Alright this is an odd one. I'm in the home stretch of getting my BK2K up and running and this is one of the last gremlins.

The problem:
While the game is on and in attract mode, the magnet under the playfield does what I'd describe as "pulse erratically". I can hear a "thud" type of noise every time it's energized and there doesn't appear to be a rhythm or regularity to it. I've placed a ball over it while it's doing this and the magnet kind of limply slings the ball around and sometimes holds onto it. Sometimes it goes away, sometimes it happens for a while. The most current iteration it sounds like a heartbeat. It also seems to sometimes coincide with the GI relay being on or off. Here's a video of the sound which also shows the weird relation to the GI relay:

During a game, hitting the Magna-save switch registers with the game and the correct sounds, lights, and display appear, but the magnet doesn't capture the ball, it works similarly to how it's working during attract mode i.e. limply slinging the ball around, kinda-sorta capturing it.

Things I've done to diagnose/fix:
- On a whim I replaced the TIP 102 at Q14
- Aux board header connectors have been reflowed and cleaned
- Aux board caps have all been replaced
- Aux board fuses have all been checked and verified to be the correct rating and have continuity
- Cleaned and filed the leaf switch contact for the magna-save button
- Verified I have voltage at the High Current Driver under the playfield
- Reflowed solder on the connectors and TIP36C of the High Current Driver

When I pulled the High Current Driver board from the playfield I ran a diode test on the TIP36C (Using this as a guide: https://vetco.net/blog/test-a-transistor-with-a-multimeter/2017-05-04-12-25-37-07) and got some results that seem to indicate this transistor could be dead. I wasn't sure if the diode and resistor on this board could interfere with that test and potentially give me bad results.

My current suspicions are either the TIP36C being dead or a short to ground. The intermittency of it is making it hard to diagnose. I guess my question is, would a dead or failing TIP36C exhibit these symptoms or am I barking up the wrong tree? Could the magnet itself be flaky? I haven't run an ohm test on that. What reading should I expect from a "good" magnet? I've attached some pictures of the High Current Driver (these are from before reflowing connectors). I can also add updated board pictures if needed but some out of date ones can be viewed in this thread: https://pinside.com/pinball/forum/topic/black-knight-2000-won-t-boot#post-5810248 ( The U21 and U22 sound roms have already been swapped to the correct socket )

IMG_7905 (resized).png

IMG_7906 (resized).png

I did some more poking around this evening and I'm still not coming up with anything.

I replaced the TIP36 just as a shot in the dark with no dice, but it did stop pulsing in attract mode so either I fixed something or broke it further.

The thing that's confusing is, the manual states the brown-violet control wire at the High Current Driver board should pass through the aux board at 5J6-2 to 5J2-2 and connect to the CPU board at 1J12-8. My setup has the brown-violet wire going from 1J12-8 down to the J1 connector on the interconnect board. I'm not sure where it goes after that but I have verified continuity between the 1J12 connector and the J1 connector at the High Current Driver board which I think makes the discrepancy between the machine and the schematic irrelevant since it's at least connecting like it should be.

As far as voltages go I see ~50vdc at the violet-yellow, ~50vdc at the brown-violet wire, and the black wire shows something like -0.1vdc at J1 of the High Current Driver. Disconnected I still see ~50vdc on the violet-yellow wire and ~30vdc on the brown-violet wire. I'm curious how ~50vdc gets onto the brown-violet wire. That seems wrong and like a lot of voltage to be going through the CPU board but I am admittedly a novice and perhaps I'm missing something on the schematic. Also I'd like to mention that I measured with several different sources of ground: the ground strap in the backbox, the ground strap in the cabinet, and the ground pin on the CPU with no difference in measurement.

As another data point, in case its relevant, the left sling was locking on when I turned the machine on. I haven't debugged it yet, but I did desolder the wires from the coil, wrapping the two violet-yellow wires with a wire nut, and electrical tape on the control wire so I didn't burn it out. I mention this only because the sling shots are listed in the same group of "special solenoids" on page 95 of the BK2k manual pdf and share the violet-yellow 50v line.

When I run the diagnostics solenoid test and it hits solenoid 15 I don't see any movement on the brown-violet wire. I'd expect this to flip-flop between high and low. In addition to the swap out of the Q14 transistor tonight I swapped out the pre-driver at Q10 with no change.

Has anyone run into anything like this? I'm kind of lost at this point. I suppose I could trace the control signal further up the chain like the 7408 at u17 or 6821 at U10 but I don't have any other solenoid problems related to those chips.

EDIT: After posting this and doing some more head scratching, I pulled the schematic back out for the CPU. It looks like the Q10 pre-driver is controlled by U17 pin 8 which is flipped on by PB6 of U10. I notice on the schematic that in addition to solenoid drives, U10 is also in charge of sound selects. This is interesting because I've had some weird sound issues where sounds 04 and 05 are missing in the diagnostic test, but eventually fire after the machine has been on for 10 minutes or so. Is it possible that U10 is my issue? I'll poke around with my logic probe tomorrow to see what's going on.

Following my theory that U10 is potentially a problem I put the game into the solenoid test and locked it to solenoid 15, Magna-save. I definitely see pin 16 of U10 stuck low when it should be pulsing. Just to be sure I unplugged all the connectors from the CPU board except power at 1J7 and plugged in the Leon test ROM. When I do this I DO see pulsing on pin 16 so I know that 6821 is CAPABLE of running that output so I don't think U10 is my issue.

Just to rule out other things I ran a diode test on U17 and nothing was out of spec and I also ohmed out the resistor array at SR-4 and that too appears to be in spec at ~550 ohms.

At this point I can only guess that something is screwy with the addressing, but in a previous thread I verified all the address and data lines as part of getting the CPU to boot so I know those are good. Anyhow it would be odd that this ONE address would be written wrong since it's the only coil not firing in the test.

Given all that I'm down to two theories: 1. Perhaps my game rom version is incompatible with my setup; I'm running the "Revision L-4 Game ROM" found on IPDB. 2. There actually IS something wrong with U10 despite the Leon ROM appearing to show it working. I have one PIA socketed already with a known good 6821 and I have a few spares as well. I could socket the U10 and try out those but it just seems pretty invasive to do on a guess.

Good point, @frunch! I should pull the board and double check for solder whiskers. Those have bitten me in the butt before.

Unfortunately I'm not sure of the prior state of the magna-save as the board wasn't booting when I got the game. It's not outside the realm of possibility that I messed up something when addressing the other issues. I'll check over the CPU board later today.

I appreciate the reminder! Always check prior work. It's bitten me before, for sure.

Last night I went over the CPU board with my flashlight and magnifying glass and didn't find anything I'd consider a 'smoking gun'. I brushed off the back and blew it with some compressed air and put it back in the machine. Since I also needed to replace the opto chips on the interconnect board I took that out and reflowed the header pins where the controlling wire for the Magna Save passes through.

Now, I couldn't tell you what possessed me to do it, but at some point I just decided to push down on U10 (with the machine on in attract mode) and doing so made the magnet start pulsing again; letting go made it stop. I repeated this a few times with the same result so I just decided to go ahead and socket U10 as it seemed like SOMETHING screwy was going on. Alas, doing so made no change aside from making it so pushing down on U10 no longer activates the magnet. I even tried three different 6821s, two of which I know are good.

Today after some sleep I theorized that perhaps the 7408 at U17 was causing problems. The first thing I did was attach a lead to the ground strap in the back box and briefly touch the tab on Q14. This absolutely fired the magnet so now I at least know that Q14 down to the magnet works. To test further up the chain I pulled U10 from the board, bent up pin 16, and re-socketed it. I put the game into solenoid test and stopped it on solenoid 15. Using my logic probe on pin 16 of U10 I STILL don't see it flip flopping high and low.

After all that, I guess I still can't rule out the 7408, I plan to do that by applying 5v to pin 9 of U17 to see if that will also trigger the magnet. If the 7408 turns out to be bad, I have some LS08s I could stick in it's place, but I'm not sure that's a good replacement. However if applying the 5v directly works, I have to move even further up the chain but I'm not sure where. The CPU? The game ROM?

Random, whacko, out-of-left field theories:

1. Maybe kinda likely: Maybe I just need to try and re-burn U27? I had to re-burn U26 as it wasn't passing verification in my eprom programmer but I don't recall checking U27. I assume it's mostly working since the game boots and IS playable, albeit without a working magna save.

2. Also maybe kinda likely: Since I still seem to be having weird sound problems where sounds 04 and 05 go missing and come back after a while I wonder if U10's neighbor, U9 is somehow interacting poorly with U10. The two are connected via U10 pins 2-9 and U9 pins 2-9 but again, feels like a reach.

3. Pretty unlikely, but maybe...?: Doing some internet digging I found a couple of cases where BK2k's with the earlier 'future' plastics, which my machine has, are running prototype ROMs, in some cases a Proto 7 and others a Proto 5. I wonder if maybe my machine would work correctly with the Proto 5 ROMs from IPDB installed? Now, the only reason I consider this is because in the manual (pdf page 95) the Magna Save is listed as being driven by Q79 even though the solenoid table on pdf page 33 lists Q14 for the Magna Save. I'm curious if, for whatever reason, the code I have is instead flipping Q79 on and off when solenoid 15 is being triggered, but I need code that addresses Q14. I plan on testing Q79 with my probe later but this whole line of thinking is a huuuuuuuge reach.

Given that EVERYTHING else works just fine aside from those two sound slots in the sound test occasionally disappearing (but coming back after the game warms up) and the magna save not working, I find it hard to believe that there's a bad component somewhere, but as I often admit I am no expert in electronics.

Anyway, if anyone with a BK2K has some spare time, I would be super grateful to hear how your Magna Save is wired and what ROMs you run.

Action items for me unless someone chimes in:

- See if I can trigger the magnet with 5v applied at pin 9 of U17
- Check over U9 and look for shorts between U10 and U9 on PA0-7
- Re-burn U27 with L-4 code
- Maybe burn U26 and U27 with the Proto code
- Probe around Q79 to see if the code is triggering this instead of Q14
- If all else fails, I can go back over my address and data lines.

Quick update:

- Applying 5v on pin 9 of U17 and pin 16 of U10 activates the magnet so the chain is good up to the PIA at U10. I have U41 socketed so I swapped those two and there was no change during the solenoid test.
- I re-burned U27 with the L-4 revision with no change.
- I didn't find any shorts between U9 and U10
- The control pin at 1J18-9 and the output at 1J19-9 are both stuck high all the time so my Q79 theory was whacko as I suspected.

At this point all I have left is possibly trying the prototype ROMs and going back over my address and data bus. I don't have a lot of hope I'll find anything there. I'll also go back over the board to quadruple check I didn't leave something undone or shorted.

I AM CHAMPION!

I fixed it! Using two lessons that DumbAss taught me, checking prior work and tracing my data and address buses, I figured out that I didn't have continuity on D6 from pin 27 at U10 to pin 8 at U16. I traced the breakdown to the through-hole under the RAM at U25 but the actual break was right at the pad at pin 18 of the RAM; I just baaaaarely lifted and broke the trace when I replaced the socket at U25.

Anyhow, I folded the trace back into place, put a tiny dab of solder onto it, and verified continuity on D6 of U10. Plugged it all back together, put the machine into the solenoid test and totally got the magnet to fire on solenoid 15. Then I started a game, held the ball over the magna save target and hit the button annnnnnd WE HAVE MAGNETIZATION!

Thanks goes to frunch for jumping in and reminding me to go back to basics and CoreyStup for reaching out in the PMs and offering assistance.

Now I will go play my game and try to stop lighting up the alphanumeric sub-forum for a while

I know that feeling of Christmas when parts for things arrive! Your previous guidance ended up helping in this case, it just took me a bit to use the knowledge. I'm getting there.

Now I'm just tracking down flaky lights and switches which I think I can handle.