Are the controlled lamps and the GI flickering, or just controlled lamps? You could have a problem with your lamp strobe lines coming off the MPU. Lamp address, data and strobe lines are connected to MPU connector J1.

Check out PDF page 40. Might also wanna check R78, R79, C49, C50.
https://www.ipdb.org/files/1755/Bally_1979_Paragon_English_Manual_color_with_paginated_schematics.pdf

You could also try swapping U10 and U11 (6820 PIAs) and see if the problem moves/changes.

Wires associated with lamps go from MPU J1 to lamp board J4. Might be worth it to try wiggling the connectors/wires while they're plugged in first, see if the problem changes or goes away, but if you wanna just repin them and knock it out as being a problem that works too.

Do you have the old lamp board you can try swapping back in?

You mentioned in your opening post that #47 (incandescent) globes are also flickering? Adding resistors to lamp sockets will make no difference if the problem also occurs with incandescent globes. The resistor load issue is related to LEDs only.
When is it happening? While the game is in attract mode or while you're actually playing?
You mention it's happening like a wave, can you video it?

Does your game have the original MPU board? If yes, please post a picture showing the chip at U14 - we need to see its part number.

What's the part number on U14? I cant read it in the pictures. Did the previous owner install all those sexy blue chip sockets?

Are the controlled lamps on the backbox scoreboard also doing the same thing or is it only playfield lamps?

Pinside doesn't directly support videos so you have to upload it on some other site. You could always upload a MP4 file to http://mediafire.com which doesn't require a user account.

The bulbs flickering and those that aren't should reveal a pattern. So we need to see/know which ones are doing it.

Unplug the U2 and U3 switch connectors and see if the lamp issue clears. When the slam or start switch leans on the grounded coin door the lamps will go wrong too.

U14 needs to be the 4049UB type or else the zero cross interrupt has problems which makes some lamps flicker.

First thing to try is this ^^^
Somebody has installed the wrong chip on that board (chip is dated 1994 so it's not original). You have a buffered 4049B chip and it needs to be an unbuffered version 4049UB, note the U in the part number.

Quench knows what he speaks of, I ran into this very problem with my Fireball Classic. Needs to be unbuffered 4049.

ebay.com link: itm

Does Jameco ship to you?
Manufacturer p/n: CD4049UBE
$0.49 ea
1,947 In Stock
https://www.jameco.com/z/CD4049UBE-Major-Brands-IC-CD4049-HEX-INVERTING-BUFFER-CONVERTER_13055.html

(maybe with shipping this wouldn't be cheaper, but I always find at least $20 worth of stuff I "need")

In the 6803 MPU they changed the zero cross detection to schmidt input chips probably because of this issue.

Texas Instruments has their own store. The MOQ might be a tube or a couple of tubes, but the price is right and you are buying right from manufacturer.

https://www.ti.com/store/ti/en/p/product/?p=CD4049UBE&keyMatch=CD4049UBE&tisearch=search-everything&usecase=OPN

a yt video that talks about slow transition times and zero crossing.

I think what also doesn't help is if you are referencing the part from most, if not all Bally Manuals from that era, they will state U14 as a 4049B version, which is incorrect - every Bally -17/35 MPU I have seen has actually had a factory 4049UB version installed.

I actually own a Fireball Classic that came from the factory with a 4049B. Toward the end of the -35 era parts were running out and not all the substitutions were 100% correct.

And yes I had to change it out due to flickering.

I think I read somewhere that 4049's were originally marked as 4049B and they were unbuffered. It wasn't until someone produced a buffered version that everyone changed the suffix. Don't quote me on this though.

I have a bunch of NOS National Semiconductor CD4049C dated 1979. They work perfectly fine in this application and also the Bally -32/-50 sound boards that uses an unpowered 4049 in a very unusual fashion in the mixer section.

CD4049CN.jpg

Application Note PDF from Texas Instruments regarding buffered vs unbuffered for those interested.

Below it the section mentioning buffered chip outputs potentially oscillating with slow rise inputs (i.e. our zero crossing input), along the lines of the youtube video barakandl posted.

Understanding_Buffered_and_Unbuffered_CD4xxxB_Series_scha004.pdf

CD4049B_Output_Oscillation.png

TC4049B and HEF4049B for sure do not like the slow transition times. Like quench said a few other 4049B have worked okay. Bally probably should have used 40106, but that may not have existed in 1977. Or 4049 was cheaper and it "worked".

That video linked even talks about the feedback resistor used. =D

The Bally -32/-50 sound board, If I make the board play a continuous tone, swap the 4049UB for a 4049B the pitch changes just slightly. Otherwise it works OK.

Yesterday when looking this up I came across a guitar forum that had some small discussion of 4049UB vs 4049B used in some guitar pedals where the unbuffered 4049 was required due to the FET gain being small enough to be biased in a linear fashion, whereas the buffered chips had too much gain and distorted. So I don't think Bally were unique in the way they used the 4049UB for analog mixing on their audio board.

https://www.diystompboxes.com/smfforum/index.php?PHPSESSID=jnnblpmeq4rqk6icv3ckik19b5&topic=116440.msg1078181#msg1078181

Phew man what a journey; great job!

Those 2 are absolute rockstars. They are both super knowledgeable and make Pinside a better place for sure. Congrats on resolving the problem!

Not to bump an old thread, but I just wanted to say thanks!

I just grabbed a FBC that was having this flicker issue, updated U14 to a CD4049UBE from the 4049B and it works flawlessly