(Topic ID: 251495)

Black Knight: PIA issues (was: MPU Lockup)


By kevinclark

33 days ago



Topic Stats

  • 80 posts
  • 9 Pinsiders participating
  • Latest reply 17 hours ago by pincoder
  • Topic is favorited by 1 Pinsider

You

Linked Games

Topic Gallery

There have been 6 images uploaded to this topic. (View topic image gallery).

test (resized).png
0FBBD37B-A2C8-4945-B7B6-7B9B2478A65E.jpeg
8344D897-0AFC-4A74-B6EE-972235C5ED56.jpeg
8C8AD6B1-1FFB-49B7-ABED-4989914AD417.jpeg
7D2256FC-750C-4476-8BC6-60088FDF4915.jpeg
IMG_6273 (resized).jpeg

There are 80 posts in this topic. You are on page 2 of 2.
#51 9 days ago

Here’s pictures of the boards

7D2256FC-750C-4476-8BC6-60088FDF4915.jpeg8C8AD6B1-1FFB-49B7-ABED-4989914AD417.jpeg8344D897-0AFC-4A74-B6EE-972235C5ED56.jpeg0FBBD37B-A2C8-4945-B7B6-7B9B2478A65E.jpeg
#52 9 days ago

Whoops…..the FET's have been installed wrong (reversed!)....turn them 180 degrees (if they are still alive). Besides this, using 0 Ohm resistors to replace the 27 Ohm base resistors is dangerous. Replace them by 1k resistors. This prevents the 6427 darlingtons from exploding when things go wrong with the FETs. (It is better to apply the modification with the voltage diver here).

Also….this driverboard comes from an older system 3 game as there are 1k series resistors installed driving the switch matrix. For system 7 games like Black Night, these should be replaced by 0 (zero) Ohm resistors.

#53 9 days ago
Quoted from MarAlb:

Whoops…..the FET's have been installed wrong (reversed!)....turn them 180 degrees (if they are still alive).

@maralb Crap. I was sure I double checked that, but my before picture agrees. Thanks! How do I test I’d they’re still alive?

Quoted from MarAlb:

Besides this, using 0 Ohm resistors to replace the 27 Ohm base resistors is dangerous. Replace them by 1k resistors. This prevents the 6427 darlingtons from exploding when things go wrong with the FETs. (It is better to apply the modification with the voltage diver here).

This is surprising to me - I thought I saw jumper *wires* on the version in Vid’s guide. Did I misunderstand something?

Quoted from MarAlb:

Also….this driverboard comes from an older system 3 game as there are 1k series resistors installed driving the switch matrix. For system 7 games like Black Night, these should be replaced by 0 (zero) Ohm resistors.

Got it, thanks!

#54 9 days ago
Quoted from kevinclark:

At this point I've used so many of your ROMs that I'm considering swapping out IC 17's socket with a ZIF so I stop mangling pins. I'm really ready for that adapter board to exist so I can put it all on one chip.

I'm still working on that board, and it'll be much more than just a simple board. Stay tuned. In the mean time, a ZIF wouldn't hurt

#55 8 days ago
Quoted from kevinclark:

maralb Crap. I was sure I double checked that, but my before picture agrees. Thanks! How do I test I’d they’re still alive?

maralb Sorry, I've RTFM and I'm testing my mosfets. Look mostly fine.

#56 8 days ago
Quoted from pincoder:

I'm still working on that board, and it'll be much more than just a simple board. Stay tuned. In the mean time, a ZIF wouldn't hurt

Can't wait for your board to come out.

#57 8 days ago

NB: If anyone else is following along and wants to try it, the ZIF sockets I got *do not fit the vias on the IC*. I'm stuck plugging them into a breadboard (which isn't making great contact all the time) and running wires, thankfully still connected so it's basically a cable with a crappy segmented connector, over to the pair of SIPs I installed to replace the RN socket that was in there. It's hacky, but now that the connection is solid I can swap ROMs trivially.

#58 6 days ago

Just some updates:

The breadboard + ZIF setup is actually working pretty ok. I revisited Leon's ROM, suspecting that I maybe had some other PIA issues and realized IC36's outputs weren't responsive. I pulled it out and put in a new socket and now everything but pin 9 (PA 7) is flashing on and off like the rest of the PAs on IC 18 and 36. Pin 9 is about 0V and when I connect it to another pin that's pulsing the new pin also goes to 0V.

Leon says (http://www.pinwiki.com/wiki/index.php?title=Leon_Borre_WMS_System_7_Repair):

If there is somewhere an output on chip IC18 / IC36 which does not go up and down, then connect it to the one next it it. If this also falls away, then you have a short on that output, trace it first, if both outputs go up and down, then you can be sure the chip is broken.

I swapped out the chip with another one I believe works and they both show pin 9 at 0V while the rest of the PA/PB outputs dance. But I don't see the short. There's no solder bridge afaict. I assume the actual input comes from the data bus, and D0-7 appear to pulse on IC36 as much as they do on IC18. I assume I'd need an oscilloscope to compare inputs beyond that.

Any tips for tracking down a potential short?

#59 6 days ago

Uhh. Continuity test shows pin 9 of IC18 is connected to pin 1 (GND). If I take the chip out, it's true for the socket. Looks like pin 9 connects to jumper 25.

Is this expected? Seems weird.

#60 6 days ago
Quoted from kevinclark:

Uhh. Continuity test shows pin 9 of IC18 is connected to pin 1 (GND). If I take the chip out, it's true for the socket. Looks like pin 9 connects to jumper 25.
Is this expected? Seems weird.

Edit: Nevermind, it’s expected. Onward!

#61 6 days ago

I've replaced IC 10 on the driver board. That's #4 so far. Glad I bought so many just in case! PIAs check out at this point except pins 10-13 on PIA 11. They're stuck high even though all the other outputs are blinking just fine. This behavior persists across different chips that looks good when plugged into other sockets.

Mark's switch matrix test (using Leon's ROM: http://pinball.flippers.info/system6repairpart5.asp) works for pins 1,2,3,5 and doesn't for 6-9, which makes sense because 6-9 are supplied by 10-13 on PIA 11.

6-9 *do* seem to flutter a bit, dropping as low as 3.8V (from a high of about 4.3V, which is what my PSU seems to supply). In contrast, TP 8 sits at about 4.25V consistently with no flutter.

Lamp row and strobes check out. Solenoids check out. Feeling like I'm in the home stretch. What's up with those 4 pins?

#62 5 days ago

Renamed the thread to reflect reality (4 of 5 PIAs bad).

Current (hopefully last issue) is pins 10-13 of driver board’s IC 11 are stuck high when using Leon’s test ROM that pulses PA/PB0-7 on the PIAs. Voltage does seem to fluctuate, but only goes down to the high 3V range, not low enough to be low.

#63 5 days ago

My money is currently on the 7406 attached to those 4 pins being fried and pushing voltage back, raising the low higher. Pulling it off the board so I can test it on a breadboard.

#64 5 days ago

Oh yeah, IC17 just sits high no matter what.

#65 5 days ago
Quoted from kevinclark:

Oh yeah, IC17 just sits high no matter what.

Nice to see you're making good progress!

#66 5 days ago

@pincoder Feels like I'm starting to get the hang of this... and then I find something new I don't understand

Last issue on the driver I'm having trouble with: special solenoids seem to work except for #6. When I ground 2J13-9 (via a jumper to 2J10 which is grounded against 2J8's grounding plate) I get no light up with the test LED connected to 2J12-9 and 5V. Mark's guide says I should walk the ground backwards through the components, presumably through 2J12-9 (which lights when I ground it directly).

Before 2J12-9 is an NPN (Q12 - TIP122). The collector seems to be getting flashed in the same rhythm as everything else in Leon's ROM, but I'm not seeing that flash on the emitter. The base appears to sit around 0.6V (according to my DMM, no oscilloscope here). The collector seems to be about 0.1V (but pulsing according to the logic probe). The emitter is about where the base is. None of them set off the test LED as being a valid ground. If I understand NPNs right, the 0.6V should be good enough to activate it and pull the collector (connected to 2J12-9) down to the solenoid ground. But I'm not seeing that, and even connecting the test LED directly to Q12's emitter doesn't seem ground-y enough for the LED. What gives?

Before that in the chain, the collector on Q11 shows dancing, but also doesn't light the led. It shows 5V on the collector, 0.06V on the emitter (which seems low enough it should light the LED?), and 0.25V on the base.

If I move back to IC 9, pin 13 doesn't do anything when grounded, 11 makes the flipper relay switch and 12 does nothing. 11 and 12 both show some dancing on the logic probe, but at different rates.

Here's one more thing I don't understand - if I ground 2J13-9 and put the test LED on 2J12-8, I get a very faint flash from the test LED. I don't know if that has anything to do with anything, but seemed funny.

#67 5 days ago
Quoted from kevinclark:

Before 2J12-9 is an NPN (Q12 - TIP122). The collector seems to be getting flashed in the same rhythm as everything else in Leon's ROM, but I'm not seeing that flash on the emitter. The base appears to sit around 0.6V (according to my DMM, no oscilloscope here). The collector seems to be about 0.1V (but pulsing according to the logic probe). The emitter is about where the base is. None of them set off the test LED as being a valid ground. If I understand NPNs right, the 0.6V should be good enough to activate it and pull the collector (connected to 2J12-9) down to the solenoid ground. But I'm not seeing that, and even connecting the test LED directly to Q12's emitter doesn't seem ground-y enough for the LED. What gives?

I think I'm just describing a broken transistor. When B has higher voltage than E and C, E and C should be connected and the current should be multiplied by some factor, right? I can establish that E has continuity with ground, but the collector never has continuity afaict. But I'm also not able to get continuity when I try it with special solenoid 5 which is flashing the test LED just fine. If I connect my DMM in continuity mode to the collector and to ground, the test LED fuzzes up instead of flashing (like I connected a capacitor to the network) but I still don't get tone, so I don't have a lot of confidence in the diagnosis. I'm wondering if I'd need an oscilloscope to detect this kind of issue..

#68 5 days ago

Q12 tones on my DMM's diode test when connecting collector and emitter with the power otherwise off to the board. Guessing that's it.

#69 4 days ago
Quoted from kevinclark:

Q12 tones on my DMM's diode test when connecting collector and emitter with the power otherwise off to the board. Guessing that's it.

I'd say Q12 is messed up. but also something else along the way (Q11,IC9, IC7). You should be able to compare the circuit of any working special solenoids to the problem circuit. You can also take note of IC6 and IC7, ie would swapping them make any difference.

I dont know how fast leons ROM toggles back and fourth. If it's too fast your DMM might not be able to take a correct voltage reading. If you want to slow things down you could use the 10-solenoids ROM. It will simply enable all of the special solenoids and keep them that way. You could then activate the ST6 circuit and your own rate by grounding 2J13-9 and make comparisions from there.

I'm not sure about the 2J12-8 activity but I'm guessing it will go away when you get 2J12-9 working properly. Either way I dont think it's a huge issue because it doesn't appear to be enough current to drive a solenoid anyway. Is 2J12-8 the only one that it happens on?

#70 3 days ago
Quoted from pincoder:

I'd say Q12 is messed up. but also something else along the way (Q11,IC9, IC7). You should be able to compare the circuit of any working special solenoids to the problem circuit. You can also take note of IC6 and IC7, ie would swapping them make any difference. I dont know how fast leons ROM toggles back and fourth. If it's too fast your DMM might not be able to take a correct voltage reading. If you want to slow things down you could use the 10-solenoids ROM. It will simply enable all of the special solenoids and keep them that way. You could then activate the ST6 circuit and your own rate by grounding 2J13-9 and make comparisions from there.

Thanks for calling this out. I'm pretty confident IC7 is also an issue. I think I finally wrapped my head around the special solenoid circuit after reading the solenoid section in this pdf: http://arcarc.xmission.com/Pinball/PDF%20Pinball%20Misc/System%207%20Fix%20Guide.pdf

And the post on how the pull up resistor is functioning here: https://pinside.com/pinball/forum/topic/understanding-system-11-solenoid-driver-circuit#post-834157

Just to say it out loud in case someone else finds thisL the 7408 (IC 6,7) is about supporting the solenoid being fired by the MPU AND the playfield trigger and the 7402 (IC 8, 9) is about requiring the blanking signal to be HIGH (though it gets inverted before it goes into the 7402 so it's checking for LOW on both sides).

My IC7 is continuously pulsing on pins 9/10 no matter what's happening on 2J13-9. But the source for pin 10, 2P1-26 (ST6), looks like it's doing the Leon's-ROM-Dance just fine so I'm wondering if there's a short in the chip that isn't enough to pull down the voltage on the interconnect but is enough to make it wobble on the chip. Anyway, that looks really wrong and IC6 doesn't reproduce that when I use analogous pins.

Quoted from pincoder:

I'm not sure about the 2J12-8 activity but I'm guessing it will go away when you get 2J12-9 working properly. Either way I dont think it's a huge issue because it doesn't appear to be enough current to drive a solenoid anyway. Is 2J12-8 the only one that it happens on?

If I leave 2J13-9 grounded, 2J12-3 through 8 all have a gentle blink on the test LED. Curious.

#71 2 days ago
Quoted from kevinclark:

Just to say it out loud in case someone else finds thisL the 7408 (IC 6,7) is about supporting the solenoid being fired by the MPU AND the playfield trigger..

Small correction here: The CPU cannot fire the special solenoids. It can only enable them. This is the function of IC6 and IC7 (AND gates). Once enabled, they are fired by the associated SWITCH SOLENOID X TRIGGER line.

So when they do fire, the CPU has no idea. This makes them behave more like a reflex. If there's an associated scoring switch on the playfield, the CPU will detect that and act accordingly.

I'll have to comment on your other stuff when I get some time later..

#72 2 days ago

Can't the program fire up to 6 of the special solenoids via the PIA control ports? That's the way I always interpreted it... and I seem to recall that games do fire them in solenoid test?

#73 2 days ago

Trolll
Yo!

#74 2 days ago
Quoted from slochar:

Can't the program fire up to 6 of the special solenoids via the PIA control ports? That's the way I always interpreted it... and I seem to recall that games do fire them in solenoid test?

Yes, even if during the game the solenoids operate directly by playfield switches, the CPU can trigger them via PIA in solenoid test.

#75 2 days ago

Agree, solenoids can be fired by either the special switch inputs OR the ST1-6 signals from the various PIAs. The PIA and special switch input are combined at the 7408 AND gates IC6 and IC7 then combined with the blanking signal through the 7402 NOR gates IC8 and IC9, then drives the pre-driver and power transistors.

#76 2 days ago
Quoted from kevinclark:

My IC7 is continuously pulsing on pins 9/10 no matter what's happening on 2J13-9. But the source for pin 10, 2P1-26 (ST6), looks like it's doing the Leon's-ROM-Dance just fine so I'm wondering if there's a short in the chip that isn't enough to pull down the voltage on the interconnect but is enough to make it wobble on the chip.

So... I was looking at IC6 and 7 upside down. Almost every chip (aside from 6,7 and 13,14) are facing left on the driver board and the chips I've got have a dot on one side and a notch in the other. I saw the dot and assumed it was another pin 1 marker and didn't notice how wrong I was until I went to sanity check my results before pulling chips.

IC7 and IC9 are working as expected afaict. The duration of the blink on IC9 is pretty short and a little hard to make out on my logic probe with daylight, but I think it's right. My TIP122s should be in later today and I'll pull Q12 and see if that resolves it. If not I'm going to grab 10-solenoids and give it a fresh look.

#77 2 days ago
Quoted from Schwaggs:

Agree, solenoids can be fired by either the special switch inputs OR the ST1-6 signals from the various PIAs. The PIA and special switch input are combined at the 7408 AND gates IC6 and IC7 then combined with the blanking signal through the 7402 NOR gates IC8 and IC9, then drives the pre-driver and power transistors.

It's nice to see that other people are also following along and watching for misinformation!

I did some testing, and I was wrong. Thank you for pointing it out. The ST1-6 solenoids CAN be fired from the CPU. See attached for test cases.

Having said that, perhaps I should rewrite 10-solenoids to also fire the ST1-6 solenoids. Thoughts? Can you think of an example of a game and condtion where the CPU would actually do this? Possibly a ball search? Does any system 3-7 game even do a ball search?

Let me know what you'd like to see in the 10-solenoids ROM
test (resized).png

Edit: "B" ST6 Enable is what I've been calling the PIA1 PB2 pin. I'll need to rename it

#78 2 days ago

No Ball search on 3-7 games. Solenoid test is the only place they fire the specials directly, and it presents a false result, since it only tests the path from the control ports, not the actual special solenoid firing path, so I'd say it's optional, or fire them, and note that this will test the transistors, but not the actual game condition path.

#79 1 day ago
Quoted from pincoder:

It's nice to see that other people are also following along and watching for misinformation!
I did some testing, and I was wrong. Thank you for pointing it out. The ST1-6 solenoids CAN be fired from the CPU. See attached for test cases.
Having said that, perhaps I should rewrite 10-solenoids to also fire the ST1-6 solenoids. Thoughts? Can you think of an example of a game and condtion where the CPU would actually do this? Possibly a ball search? Does any system 3-7 game even do a ball search?
Let me know what you'd like to see in the 10-solenoids ROM
[quoted image]
Edit: "B" ST6 Enable is what I've been calling the PIA1 PB2 pin. I'll need to rename it

No ball search that I know of but some games will fire solenoids at the end of a ball. For example, Firepower will fire the "lock" solenoids to eject any locked balls at the end of a game or whenever a ball is detected in a saucer at power-up or during game over.

Black Knight programmatically fires solenoids in the 2 ball lock areas to release balls.

Blackout programmatically fires the solenoid in the saucer

This type of test would be a valuable add IMO.

#80 17 hours ago
Quoted from Schwaggs:

No ball search that I know of but some games will fire solenoids at the end of a ball. For example, Firepower will trigger the "lock" solenoids to eject any locked balls at the end of a game.
Black Knight programmatically fires solenoids in the 2 ball lock areas to release balls.
Blackout programmatically fires the solenoid in the saucer
I think Gorgar programmatically fires the solenoid in the "pit" (not sure on this one).
This type of test would be a valuable add IMO.

Awesome. Thanks for the input guys. I'll see about making the change in the next release.

Promoted items from the Pinside Marketplace
From: $ 40.00
Cabinet - Other
Rock Custom Pinball
$ 8,999.00
Pinball Machine
Classic Game Rooms
There are 80 posts in this topic. You are on page 2 of 2.

Hey there! Got a moment?

Great to see you're enjoying Pinside! Did you know Pinside is able to run thanks to donations from our visitors? Please donate to Pinside, support the site and get anext to your username to show for it! Donate to Pinside