One set of 4 are connected to one side of the 6.3VAC transformer winding, the other set of 4 are to the other side -- once they go through the relay on the power board, which allows for computer control flashing of all GI.

So they're meant to be all together, and even better if you form a perfect solder bridge across all 4 pins when soldering for best conductivity to these high amp connections.

Good thing you went with the friction lock AND square post headers. I've seen a lot of these replaced with thin round header pins that barely give any contact points to the connector.

Be sure you buy a set of 8 Trifurcon pins (GPE has em) when making a new connector -- especially if you are planning to keep incandescent bulbs. The trifurcon touches 3 of the 4 sides of those posts ensuring highest conductivity which will keep them from degrading like the original ones.

GPE has the housings and pins. Total cost will be under $2.00 to replace it. Much easier pushing the pins into a new clean connector than trying to pull old ones out of a used one, and as noted above, a Panduit IDC won't accept a molex pin.

IDC stands for Insulation Displacement Connector. The pins are permanent inside the connector and pushing the wire, not stripped, forces the back of the pins to cut the insulation and make contact with the wire. Quick and easy and why they used them. Molex with regular or trifurcon pins require wire stripping, proper crimping, then a quick slide into the connector from the back to lock them.

Molex is just one brand of connector. Amp is another common one (more frequently used in video games). Panduit is the common IDC connector brand. There are different "series" and different pin sizes and shapes. There's a lot of variation out there, but luckily with pinball, most of it is .156" spacing. but older Bally's used .100 spacing. Get into video games and there are untold variations to deal with there.

For your display issues, I believe your glass is in good shape on the left one. It appears to be a shorted driver chip causing the additional 7 segments (and at least two of the basic 7 segments) to be turned on all the time, albiet at a slightly dimmed level.

Definitely swap P1 with P2 and see what you get. If P2 is still dead it likely is the glass--but do a test against a plasma ball to be sure. Based on the P1 results, the problem will either be located in the P1 area or P1+P2 area.

I believe you can safely swap the alphas into the numerics because the pinouts are identical on the driver board to the sub board, only the first 7 segments will obviously be driven however. But -- doublecheck with someone smarter than I to ensure that they haven't done something bad like tie the extra pins to GND/+5 to the glass that would cause a problem.

Given that you have 2 missing and 1 (likely) bad glass, it probably would be best to just get the LED replacements, and then sell off your two existing functional ones, it will cost less than trying to get 3 guaranteed working ones that will already be used in most cases. (PS: I may be interested in the credit/match if you do decide to sell-mine has pox and not fully lit segments). Also, you will need to still fix any problems on the driver board, even if you get LEDs.

Remember, High Speed has a completely different, much more basic sound board than all the other Sys11 games. I will double check on mine tonight, but I believe it does NOT make a sound when powered on, other than perhaps the coin door lockouts engaging (if it even has them, or is that on my Bally Vector?). My F14 makes a nice "ding", and my Jokerz (a much more complex sound board) makes a weird "bzzring" sound on power up.

Plasma ball test:
Video: http://www.youtube.com/embed/l8Eafl-FHAI
Thread: http://pinside.com/pinball/forum/topic/testing-displays-with-a-plasma-ball

(Never do this with displays connected to a board with active circuitry on it like later Sys11 boards, otherwise you will destroy the ICs with static--only safe to do on just the glass or glass connected to a pass through PCB (no active components) like High Speed has.

You won't fry the new LEDs -- especially if you pull the fuses for the 100V lines on the power board. Then you're left with just +5V on the display driver board. But yeah, to save sanity, try to get them working with Plasmas first, then the rest is drop in and go.

I just powered mine up -- I get a knocker clap and the coin door lockout coils engaging -- no beeps boops or bzz's of any type from the speakers.

For those of you who haven't done the safety fuse mod for the BR's, here is what mine looks like. No guarantee your BRs are oriented in the same direction but this should help anyone needing to do the mod:

HSBRFuseMod.jpg

Correct -- the display with white speckles in it definitely is more than outgassed -- it experienced a catastrophic seal failure (epoxy separated) and is dead. I have seen that in most every dead display. Don't toss the thing, someone can use the board at least to solder on a new glass if they have one. Also be sure to carefully peel (using a sharp utility knife blade) the black surround off the front so you can transplant it to a new one if needed. Don't heat it (unless perhaps on low from a blow drier) or it will stretch, don't pull it off at an angle or it gets all rippled and curved and won't reseat on a new one.

Good note on the phantom segments not being consistent -- more indication it is not a glass problem but rather a driver problem (best guess though theoretically it could be either but a lot less likely). Also note your dot's didn't replicate the same either. Next thing to do is to see if you can figure out which primary numeric segments might be triggering the secondary alpha ones if that leads you anywhere. (e.g. if the top flat one also triggers the upper right angled one consistently, figure out which traces are involved). Get into test mode and see if you can get different text appearing and see if anything can help lead you to a specific failure source.

Regarding low voltage: Extra ghosting would not show up there. If you have good displays you will barely notice a drop in brightness.. basically equivelent to a good but used display next to a brand new glass. However as they lose more gas or material on the segments, lower voltage does cause the segments not to fully light up, sometimes only half of it lights but tends to 'grow' as they warm up. (This is what happened to my credit/match one once I dropped from 100V to 91V)

Regarding cables. Almost every time someone / some guide suggests cables or caps as a problem for an issue I have it turns out to be not at all related. At least with cables, you have a total of 5 identical ones in your backbox. There is no need to buy new cables, there is no way all 5 would be identically bad. Swap them around (ensuring you align the red pin 1 stripe). Borrowing another display or taking yours to another Sys9 or early Sys11 machine would also narrow it down correctly.

HS: Top 2 are 7 digit alpha's, bottom 2 are 7 digit numerics, score/match is a 4 segment numeric.

That is an anti-rust coating, and original. It's yellow zinc aka zinc chromate: http://en.wikipedia.org/wiki/Zinc_chromate . Most other parts are in a lighter color zinc coating (technically plating) instead.