I was having trouble getting waveforms not to be a mess (you thought that maybe my ground wasn't connected). At first I thought it was noise or something on the machine, but as I played around with the ground lead and got it better connected, the waveforms cleaned up. So my whole rabbit hole about noise may have been a red herring. Did I just mix metaphors? I'm starting the day perfect.

I do have a Bally solenoid driver board! It's not bulletproofed or refreshed in any way, but it works. I'll put it in today and take some more measurements.

In a conversation I had with Alltek, one of the suggestions was to isolate the LDB from the backplane ground, just to experiment, so at one point I had the Alltek screwed down and the Bally "floating" in front of it loose. Didn't make any difference to anything, but when I was chasing those ugly waveforms I thought maybe it had something to do with it...It didn't.

I would, but given I've swapped the MPU and LDB in various combinations, I feel like I'm chasing my tail. The only thing I could really try is the rectifier, but why would the rectifier getting warmer impact the voltage from the MPU to the LDB? Anyway, feel free to suggest heating up anything, I'll look like a fancy hair stylist and run around this thing with my son heating every element if that gets us to the solution.

I was talking about the LDB. Maybe I've got my connectors wrong...Top left and bottom left, the only connectors on the LDB we need to perform the test. The lower left brings power, the upper left to the MPU (maybe that's J4).

Yes, tight as can be. I've also checked and re-checked resistance and continuity between all these ground leads. That being said that is the opposite of a dumb question, because at this point, the problem is something that is going to be completely dumb and silly and we just need to find it.

Maybe something else in these photos will give you a clue:

VH1A6214.JPG

VH1A6216.JPG

VH1A6217.JPG

While we're on that topic, let me present some photos of the machine, outside and in. You might call this the big reveal. Feel free to pick apart the photos if you find anything strange or have any questions, maybe it will also lead to a discovery in our lighting ghost!

VH1A6186.JPG

VH1A6191.JPG

VH1A6193.JPG

VH1A6198.JPG

VH1A6205.JPG

VH1A6211.JPG

VH1A6212.JPG

VH1A6215.JPG

VH1A6219.JPG

VH1A6220.JPG

VH1A6222.JPG

VH1A6223.JPG

VH1A6224.JPG

VH1A6225.JPG

VH1A6226.JPG

VH1A6227.JPG

VH1A6239.JPG

Thanks, but my son deserves a lot of credit as well. We worked hard.

It is the lug, which has a blue strip on the crimped area. Basically, anywhere the ground braid was screwed into something, or originally had a yellow plastic covered ground strap, we replaced it with a lug and braid. What's great about this is that the braid doesn't disintegrate over time and it's very serviceable. It's not isolated from the cage, as it's connected to the metal slide base for the lamp panel door, which is also connected to the cage.

IMG_3232.JPG

Not surprisingly, I plugged in the SDB and it didn't work. I'm doing some early tests to see if it's just a fuse or a solder joint. I won't spend too much time on it but I want to see if I can get it working.

Wow, thanks for the complement. I think High_End_Pins could bang this out in 2 weeks. It took us 10 months. That's experience for you!

I got the old SDB working. Unfortunately there are some bad traces and solder joints... and while I want to fix it, I just got it working enough to know I'm getting the identical problem as with the new SDB. We now know the blinking lamps are not caused by anything specific to the SDB either.

I moved the machine to a new location, with power I trust more and better lighting for working. We're calling the restoration COMPLETE, but we're going to keep working on this blinking light mystery here in this thread.

As with all of these restorations, it couldn't happen without all your support along the way, so thank you so much from both of us. There were times when we literally thought we couldn't make it this time (automotive paints, omg) but your encouragement kept us going.

Thanks to this amazing community! I probably forgot a few, but a quick shout out to a few of you amazing peeps:

MustangPaul Arcane vid1900 Atari_Daze Bryan_Kelly High_End_Pins Quench barakandl Inkochnito mof JethroP @jjsmoth pintechev radium g94 cottonm4 Xenon75 BJM-Maxx jibmums Da-Shaker rack-em-up Pinzap Elicash Grnrzr chill lb1 steamfitter Joey_N pinkid the4horse pinballinreno Tommy-dog Tsskinne Madmax541 legtod2 RoyGBev Groucho hawkmoon LynnInDenver NimblePin Squeakman Mk1Mod0 Redketchup wm6929 heni1977 wolffcub Skins

You folks are the best.

Now help me find this source of lamp flashing. LOL!

Right now I'm going to quietly put the rotisserie away until I've negotiated more space in the house for pinball machines.

Quench and barakandl ... what do you make of this.

Alltek diagnostic board connected, same wiring/setup, either Alltek or Bally LDB, I can light individual lamps or entire SCRs with no blinking, one by one. That's the diagnostic board telling the MPU to light a lamp. MPU tells the LDB/SCR to turn on. No problem.

Game boots up normal code, no diagnostic board, Q14 gets those intermittent wave forms on the gate leg like this:

Screen Shot 2018-10-06 at 12.45.33 PM.png

I have not re-terminated the power going into the LDB. Before I do, is there a way to measure that and check it to see if there is anything weird arriving there?

If not that, how can this be anything but a game code bug of some kind, related to interrupts or something? Speculation, but if so, maybe something invisible up until now for two reasons:

1) No one restores/messes with a Future Spa. It's like the kid who gets picked last for dodge ball. Maybe I'm the first to see this?

2) In most circumstances, these games have incandescents, and maybe when there is enough draw (the playfield connected, the lamps all 44s) it masks the intermittent off state of these phantom SCRs.

Is that possible? Just throwing it out there.

I'll give it a shot. Yes, I have an EPROM programmer...and probably some blank ROMs as well. Note I'm not switching ROMs between boards, the Alltek has all the ROMs in a single chip.

Ok new information. Remember, we found these bad waveforms with no lamps of any kind attached. So take this all with a grain of salt.

Before I began this next test with a heat gun, I made sure all the lamps that have EVER blinked inappropriately were replaced with 44s. Just because. As you can see below, I never got to a heat gun.

With no warm up period, immediately after powering up, all the lower FUTURE SPA and upper UTURE rollovers lamps now behaved appropriately. Those lamps now never blink in any weird way. Fascinating. All we did was increase amp draw going to 44s from 47s.

Also with no warm up period, the upper rollover F and the center target SPA all blink inappropriately during attract, lamp test, and game play.

After a short period (let's call it 3 minutes), ALL the lamps begin to behave appropriately during attract and lamp test... which is great... but...

During GAME PLAY, the upper rollover F and the center target SPA all blink inappropriately.

What do all of those have in common?

*They are all address 0 on the four decoder chips.*

My goal is to get LEDs back in here and working eventually, so the question is: Why would going to 44s change the behavior? Also, what would effect only address 0 on the decoder chips?

I've also verified that other Future Spas out there do not have this problem, so it's not going to be in the code.

That's the plan, until this is solved. Note, the majority of sockets didn't flicker, and those higher addressed lamps still have LEDs. I can switch everything out to incandescents but they seem to be operating independent of these tests pretty well without any impact to testing so far. I'm happy to swap more out though.

Yes, but in this case, the incandescents flicker in game mode, and don't in lamp test/attract! Different behavior than you're mentioning there.

I'm a network engineer originally. My brain tends to think about signal clarity and things I've seen in the past.

I've definitely seen situations where data being sent worked fine when 1's are sent, but failed when 0's are sent. Or maybe a better way to put it: If you put payload on a line, the payload arrives fine. If you remove payload, and send nothing, noise isn't pushed out by the payload, so you get noise. It makes me wonder if something like that is happening here...Perhaps I have noise, which is drowned out most of the time, but when competing with a 0 addressing, sometimes corrupts that addressing.

Pure speculation. However, it would explain why other games work and mine does not, because I've somehow introduced noise.

You are correct, that was the behavior before. At the time I wrote the above post, the behavior was different. It's also colder where the machine is now, which may have an impact on this.

Essentially, before the 44's, more lamps showed flicker, always. After I initially installed the 44's, attract and lamp test were showing no flicker (after warm up) for FUTURE and SPA center target. Game play was showing flicker for F rollover and SPA center target lamps.

Now we're back to before warm up, SPA center target and FUTURE (upper rollover) all flickering, always. After warm up, SPA center target and F (upper rollover) flickering, always. It may be related to the temperature, though.

@zitt thank you for the excellent information. I have a couple questions.

@zitt can you remind me how to best check the resistor values with a multimeter and what they should measure? I'm a complete amateur with this tool and I want to make sure I get it right.

This particular direction intrigues me because I did have to do my own soldering on the back of the rectifier board and this could be where I created a problem.

Could this be possible since I've swapped both the MPU board and LDB board without any change? I think that pretty much proves it's not something ON one of those boards.

Are you talking about the J1 connector on the MPU? I can check. Would a continuity test reveal a bridge between those lines?

I think swapping LDBs without results also would steer us away from this, right? I've been getting the same result with original Bally board and a brand new Alltek.

I've reterminated and recrimped each of the four address lines and the strobe line between the MPU and the LDB, without a change. I was worried there may be attenuation or a short between those wires in the harness, so I've also tried creating a dedicated AD0 line and a dedicated strobe line between the two boards with a separate wire, without a change. Should I bother with verifying the resistors on the LDB and MPU given the board swap yielded no results?

Same question as above. Thanks @zitt !!

Thank you! In all seriousness, this game is really fun and a fantastic commentary on the state of things in 1979. One of the weirdest years in America to revisit, culturally. In all its glory, this game really is amazing to look at.

Let me know if you run across any issues in the process, I'd love to follow along!

I love this, keep throwing stuff out! Since we were able to reproduce the flicker with only a single lamp, one lead connected to SCR Q14 and the other to TP1 on the rectifier (your repro board) during lamp test, I figured it must be voltage related but not specific to draw on the playfield lamps. Is that a false assumption?

Right now, I'm operating with the Alltek board with 6.3vdc rail connected. That adds a 470ohm resistor to every control lamp, by default, whether it's LED or not (and now the lamps in question are 44s). Wouldn't that be enough to make the SCR latch?

The board is new, the connector is new... I doubt it's had more than 15 cycles. I can repin the board/connector but I feel like the above point (flicker with a direct socket attachment to Q14) might have us looking elsewhere?

If I had to do it again, I'd seriously consider borrowing someone else's paint booth. Temporary booths are a total mess! I have a post above about my lessons with painting here:

https://pinside.com/pinball/forum/topic/future-spa-father-and-sons-second-restoration/page/4#post-4494531

The main reason I went this route of automotive paints (besides the AMAZINGLY cool feel to it when it's done) was because back when we did our BoP restore, I was told over and over again that the colors I wanted didn't exist in traditional paints, and I should go for automotive. The purple we got for BoP was a royal PITA to produce, and professionals all out refused...we got Home Depot to make it in latex and it got really close.

In contrast, when I showed the original pink chips from under the coin door frame and lock down bar to pros (good pros), they said that they recognized the pink and it was based on a DuPont color that had a neon-ish component that was pretty much banned after the 70's. To recreate it would be impossible with photographic paint matching in general, but that I could get close with automotive.

I detailed my experiences up in the thread, but suffice to say I got REALLY close with swatches. The automotive guys laughed and laughed at this because it's a color pretty much reserved for "pimp my ride" type or the dragster world. It's also pricey.

The secondary reason I used automotive was because I wanted to learn how to shoot clear. I think with an amateur hand and amateur booth, as long as you spray on a horizontal/flat surface (next time I'll put the cab on a rotisserie), it can be great. With a pro booth and some experience, spraying sideways isn't that bad.

Anyway, so then I have these automotive paints and jumping into the @high_end_pins type of automotive process with no experience was incredibly fun and challenging. I'll think twice before jumping into that process again in my own garage!

Don't leave out Buck Rodgers in the 25th Century. Come on.

Here is a good photo of the rectifier board. I assume you mean testing the ceramic resistors? (How exactly should I test them?)

072418091133.jpg

Also, isn't the problem that the SCR won't turn ON?

For now, I'll measure in circuit, because the alternative may introduce more problems if I'm not careful. I'll do it if we have no other path to pursue. Meanwhile, the left one measures 27.3 ohms, the right one 600 ohms, I think.

Agreed. This is where I'm focusing right now.

The flickering occurs only when the lamp should be energized. When it's off, it's off.

That's right. We've been able to verify that if we "hard wire" the SCR to ground, the light stays on solid. When we rely on the MPU to turn it on, it doesn't stay latched and pops on and off. The cables between the MPU and LDB have been re-connectored and re-terminated. The MPU and LDB have been swapped. My gut tells me it's related to power somehow. I did find a loose connector on rectifier J3 pin 4 heading to LDB J4 pin 1. It was not secure in the housing. I got excited for a minute, but it made no difference (and I was testing solid voltage to the LDB anyway). Maybe I need to look closer at a specific pin/wire?

Yes, 100% confirmed. Also, for LEDs, I have the Alltek hooked into the 6.3vdc rail and they are 100% stable. However, for troubleshooting, they've all been replaced with 44s, with and without the Alltek and the extra resistors.

No, I get it. Believe me, that was my first thought, so we eliminated it from the testing. If you want to see the extent with which we've been troubleshooting this, please visit this thread here:

https://pinside.com/pinball/forum/topic/bally-as-2518-club/page/25#post-4613339

Ok, here's a thought. Earlier, I put new connectors on the entire harness. For the rectifier board, I used high current crimped ends for the higher current pins (regular molex for the remainder) and a GPE larger style block connector.

Maybe I should re-terminate all these connectors and stick to Molex and normal .156 trifurcons all around...and while I'm at it be obsessive and see if I can find something goofy, like a bad crimp. What do you think?

At risk of exposing my embarrassingly bad soldering skills, anyone see any obvious problems here? All the test points test fine, but remember we're looking for a ghost. Something that would effect voltage heading to the LDB SCRs.

IMG_3258.JPG

IMG_3252.JPG

IMG_3257.JPG

Yeah, I think I need to improve the rear solder joints of the board. I'll hit them today and make sure they are perfect.

Also, the original rectifier had a 20amp fuse for the feature lamp bus. This one has a 10amp fuse (but the intention is to go LED eventually anyway). Could that have anything to do with this latching issue?

I hear you, but I get the same result with both the original and Alltek lamp board (and same for MPUs). This tells me somehow I’m responsible for this. I just can figure out where. If it was components, what are the chances it would be identical on two lamp boards?

Beyond just voltmeter tests of the rectifier board, are there any ways to test the integrity of the power and that it meets the specifications? Waveforms?

With incandescents, Q14 lamps did not flicker with either lamp board before teardown.

Since reassembly, with the only components changed being the connector housings, trifurcons and the rectifier, I now get Q14 flicker.

This has to be a mistake I made.