Could you post a picture of the 7810 RAM?

It's indeed jumpered for two 2732 EPROMs.

Unless the windows had access to low level UV light I wouldn't be so sure they're erased.

Maybe the '7810' is the manufacture date code. 10th month of 1978. Like Tuukka said, post a picture.

OK! The 7810 is manufacture date, and 620-30 is Bally code for 6810 RAM. So you have the correct chip.

https://www.marcospecialties.com/pinball-parts/6810

U2 ROM "E-838-14" is Eight Ball Deluxe. What makes you think it isn't?
Per the Bally parts catalog:

ROM_EBD.jpg

As Quench table shows, the E838-14 is a 9332 4kx8 ROM and requires E13-E15 jumpering.

If I remember correctly, the 9332 is not exactly pin compatible with a same size 2732, but more like a 2532. That's why the common U2=U6=2732 jumpering does not match similar size 9332 jumpering.

Anyway, the table sure is correct for original Bally masked ROM's.

If both U2 and U6 are 9332 or 2532, then for some reason, according to https://techniek.flipperwinkel.nl/ballyss/rep/index3.htm#35roms the jumpers should be

4-12, 7-8, 10-11, 13a-14, 16a-34, 29-33, 31-32 and cut 13-15

The U2 ROM in your picture has "9332B" on top line. Most likely your U6 has similar marking.

Your E13-E15 jumper is in correct place, if it is needed. There are differences between the two jumper tables.

While this might work, the signal mixing to U2 is more convoluted. The jumpers listed above in the Bally parts catalog is logically a cleaner solution.
I actually prefer a slight variation to the Bally setting which allows U2 and U6 ROM types to be independent but I wont confuse matters here.

Maybe I should take on a weekend task to look up all the E-jumpers on schematic and make clear to myself what they actually do!

Edit: also document those of course.

Usually the last digits in part number tell the size in kilobits. So 2532, 2732 and 9332 are all 32 kilobits, or 4 kilobytes (4k x 8). 2716 would then be 16 kilobits or 2k x 8.

The jumper locations on the board use larger solder pads to make it easier for techs. The smaller pads are just locations where signal traces connect between the top side and bottom side of the board via copper in the holes.

The ROMs for the 84 version are no different to the earlier games.
U2 is an E838-15 which is specifically Eight Ball Deluxes "personality ROM" and U6 is a common operating system ROM used in a number of games whose code is E720-52
Personality and operating system ROMs are specific to each other, i.e. you cannot use a different Bally operating system ROM with an EBD personality ROM.

Might be E838-13. From memory this is the same as E838-14 but just simply a different ROM package (EPROM vs mask ROM)
I'm not sure if it's known what the difference between the E838-14 and E838-15 code is. One day I'll compare them.

Post a nice clear high res picture of it.

Here you go, I did it years ago - I would post it in plain text if Pinside didn't swallow spaces:

MPU-35_ROMs.png

Thanks! And to further clarify what is being done, the pin definitions:

pin 9316    2716    2732    2532    9332

18  CS2     /CE     /CE     A11     A11

19  A10     A10     A10     A10     A10

20  /CS1    /OE     /OE     PD/PGM  /CS1

21  /CS3    Vpp(+5) A11     Vpp(+5) CS2

22  A9      A9      A9      A9      A9

Wonder why the A9 pin is configurable, when it is the same in all types. A10 is fixed.
(Also wonder why the forum software makes backtick formatted code block doublespaced.)

A9 is configurable for compatibility with 512 byte ROMs that came from early -17 MPU board games.

part #

big giveaway it's an eprom is it has the window in the center so even if a label is on it you can usually feel the window through it (or pull the sticker back).

There is such a thing as prom's (one time programmable) - those you'd have to go off the part #.

You REALLY need to get eprom burning going Rich.... I know it's a seemingly rough learning curve, but I couldn't imagine at all not having a burner here for all the stuff I'm into. Hardest part at this point is sourcing the eproms, which I'm going to have to do again soon as all that vintage I have is dying quick.... couple bad bytes here and there. Of course, no one expected a 40 year old eprom chip to being used 100's of times for crazy software testing.

During my learning experiment... I found it's best to convert the board back to original stock config with ?2716? roms. Then and only Then; apply the fixes for larger roms. You never know what mistakes were made during the previous conversion and the manuals don't really give "final" jumper configs for the final configuration.

They likely used EPROMs at the start of game production and once code was proven to be good switched to cheaper mass production mask ROMs.

Yes and no. There are some jumpers that kind of configure both U2 and U6. That's why I mentioned above a jumper variation to 9332 chips without going into detail. Specifically if you change E13-E15 to E13-E36 it makes U2 and U6 independent.

Someone's tried gunshot repairs by blindly replacing logic chips. Shame on such a clean board. The problem now is finding errors in their work.
Swap the PIAs (U10, U11) for known good chips.

The board with the locked on LED, you need to determine if the valid power circuit is coming out of reset, the clock circuitry (U15, U16) is giving the CPU 'sense' and that the CPU is attempting to talk to the U6 ROM.

These are nice looking boards that deserve to live again.

It is worth attempting repair. You need at least a multimeter, but a logic probe would be a great help. You can make a simple probe with a high-efficiency LED (most today are) and a 1k resistor. Connect the led + side to +5 volts via 1k resistor and then probe the signals by connecting the led - side into signal. Led glows brightly when the signal is at 0, led is dark when the signal is at 1, and with pulsing signals you will get intermediate brightness. Try it out with a working board to get to know it.

(It is better to make the led light from 0 instead of 1, because the TTL chips output pull down to 0 much better than they pull up to 1. You do not want to make extra load to 1-level with the simple probe.)

I think the Bally boards are by far the easiest to repair on bench, due to simple architecture and well working diagnostics. First try to get the onboard led to turn off at boot. If your ROM jumpers and ROM itself are ok then it only needs a reset and clock signals on CPU (with the homemade probe you might see pin 40 of CPU U9 flash the probe at power on, then stay dark, and at pin 3 the probe should light about half brightness).

It is possible that the onboard led driving PIA U11 or the led circuitry is bad, but since your board looks so nice and no battery damage, it is not likely. But you could swap U10 and U11 to be sure.

Usually I like to jumper all my boards to use 2732 in U2 and U6, because that is the most available chip today. I recommend getting an Eprom programmer and maybe also a UV erasing lamp. A TL866CS programmer can be found under $20 from eBay, but beware that not all versions are able to program the old eproms requiring 21 or 25 volts.

I've been able to fix all my boards this way. An eprom burner and stack of 2732's are required if you want to fix more than 1 or 2 boards. If all the jumpers are right and there are no issues with the traces on the ROMs and RAMs and the ROMs are good 2732 burns, you should be able to get past the solid LED. Sometimes on the bench I have to short U9 (CPU) pins 39 and 40 with a screwdriver blade to get it past the solid LED. I think the bench power supply doesn't quiet down fast enough sometimes, but these boards always boot in-game.