Hello everyone!

This is my first post and I hope I'm putting it in the right place. I've been a pinball machine owner for 72 hours now and my hands are already dirty.

I have a 1981 Bally Medusa. I've done a lot of troubleshooting, so I'll try to be as brief as I can, while still being thorough.

The playfield numeric display has significant flicker and most of the time the digits displayed are +/- 1 from their actual value. For instance, the machine starts the game showing 11 11 14 when it should show 00 00 05.

The right-side number is for "shield of the gods" and the machine actually counts that correctly, the display is just wrong most of the time.

I've checked and reconnected most/all connectors a few times. No change. I pulled the board and checked all solder connections with magnifying glass and using magnification on my phone's camera. I found no problems.

hitting the sides or front of the cabinet does not produce any change. However, if I hit both flippers at the same time and repeat that 1-5 times, it will usually correct the display of numbers and sometimes the flicker for 5-30 seconds, then if messes up again. I suspect this has something to do with causing a temporary voltage sag and that somehow resolves the problem briefly. Since hitting the cabinet produces no change, I do not suspect a loose connection.

I have one other 6-digit display that is up on the scoreboard. However, it was a STERN part and I was originally hesitant to swap them. I finally did it anyway and it seems to have "resolved" the issue. The playfield display now shows the right numbers with almost no flicker. (All displays on my machine flicker a tiny amount and I *think* its normal?) The scoreboard display only uses the one digit on the right for "ball in play" and it shows correctly now too, with no miscount and no flicker.

Somewhere in my troubleshooting, I finally go the nerve to test my power outputs as well:

Power down machine and open her up.
My board is an AS2518-54.
The transformer looks like an anchor from an early battle ship in both size and age.
Unplug J1 through J4 (I think these are all the outputs).
Power on the machine.

TP1: 8.12 VDC
TP2: 182.5 VDC
TP3: 13.97 VDC
TP4: 7.48 VAC (Fluctuates 7.46-7.49)
TP5: 46.3 VDC

NOTE: I also powered down the machine in between checking each test point and I used alligator leads... I can be clumsy.

I'm down to assuming a problem with that particular display board, part number AS-2518-21.

Please let me know your thoughts on my troubleshooting, anything else I should check and if I should repair or replace that display. Are LED boards a simple drop-in replacement?

Thank you everyone. I've already spent hours reading in these forums. This is an amazing resource!!

EDIT (SOLUTION):

1) Quench is amazing.

2) In my case, display flickering on all displays required adjusting RT1 trim pot on MPU board until DSP-INT test point went from 380hz to 420hz.

3) I finally did what Quench and 10,000 other forum posts told me to do and resoldered the header pins on my troublesome display board that was showing incorrect numbers. It fixed it.

EDIT 2 (Pulsing/Flickering controlled illumination *explanation* (not resolution)):

The voltage on the C.I. circuit fluctuates in my machine by about 1 volt, depending on the number of lights on at the time. This appears to be normal operation and electrically expected. What I see is brightness variation in the controlled lapms that stay on for long periods of time, especially noticeable on the scoreboard (IE: Ball in Play, Tilt, High Score To Date).

Quench, thank you for the welcome and reply!

On the solenoid driver board: TP2 is 185 VDC steady. TP4 ~250 VDC. TP4 wobbles +/- 1 VDC.

I haven't changed any settings on the machine. On the MPU board, S27 is "off" and that appears to indicate display of credits is off. It was delivered this way.

All lamps in the machine are currently incandescent. I am strongly considering adapter boards and LED bulbs, but I'm undecided.

The MPU board is new, according to the guy that sold me the machine.

All four requested pictures are attached. (Forum says it resized my images, so I hope quality is good enough.)

MPU Board (resized).jpg

Trouble Display Solder Side (resized).jpg

Trouble Display Topside (resized).jpg

U12 (resized).jpg

Quench, I'm sure you already know this, but you are a Rockstar! Thank you so much!!

I dug out my portable O-Scope and scratched my head for a minute, then realized my DMM had a frequency function.

The MPU was tuned in at about 380Hz. I adjusted it right up to 420Hz and the flickering I could see has virtually vanished.

I flipped 27 on after reading your first reply and played many games with it on. Everything with the "bad" board seemed to be working fine... until...

Interestingly enough, when I cranked up the frequency, the board that was previously having trouble (now displaying the ball number and credit count) started acting up again. I pulled it and looked closely at the solder joints again. I'm 99% sure what you pointed out in the pictures is a strange shadow or reflection. I checked resistance from each pin to another point on the board, while trying to wiggle and move the pin around and didn't see any anomalies. I may try a resolder in the morning, anyway.

My gut tells me the MC14543b chip is bad, but I'm guessing. The fact that it seems to be off by 1 digit, then correct itself is very odd. When I moved it closer to the MPU (much shorter wires) the problem went away. Then, when I increased the refresh rate, it started acting up again. I read some other posts about the resistors and transistors. However, in this case, when something goes wrong, it happens to all of the displayed digits at the same time. They all go up by 1 or return to their correct value simultaneously. That would lead me back to the common component decoding the signal - the MC14543b. It seems like a timing issue of some kind, which I assume is what that chip handles. What are your thoughts? Could that be it or am I out in left field?

TLDR: Take your 6 pack of beer from me and let's never discuss this again...

I don't have any soldering braid and I'm mediocre at soldering anyway, so I just reflowed and added a touch of solder to each of the header pins. I got finished and thought, "Quench specifically mentioned p20." So I went back and hit that one a second time, just in case.

Lo and behold, the damned display works again.

For the sake of completeness, I put it in display test mode and these were the results:

GOOD BAD
111111 111111
222222 333333
333333 333333
444444 555555 *
555555 555555
666666 777777
777777 777777
888888 999999
999999 999999
000000 000000 **

*In the first cycle, one of the 5s on the bad display dropped to 4.
** In the last cycle, some of the 0s on the bad display bumped up to 1.

I cleaned it last night before giving up for the day! It was definitely tarnished!

I'm sure this is a breach of etiquette, but here I go. I'm having another problem - I can start a new thread, but I figured maybe some of the things you've already had me check were relevant. If it's better to do a new thread, please let me know.

The controlled illumination bulbs on the scoreboard are pulsing in much the same way at the displays were. They've been doing it all along. I think they all turn on and off when they're supposed to, they just pulse rapidly when they're on. The GI on the scoreboard and playfield seems fine. All the lights in the inserts that are controlled move so quickly that I can't tell if they're pulsing or not.

PS: I'm choosing the word pulsing because "flickering" feels like it has a random component to it. This isn't random, it definitely has a rhythm to the pulse.

I read some other posts about issues on the MPU, but it seemed like most of those were also associated with other issues, like game resets. I don't think I'm having any of those other issues.

All of my bulbs are incandescent.

Bounces anywhere from 116 Hz. to almost 160 Hz.

I tried a couple different ground points as well as repositioning the probe a few times and varying the force I was applying to the probe. I couldn't get it to stabilize on a frequency for more than a half second or so.

I did not have this problem when measuring the frequency at DSP-INT. That was a rock solid reading until I started adjusting the pot, so I think I'm using the meter correctly.

EDIT: I'm sorry, I omitted the results about playfield flicker. I did not put it in test mode, but I was looking at it last night. The "Gorgon" lights on the playfield stay illuminated for several seconds at a time. The same flicker does appear on those as well.

EDIT 2: I just couldn't leave it alone. I wanted to see what all of that fluctuation was, so I grabbed my cheap scope. Oddly enough, the O-Scope shows a perfect 120Hz. I checked it again with the DMM. DMM registered the DISP_INT exactly like the scope does, but the DMM goes haywire trying to measure the zero crossing... I have no idea why (or if I'm using my O Scope correctly.) Images attached.

DISP_INT (resized).jpg

ZERO_CROSSING (resized).jpg

I don't seem to have any other unusual behavior from the game that I can detect. The lights seem to come on and go off when they're supposed to. Clean, crisp on/off as expected.

I'm attaching pictures and video. The flicker is difficult to see just during game play and video, but its definitely there. It is much more detectable on the scoreboard than on the playfield lights. The "Blue_Flicker" is a close-up of one of the field inserts flickering. I think you can even see the same flicker in the in-field score panel here, as well. The Scoreboard Flicker is a side view of one of the scoreboard baffles. It makes it easier to see the flicker in the video. Then, as you requested - gameplay... Don't judge me. LOL.

I'm poking around, myself and probably wasting time... but there are some things I found odd that I figured I'd throw in here:

1) I have at least 3 sets of lamps under the playfield that are "bridged" and it doesn't look like they should be. For instance, I'm sure they are supposed to be individually activated by the lamp driver, but someone has added another piece of wire between two nearby lamps. I've included pictures. I thought maybe if its activating one lamp, the other wire is possibly feeding voltage back to the display driver and causing unexpected results. I didn't desolder/clip any of those yet.

2) I put my scope on the flickering scoreboard light. I've included pictures of the probe locations, as well as the scope results. The -4.90 correlates for me, as that's exactly what I measure on the test point of the lamp driver board (Granted, it's a +4.90 on the test points.) However, what's up with it showing a 1,000 volt sag in between? That seems absolutely bizarre to me... and if any of this is irrelevant, just ignore it. I like poking around. =P

3) I found one of the "controlled illumination returns" (fancy phrase for ground?) pins on the header with a broken solder joint and repaired that. No change.

4) Again, poking around, I put my DMM across the C1 capacitor on the display driver and the results aren't what I'd expect. It begins to show a Farad reading and then goes back to 00.0. I have bags of ceramic caps, both smaller and larger than 0.01uf and my meter reads them all perfectly. However, I did NOT desolder the C1 capacitor and test it alone, out of the circuit.

VIDEO:
Scoreboard:

Blue Light:

Gameplay:

Lamp bridge 1 (resized).jpgLamp bridge 3 (resized).jpgLamp bridge 2 (resized).jpgScope_Probe_location (resized).jpgScope_Results (resized).jpg

Thanks, Skidave. The video links may work now. I told YouTube to change them over, but they were still set as draft. I've read the lights are supposed to pulse/flicker when the machines are in attract mode, but this continues during gameplay as well. The flicker that I'm seeing also seems to affect the numeric displays.

If it's normal, I can live with that. However, I want to make sure I don't have something (IE: a power problem) slowly cooking parts of my machine. I'd also like to upgrade the machine to LEDs, especially under the inserts. However, I don't want to start attempting modifications if I'm not building on a solid foundation.

I appreciate all the help! Thank you.

Sorry about that. Quench fixed my initial problem with the displays and I asked for additional help with the lights flickering. I've removed the resolved tag for now.

I marked the lamp holders and removed the 3 bridges I could find. I have many lamps that need to be replaced - they're on order. So, I don't know the final results of removing the bridge or if I have other problems on this board. When my lamps come in, I'll get them replaced and put her in test mode. Removing the bridges did not change the flicker/pulse I'm seeing.

Yes, and I'm guessing that's the reason I can't read the cap. I tried several others still connected to other boards with the same result.

I picked up on that, not that I completely understand it. Give it 18 volts, but the RMS is 5 or something like that. It blanks everything and then quickly turns back on the ones that it needs on.

However, what I'm seeing isn't a steady pulse on/off/on/off, like I can see the frequency pulsing it. They get noticeably brighter and then darker, like the voltage isn't steady or the power is dirty or something. I'm not sure. Maybe this is normal.

I'm back to poking around with the O-Scope on the flickering backboard lights. I'm assuming I had some wonky settings the first time around, like set to 100X probe when my probes don't do 100X.

Anyway, I got a better read. They pulse at a pretty steady rate, near 120, sometimes dropping to about 115. However, the flicker I'm seeing happens much more often than this slight frequency drop. So, I think the frequency drop may be contributing to it, but I don't think its the source.

What I'm seeing is the bottom of the waveform is nearly flat. However, the top of the waveform peaks vary by around 0.25 to 1.25 volts and sometimes for several cycles. I think this top end voltage fluctuation is what I'm seeing in the bulbs.

Where the heck does the lamp driver board get 10 volts? Then it hit me like a brick - it doesn't. The lamp driver is connecting the ground side. So I can ground my O-Scope to the ribbon and hook the meter to the power line of the flickering lights. Bingo. I see the exact same thing in the wave form, except no variation in frequency. So, the slight frequency variation is due to the lamp driver board, but if the voltage problem is causing this flicker, that should rule out the MPU and the lamp driver boards.

Now I'm pouring over schematics again trying to figure out where the CI lamps get power. Got it. TP1 on the power supply. Schematic shows TP1 at 6.5 volts. When I measured it with all output connectors off, I got 8.12. I'm going back to raise the playfield and check TP1 with the game on.

My DMM gives a reading of 7.41 to 7.48 VDC on TP1 of the power supply board.

I get the same VRMS on my O-Scope as the DMM reads, however, the scope picks up that same voltage variance on the top end of about a volt.

I'm guessing this is the flicker I'm seeing.

What do you guys think? If this is the problem, how do I clean it up (or can I)? New transformer? New board? I'm including some pictures. It looks to be in pretty rough shape down there.

I figured I'd go ahead and get voltage readings on the power module with the game on. I used my DMM for all readings:

TP1: 7.41 - 7.48 VDC (This is where the CI light flickering is coming from, I think.)
TP2: 237.0 - 242.0 VDC
TP3: 15.39 - 15.44 VDC
TP4: 6.72 - 6.74 VAC
TP5: 44.9 - 45 VDC

Thanks again for all of the help!

POWER SUPPLY:
Power1 (resized).jpg

POWER SUPPLY:
Power2 (resized).jpg

POWER SUPPLY:
Power3 (resized).jpg

VOLTAGE TOP:
Scope1 (resized).jpg

VOLTAGE TOP:
Scope2 (resized).jpg

Do you have a scope you can put on TP1 and see what the output looks like?

Quench

So the flicker I'm seeing in the Medusa seems to be "part of the package." I have about 90 controlled lamps and as the voltage fluctuates, I'm seeing the CI that stay on for long periods of time reflect that voltage waver. This seems to be the case and there's really no problem to be resolved.

I have some warm white PCB/LED and some ghost buster warm LEDs on order from Pinball Life. Their tech says these should work really well with no board modifications, as they have built in rectification, resistor and capacitor to help with latching, ghosting and flickering. If they work, it should help reduce current draw and smooth out the voltage, along with the capacitor in the bulbs themselves helping to smooth that out.

I'm just going to replace the board. I'll fix this one later or something. I don't have the thermal paste, spacer or screw that's supposed to be in there. I can't just paste a big ass heat sink on top of it? XD

I 100% respect this position! I've been a computer tech for 20 years. A long time ago, I quit using terms like "It will..." and replaced them with "It should..." I am 100% convinced the voltage fluctuation is what I am seeing and your explanation of that fluctuation makes perfect sense.

Yes, I did. We had a great discussion. Our initial discussion was about the PCB LED that replaces the lamp and lamp holder. He said they're designed to work in the older machines as-is, with no upgrades or board changes. He says they are fully rectified to work in AC or DC circuitry, have a resistor to help with the latching issue and a capacitor to smooth out the incoming voltage. As our discussion went on to the scoreboard, he said "I'd suggest using our anti-ghosting LEDs there, as they have the *exact same circuitry built in.* I even called him this morning to confirm the anti-ghosting LEDs I chose had all of those features, since I chose the economy ones. Their site only mentions anti-ghosting, but the engineer seemed to know his stuff. Time will tell the whole story and I will only be installing a couple of each. Even after that test, it'll probably be a few at a time. I don't have the patience to sit under there and remove/resolder 90 lights at one go.

Okay, fine. FINE! I wasn't going to toss it anyway, but now I'm convinced I should fix it. I'm sure I can easily get thermal paste and find the proper screw. What is this bakelite spacer and how/where would I get one? I assume its a pretty important component to get the tops of the rectifiers flat and aligned, so as to maximize thermal transfer to the heatsink.

Rikoshay Thank you. That seems like a really good idea. It doesn't seem like the height of the stand-offs is critical, so long as the space between the BR and board are adequate when the screw is fully unthreaded. It seems like you could screw down the rectifiers onto the plate and then solder them in place. As you say, no strain at all, still extremely serviceable and if you solder the BRs in last, everything is perfectly spaced.

I went ahead and ordered another board. When it gets here, I'm going to drop that one in place and that will give me time to rebuild the original properly. Was your rebuild recent? What did you use to clean the board before rebuild? Were you able to find the entire set of components as a kit?

Quench Just curious - is it safe to do the following?

1) Shut down my machine
2) Unplug one or more connectors from my lighting and/or aux lighting board (As in an entire "J" connector)
3) Power on the machine

Obviously the lights attached to those connectors will not function until I shut down the machine, reconnect and power back on... I don't see any chance of damaging the equipment here, unless I physically break something in the process... Would just be nice to also have an expert opinion.