Displays, Sound, Solenoids and Switches all suggest that there is no current problem with any individual component. All switches activate the correct value and only the correct value as per the documentation.
- Machine accepts credits, starts a game and plays. All switches work, all solenoids fire as required and the score increases proportionally to the shots being made as far as I can tell.

THE ISSUE: After a random amount of time, 1,000,000 or 3,000,000 points gets added to the active player's score total, irrespective of any shots being made during the game. This can happen multiple times during play. Gameplay and scoring continues as normal between these incidents.

I'm not as confident as I once was that the issue is with the switch matrix. That's the main reason I started looking for alternative answers.

This video I took illustrates the issue better than I can describe it.
https://www.youtube.com/embed/V55jTQiQK_M?autoplay=1&rel=0

I'm still trying to define the problem. Also, defining the problem is much easier when you're knowledgeable about what systems interact with what other systems and in what capacity. Essentially what I've been trying to do is to get assurance that a system I have mentally ruled out, should legitimately be ruled out as a cause for concern.

Out of all the reading I've done regarding Switch Matrices and their logic and the steps used to troubleshoot them, my conclusion was that based on the evidence that I can see and understand, there's decreasing chance that the issue is due to a switch matrix wiring/diode problem and more with something on the boards. I've been trying to find the edges, so to speak, which is how I learn.

Just for clarification, what I see in the video is that when the 1 million first hits, there's no drop target that gets touched, the ball bounces off the post at the base of a ramp. 1 million points is awarded, the ball drains then, 3 million is awarded, the score is sitting at 4,022,210 at which point the three steps of the bonus countdown plays and the score increases in 1000 point increments four times, for a total of 4,026,210.

Does that still jive with your chip theory Hans? Don't get me wrong, I hope it does.

Patofnaud, that is good information but in this instance, it's also information I've already used, to great effect actually.

Earlier, I did do a systematic test of the switches while in diagnostic mode. Originally, the left inside rollover was activating when tapping the playfield in front of the left kicker, that's since been adjusted and is no longer cause for concern.

I've engaged each switch individually during test mode and ensured that they all registered correctly within the self-test.

I've even activated multiple switches at the same time to ensure that only those switches were listing as being active during diagnostics, at no point during the testing has there been any out of spec response from the playfield hardware, it all appears to be working correctly.

The driver board is correctly built. W9-W16 are all 0 ohm jumpers as per the sys7 spec requirements.

I've also replaced all the male ".156 connectors on each PCB as they were the original round pin style and many of them were bent, plus the soldered connections were very questionable.

Long odds on odd problems seems to be my stock in trade.

I've pressed the diagnostic button and had it pass, both on the bench and once the boards were put back in the machine, is that what you're referring to Hans?

If there's another self test to run or initiate I'm not aware of it.

I'm in the midst of doing a shot by shot tally of all the points scored in the recording for all 4 players. I'll post it in its entirety shortly.

First, apologies for the length of this post, but hopefully it's instructive.

EDIT: Informed there was an error in the operator's manual. The rollunder (switch 14) on the left middle ramp point value is actually 100 in the game and not 500 as listed in the manual. Hence, there is only one spot (where it scores 10 points) that I'm concerned about now.

I will do what you're asking Terry regarding the switches you questioned me about and barakandl, if for no other reason than my own peace of mind and yours. Unfortunately I can’t yet answer your question Terry, I haven’t checked that issue since I put the boards back in. barakandl, we must be seeing different things when looking at the footage, because I don't see the score jumping like you do, at any point.

I’ve dropped the speed on the video and analyzed it like I was 15 again and trying to decipher scrambled softcore late night porn… ahem… anyway, this is what I’ve come up with. What follows is a shot by shot running commentary with accompanying points value to the best of my knowledge/ability based on the Operator’s handbook ruleset and switch value point totals.

Player 1: Ball is launched, passes through the upper play field without hitting anything, rolls under switch 14 for what the operator’s manual says should be 500 points/mystery (only 100 is awarded) and then alternates between the left and right kicker 3 times for 10 points each. Score is sitting at 130. Ball does not appear to hit any more switches until tripping the out hole switch after draining. Multiplier adds (x1?) 1000 points and player 2 begins their turn.

Player 2: Increments up to 2,600 points before the ball is even launched... this seems strange to me. The initial increment is 1000 followed by 100 point increments up to 2,600. You can hear the audio count it out before the player 2 ball is launched. It looks like the ball then hits 2 drop targets for 1000 points each (slightly unclear) bringing the total to 4600 points and then hits a drop target on the right for another 1000 points (5600 points total and so far). Three shots of the pop bumper for 1500 points (500 per) takes the score to 7100. This is where it gets a bit sketchy. Following this, there is definitely a 500 point pop bumper hit (7600 total) plus maybe another 1000 each for the last two drop targets on that bank lights the magna save and brings the score to 9600, an additional 10,000 points is added for dropping the full bank (new total of 19,600) and then the pop is hit a final time and the score sits at 20,100. One final hit on each opposing banks of drop targets on the upper play field brings the score up to 22,100.

On the way to the lower play field, the ball should, by rights, activate the roll under (switch 14) and post up 500 points I think, when in actual fact, that switch activating only produces a total of 100 points, raising the score to 22,200. This seems like the first inconsistent scoring since the very beginning of Player 2’s turn. The ball is then shot, hits the ramp post and suddenly the score jumps to 1,022,200 points. A kicker is activated (adding 10 points) then after draining, the score jumps another 3 million points (total is now 4,022,210). Prior to activating player 3s turn, the score increments up another 4000 points incrementing by 1000 for a final total of 4,026,210.

Player 3: Before tripping any scoring switchings, player 3 is awarded 1000 points. Then, the score increases incrementally by 100 points (18x) for a total of 2,800. Now, during this run, the roll under gets activated I think, as the ball travels from the upper play field down to the lower one via that ramp, but at no point does the score jump by 500, so I’m willing to bet that, like in Player 2s turn, the switch activation only merits a 100 point increase. I can’t know for sure however just from the video as the timing is too difficult to determine. Following this, the turnaround (switch 23) is activated for 5000 points bringing the total to 7,800. The ball is shot back up the ramp (100 points added) making the score 7,900 it falls right back down adding another 100 to get the total to 8,000. I lose track of the ball here, but it looks like 1000 is added to the score twice for some reason making the total 10,000. I was thinking maybe an inside rollover was tripped and the ball was left to drain, but those are supposed to award 2000 points in one go, so I have to assume the score wouldn’t increase via 1000 point increments. Possibly it has something to do with hitting the turnaround? But I can’t be sure, I’m not 100% on how those bonuses are applied.

Player 4: The game begins identically to player 3. Before tripping any scoring switchings, player 4 is awarded 1000 points. Then, the score increases incrementally by 100 points (18x) for a total of 2,800. The ball is finally launched, 2 drop targets are hit for 1000 points each and the score sits at 4,800. The ball eventually makes its way into the lockup trough and for some reason (I believe there was already a ball in there from an earlier game) player 4 is awarded 10,000 points (5,000 per locked ball maybe? It could also be a multiplier being active I’m not aware of though) bringing the total score to 14,800. Two drop targets are hit for 1000 points each (16,800) then the ball makes its way down to the lower play field, tripping switch 14 (another 100 points) and then the lower play field eject hole shot is made, locking a 2nd ball and initiating multi ball.

At this point it looks like multi ball is activated (2 balls) so points scoring is doubled. The first ball is ejected out of the upper play field lock and 10,000 points gets added (26,900) I think this is from making the lower play field eject hole shot to initiate multi ball in the first place, as the 5,000 points you’d normally receive weren’t added immediately, so likely this value is applied after the 2x multi ball multiplier. The ball travels to the lower field across switch 14 again adding 200 points for a total of 27,100. I lose track of the ball here, but it likely drains, and then the ball in the lower play field lock is ejected and 10,000 points are added bringing the total to 37,100. A kicker is hit for 20 for a total of 37,120 and the ball travels up and then back down across switch 14 but only registers 10 points for some reason. (37,130 total now) I’m very confused about this. Additionally after this, another 10 points are added, maybe multi ball scoring has stopped so it was likely a kicker hit based on the ball trajectory, that added the the points, for a total of 37,140. The ball is then left to drain, a final 1000 points get added and the video ends with Player 1 becoming active again.

Phew…

Hmm, good to know, I'll update that in the handbook.

It would be one thing if the displays just showed the wrong amounts, it's the fact that as far as the game logic is concerned 1,000,000 points and then 3,000,000 points are being scored, seemingly independently from any switch input being triggered. It's being processed by the CPU as legitimate because the resulting rewards from making those kinds of totals are being activated during gameplay.

Isolating and resolving whatever is causing this problem is my focus.

I remain very curious about why the player 2 score jumps to 2,600 and the player 3 and 4 score jumps to 2,800 points before the even enters the playfield proper. That has me completely stumped.

The 5101 is definitely socketed.

I'm thinking that grabbing a pinforge 5101 adapter might be the best option.

One of the octal buffers (74LS245) was taking perfectly clean signals and dirtying them up a little bit when I was originally trying to bring this thing back to life. After installing some known working ROMS, the boards powered up and things seemed to be working fine so I didn't look into it any further.

I've ordered 5 new octal buffer chips because they're like $4 for 5 chips, shipping included and it will be a cinch to replace since it's already socketed. I will likely also order a Pinforge adapter to replace the 5101 ram completely. Will update once the parts arrive and I get some time to work on it.

Is there a known good supplier of 2114 ram or equivalents that anyone's aware of? I haven't been able to find anything in the usual places and ebay is always such a gamble...

I don't actually know, there's no obvious listing in the manual.

It's listed as Williams Part # 5340-09409-XO | 2114-45 1K x4

Anyone ever used Jameco?
https://www.jameco.com/webapp/wcs/stores/servlet/Product_10001_10001_38957_-1
For around $25 they're saying I could have 6 shipped to my door. So far that's looking like the best bet as far as parts/cost goes.

EDIT - Hah, nice Hans, I was literally just checking there. That now looks to be my NEW best bet for sourcing them.

Indeed, I've ordered from there before and have nothing but good things to say about the experience. When I said "new best bet" I meant for this specific piece, not because I wasn't already familiar with the site.

Cheers folks, really appreciate the continued help and like I said earlier, I'll update with findings once the parts arrive.

Figured I should update this mess again.

With all new Pia's on both the driver and mpu board, a new 74ls245 and all new 2114 memory chips, there's been no change in the behaviour of the system.

So I thought to myself I need to divide and minimise the possibilities of what's causing this weird jumps in scoring.

I powered up the game, added credits and then disconnected the switches at the board. Then went around and activated a bunch of the solenoids (flippers, bumpers, pop etc) obviously no score showed up but my theory was that if there was something influencing the switch inputs unrelated to the switch circuitry then it might reveal itself. No luck, there appeared to be no issues and all the hardware was working normally.

Within 20 seconds of reconnecting the switch matrix the scores jumped. I was not playing just watching.

The only active switches were the ones registering the balls in lock and in the shooter lane.

Going to mess around a bit more. Feel free to chime in with any ideas please. Even if it's just random brain droppings : )

Hah, ok NOW I've got some actual hard data.

Everything connected
Started a game
Removed all three balls
Let it sit for ~5 minutes (no issues, score did not change, no problems detected)
After the 5 minutes elapsed, I banged on the flippers for ~2 minutes (again, no issues)
I tapped the left outlane switch once, the correct 5000 point value showed up
I then immediately disconnected the switch matrix from the driver board (2J2 and 2J3)
Exactly 1:00 later, the score jumped by 1 million
Exactly 1:00 after that, the score jumped by 2 million in 1 million point increments
Exactly 1:00 after that, the score jumped by 3 million in 1 million point increments
Exactly 1:00 after that, the score jumped by 4 million in 1 million point increments
etc etc etc

So... what the hell does that tell me?

It really is the strangest thing.

@ Hans, it doesn't do anything in switch test mode. I can bang on switches for minutes at a time with absolutely no problems. They all register correctly and there's no action on the scoreboard other than the switch activation indicator number.

@ Toyguy, I'm trying man, believe me. I live in a smaller city and unfortunately I'm not well connected with the local hobbyists. I thought I had a line on someone a while back but wasn't able to make contact at that time, I guess I should try again. Although at this point, maybe dropping $300 for a Rottendog combo board isn't the worst idea in the world.

As for the RF interference... other than the wireless router in the basement, I barely get cell service next to the pin. So I'd say it's likely not that. It's the fact that it's so predictable tends to make me think it's a hardware problem being produced or activated by the game code/software.

It just seems to be SO close to working. It'd be such a shame to stop the process now.

I really want to solve the issue. I don't know if I can properly communicate just how much I want to solve the issue.

@ Mackin - It's an early version with the Sys6 power supply. I have tested with the soundboard disconnected. I have not tried testing without the GI yet.

I did go through all the "Before you apply power" PowerSupply tests as per Clay's instruction in his guides, so I'll have to refresh my memory on whether or not that diode on the GI relay got looked at. Q34 and Q35 I believe tested OK if my memory serves.

A few hours of study/work to figure out how to properly test all the components does not phase me in the least.

In addition to the probe and my DMM, I have a basic chip tester, I've applied it to everything it recognizes. As far as it's concerned, the chips are alright, but I know that it's cheap and not 100% reliable. Where at all possible, I've sourced multiple spares and tried them too to see if there was any change in behaviour.

All the sockets are newly installed and I've pinged all the connections with my DMM for proper continuity and confirmed no shorts.

IC1 is socketed and I've tried 2 CPUs with no change in behaviour.
IC2 is not socketed and has not been tested yet.
IC3 is socketed and passes according to the chip tester.
IC4 is NOT socketed. I could socket this chip and swap with IC3 to see if that makes any difference.

IC15 is socketed and passed in the chip tester, although I have plenty of old arcade boards, I can source spares
IC16 is socketed and I've tried a few 2114s with no change in behaviour
IC17 is socketed, although I haven't tried my spare ROM yet so that's something to look at.
IC18 is socketed, multiple PIAs have been tried with no change in behaviour.

Testing the driver board transistors is something I've yet to look at and none of the 740X chips on the driver board are socketed. Could you expand on the methodology one would use when testing the 740X chips?

Ahh, yeah ok. Also I already have a copy of the guides, was just wanting to make sure you weren't referring to a method of testing the 740Xs or the transistors that wasn't already outlined in them.

Thanks for the info, I'll update the thread once I do another round of testing.