Firepower System 6 Williams Switch Matrix problem driving me crazy
Ron Lyons
Help! I've got a Firepower I just picked up the other day that I'm
bringing back to life, and It's trying to kill me.
Anyways, the switch matrix has an issue that's making me pull my hair
out. I've swapped out cpu boards and driver boards several times, and
the problem has presented itself on 2 or 3 different boards. I'm
fairly certain that the problem is not on the driver board, because
I've swapped it out with 3 different ones, they all do this.
My problem is, when you hit more than 1 of the standup targets in the
middle of the playfield (1-6) sometimes, you lock the game up. No
other switches do this. You can hit all the switches you want until
you're blue in the face, and it won't do it... but when you start
hitting multiple standup targets in those two banks, it'll work fine
for a few seconds, then the whole machine goes dead (except for
G.I.).
I can't find any rhyme or reason to it, all of those switches are in
the same column, and only share that column with each other.
I was sure I figured out a way to find it... so I took the solenoid
fuse out. It still does it. I took the G.I. fuse out. It still does
it. I took the CPU controlled lamp fuse out, and it still does it. I
took the flipper fuse out, and it still does it. I unplugged the
sound board... and it still does it. I took the display fuse out...
and it still does it.
So it's not doing it because it's shorting power from the lamp matrix,
the G.I. voltage, or the solenoid voltage. I can't find anything
shorted out under the playfield, and I believe I've tested all of the
diodes on every switch in the game!
If you go into switch test, it does it there as well. EVERY SWITCH
works, if you hit them one at a time. If you hit two at a time,
sometimes that works too... and then sometimes it locks up the
machine. In Switch test, none of the cpu controlled lights are on,
either, so it's not a "when this light is on and this switch is
pressed" type of thing.
Is this likely a bad diode? Why is it doing this, can somebody
explain to me what I should be looking for? I know the driver board
strobes the switches, is that 5 volts (or 18 strobed down to 5) and is
it resetting because 5volts is hitting 5volts or something?
Any ideas are greatly appreciated! I'd like to get this sucker up and
running.
Ron Lyons
New roms, new rom sockets, new 40 pin connector, new header from the
power supply, 4.9 volts on the pcb. I don't think it's a 40 pin
connector issue :) (could be, you never know)
Ron
Hans
I know this may sound odd, but do you have any system 7 games laying
around? The system 6 switch test is more basic, and only shows the
most recent switch pressed. It won't show a stuck switch if it's
being masked by another one. A system 7 test cycles through all the
switches that are currently activated. Plus, a system 6 will still
show the switch on the display, even if you've let it go. A system 7
will clear the display when the switch clears.
If you throw a system 7 board (with system 7 ROM's) into it, only
connect the switch connectors to the driver board, and see what it
does.
Whatever it is, the problem seems isolated in column 2. Any specific
pattern which standups does it? Is it only 1 and 6? 2 and 5? or any
pairing?
Probably wouldn't hurt to shoot continuity from the individual
switches back to the wiring connetor at the driver board (with the
connector NOT installed on the driver). Make sure that column 2 isn't
somehow showing up on another column. Same with the rows, make sure
each switch only shows up on the correct row.
-Hans
AVP Pinball Division
news:c5558cc2-2c9e-4212...@k15g2000prk.googlegroups.com...
Whoa, take a breath... In switch test, hit only one standup at a time. Is
only that switch's number showing? Now hold that one and hit a second, then
release the first. Find the exact combination that causes the lockup.
Check the diodes on those switches and the harness leading to them for
shorts or broken insulation. If that checks out good then go back and
triple check all of your work on the sockets and interconnect. Last of all
change the 6821 responsible for that part of the switch matrix.
--
Pistol Pete
AVP Pinball Division
Towson, MD
410-583-9200
www.AVPpinball.com
ser...@AVPpinball.com
Hans
wrote:
> "Ron Lyons" <rlyo...@carolina.rr.com> wrote in message
^ what he said. I need to stop giving advice when zombified from
only having two hours of sleep.
His version is much simpler to start with than mine.
-Hans
Ron Lyons
one and three, sometimes that will do it. Sometimes 1 and 2 will do
it. Sometimes 4 and 6 will do it. You can't catch it very easily
because if you sit there and do it slowly, it doesn't do it. I've
double, triple, quadruple checked the 40 pin connector, 6821, and
connector on the board, and then on two other boards. And then with
another cpu on all three of those boards. I suppose it could be a
6821 problem, but that doesn't seem likely.
I'll try tracing those wires down through the cabinet and see what it
says.
As for the switch test, it's not doing me any good, none of them show
up twice, and if you take all the balls out of the game it doesn't
show any stuck switches. The only thing in that column is those 6
switches and all of the diodes test good... I'm going to try testing
continuity on the wires though next, we'll see what that does.
Thank you,
Ron
firepower
One comment I have to make is that those 6 standups (1-3 and 4-6 are
each a "pair") do something else. They fire solenoids 2 and 3 which are
reported as "unused" in all the manual lists. But the code is there and
running... for the drop target reset coils.
Just needed to say that. You don't have wires underneath that would
have gone to the DT reset coils? I say that because they removed the
drop targets and went with standups late in the production day... some
games still have the wiring in the harness for the switches.
Just a shot in the dark and something to bear in mind and to rule out.
I know this sounds weird, but it might be the solenoid PIA being
triggered, or even switch wiring you don't expect to be there hidden in
the harness!
I would expect you to have to complete the sequence 1-3 or 5-6 for it to
"lock up" if the phantom solenoid #2 or #3 is the cause... but maybe
you aren't paying attention to which arrows / switches have been "made"
before? Can you just hit 1 & 2 from a new game start and it freezes?
Is it possible to connect some LEDs to the tabs of the power transistors
with a current limiting resistor and tie them back to +5v (or +12v) as a
test? See if the lockup is right when they light?
You can see the "resets" happening here in a video. I wired sol #2 and
#3 to LED flashers. The flashers are wired "hot" like a coil, so the
solenoids can ground them and they flash.
Prototype: http://www.youtube.com/watch?v=0PNTTdCzt6I
Under the plastics: http://www.youtube.com/watch?v=Oc3CY2v5ayg
I later changed the Flashers to bright white 5+1 LEDs - so they are more
visible to the player. It was a lot of fun to figure out and hook up,
still on the game as a "feature".
-Richard
Pinball Plus
Have you tried disconnecting the coin door and see if it still does
it . Ive solved lots of sw issues on system 6-7's by doing this . Lots
of areas for shorts to affect the matrix
Ron Lyons
hit just one target, or a couple, it won't reset. If all three have
been triggered, though, sometimes it will reset. Hmmm. What do you
think about me cutting the transistors off the board that run
solenoids 2 and 3?
Ron Lyons
>
> - Show quoted text -
I just tried it with the coin door unplugged, still did it. Pretty
much confirmed though that it has to have all three targets 'scored'
to do it. Since the targets don't physically 'stay' connected... I'm
thinking this has to have something to do with that solenoid that
fires to reset the drop targets!
Richard, was there somekind of switch that told the computer that the
targets reset? My issue doesn't appear when it first lights all
three, just if you start hitting two targets after all three are
'lit'. So why would it need to have already 'fired' the solenoid,
before the issue pops up?
Thanks,
Ron
Pinball Plus
Isd the game acting like a slam tilt reset or does it lock up
(displays go blank etc.)?
Ron Lyons
I just switched the driver board out with another one (again) and it
still does it. You can play the game fine, but if the ball hits two
targets, you've got about a 20% chance it's going to lock the game
up.
It doesn't have any extra wiring past the head connector, for the drop
target coils or the extra switches (I read on Richard's site there
were two point ones, and two that showed all three targets were
down). All the 6 switches look like they're wired properly, all the
diodes are facing the right way, and all the diodes test out right (in
circuit). None of the diodes are touching another blade, etc.
Thanks,
Ron
Ron Lyons
runs through, and checked continuity on each pin to every other pin,
and they're all open.. with the balls out of the game. No closed
switches, no lines shorted to each other (on the playfield). I'm
going to rework from there up to the board next. It's a pain.
Ron
firepower
I wouldn't - you have tried different driver boards...
Really what you are after is finding a short in the wiring harness to
+5v or another part of the circuit that the solenoids are grounding.
Much better to troubleshoot with them there!
Firs thing to check, are the tabs of 2&3 running at a voltage with
respect to ground? They shouldn't be, as no wires should exist at
Driverboard J11 pins 5 & 7. This was just a shot in the dark, and
should only be considered so it can be ruled out. There were early
production games that had the extra switch wiring at the targets.
Sounds more and more like the elusive blanking is kicking in, and that
puts the problem of the CPU or pin 37 of the driverboard "40-way". I
wire up an LED to blanking on my driverboard interconnect- Clay covers
this. You can also use a logic probe on TP #4 of the CPU.
Check if blanking goes low when the problem happens (lockup)?
Also what LEDs are lit on the CPU after it locks? Both lit?
-Richard
firepower
The series (or all targets down) switch *would have* been at the switch
positions below target #3 (sw 20)and target #6 (sw 24) marked as not
used in the switch matrix. That's some of the extra harness wiring that
can exist.
If you have nothing at J11 pin5 5&7, then there can be no extra #2/3
solenoid wiring, But with the switch matrix, you will have the row and
col wires going out to the playfield, just probably not the Row Wht/Yel
and Wht/Gry wires near the switches. There is a 1-6 "target wiring"
diagram I did years ago and it's hidden on my firepower site here:
http://www.firepower.2ya.com/red-bullseye.html
Worth checking, but I'm going to bet your switches are wired up right.
On to another idea lamp matrix:
I just thought given you are not finding any problem with target
switches, you need to consider other things those targets complete.
One is firing of those two solenoids, another is lamps which that were
out that get lit. So here's something in feature settings to check to
rule something to do with the lamps getting lit...
Set the function #31 to 00 - this sets to light only one eject hole
(lock) on complete of targets 1-6, and no "target memory" ball to ball.
Then do your test.
When does it happen? Usually it (randomly?) lights 2 locks at a time
which would make finding a lamp fault harder.
Obviously the insert lamps to consider and check would be the ones for
the locks being shorted there isn't much room at the eject holes. So
maybe... the ones under the orange arrows 1-6 are also worth looking
over although less likely from your explanations.
Next, it's GOT to be the RAM or game code:
By now, you will have eliminated almost all the other possibilities.
Be sure the code is good for All the Eproms and the memory at IC13,IC16
and IC19 is also working perfectly. I think firepower (and other games)
use IC19 as "scratch memory" when games are in prograss, to track
multiple player achievements (lamps lit, locks made etc...) So *all*
the memory including the 5101 at IC19 has to be perfect. It's not just
being used for audits and settings!
I've seen all sorts of strange and wonderful glitches on this game when
the working memory is getting corrupted on a Firepower!
The 5101's are notorious for getting zapped due to static (or power
spikes), as they are CMOS RAM. Leon's test ROM can confirm if the
memory looks good, but as IC13 and IC19 are usually socketed I just swap
them over for another game's chips when I
Lastly, are the Eproms (game code) and the MPU board - the worst problem
to tackle:
I know it's a pain (from experience) with Firepower as the "combo rom
mod" means it isn't a straight swap with another Sys 6 game. I have
replaced the System 6 board in my Firepower with a system 7 board. It
works so much better. I don't know why, maybe the 2114 RAMs or better
chip addressing? Also then I can swap to a board from my BK (or another
working Sys7 game I own) if I have a fault.
If you are running any of the original PROMs on Sys 6, change them out
for new Eproms first, before swapping the MPU! The programming in the
Factory Green Flipper1&2 PROMs rarely go wrong, but the do legs tarnish
and rot. It sounds from all the board work you did that you wouldn't
skip this important step. It might be a cool idea to just to pull all
your Eproms and verify them against the original code in a ROM burner,
if you get this far and haven't cracked the problem. If it's not
memory, it's code or a glitchy CPU?
So maybe try another MPU board, if you can get one and rule that whole
board out? I'm becoming more certain the fault will be somewhere there.
A system 7 board is easier for a test that modifying another System6 to
run the Firepower "combo ROM". But for Sys7 you need a different
Eproms, a 2532 Eprom for IC17 (the "combined" green flipper ROMs), and
two new 2716 Eproms -you need the original "game ROM" at IC20 and the "3
proms code" in IC14.
If you have an 2716 original game ROM, that can go at IC20 instead in a
system 7 board and save some trouble. As I said the factory ROMs are
either good or bad, but the programming doesn't usually change. The
PROM starts out as all "1"s and is burned to "0" for each bit they
program like opening a "fuse"!
Easiest is the method from the pinrepair.com guide. You burn the above
Eproms you need and then can leave the System 7 board jumpers at their
default.
To digress a bit (sorry to spam you if you're not interested just ignore
this paragraph). I cover the above as "Firepower Method A"... you can
also look at a "Firepower method B", which runs the *original* 2716
Green Flipper 1&2 Eproms (or Proms). However a "combo rom" needs to be
written out to a 2532 Eprom, and you need to change jumpers away from
the defaults. the upside is that you burn only one Eprom and this is a
lot closer to true emulation of System 6 on a System 7 MPU. And the
diag button works. My page is here:
http://www.blackknightpinball.com/system6-games-on-system7-mpu.htm
Note that you are actually changing over the OS, the Green Flipper ROM
code when re-burning for a System7 MPU. So if the fault follows you to
Sys7 - you might need to also change out the rest of the game code for a
new one. At least you would know the fault is there. Or not.
I'm on a different time zone (GMT) so it sometimes takes me a long time
to reply or I can reply at times when most folks aren't around. Hope I
didn't get to far off topic. I love that game and it has taught me
whatever I know about pinball technology. And I'm still learning.
Best wishes on finding your firepower gremlin.
-Richard
Ron Lyons
wrote:
>
> - Show quoted text -
Thanks Richard;
I'm going to work on it again today. I worked on it for 8 hours
yesterday. 8 hours over a switch fault.
I've got the 2532 'mod' done to my system 6 board, so I've got new and
verified flipper roms, and then the 2532 mod installed on the board.
I also did the same thing with another cpu.. so I've got two cpu's,
and 3 different driver boards that it's doing this on. I know it
sounds like a board fault, but I'm scratching my head trying to figure
out how it's doing the same thing over all those boards. I know a lot
of people say that the 40 pin interconnect is always the problem when
you have an issue on one of these games, but I've already replaced all
that on all of these boards, and have checked continuity between the
boards over and over again, and it's how it should be.
I'll check some of the things you mentioned today and see where it
ends up. I don't think it's a short problem to the lamp matrix,
though, because it does it with all the fuses removed. Maybe to the
lamp ground though, I don't know. You can remove every fuse in the
game except the 5 volt, and it still does it.
The extra wiring for the extra switches isn't there at the switches
under the playfield, and it doesn't have the u cutout for the point
target. The two solenoid wires don't leave the head connector. I
think today I'll try burning another rom just in case that one's
trippy. It was doing this with the original '6' rom setup, though, so
I don't know kind of seems like i'm wasting my time with all this.
Also I put a new 5v filter capacitor on the power supply, and new 5v
diodes. No dice.
Ron
Hans
Have you tried disconnecting various segments of the wiring harness?
Rather than just yanking the fuses, try pulling all the connectors off
the driver board other than the two switch inputs. If it still happens
then, you might want to even consider pushing the pins out of the
connectors so that only column 2 and those 6 rows are feeding into the
driver board.
Then, if it's STILL happening, I'd pull everything off the driver
board and see if you can cause it to happen by manually tripping the
driver board up on the connector.
Start removing enough stuff, and eventually you'll be able to track it
down to a certain element of the harness, or to the boards.
Are you sure it's because of multiple switch closures? A few things
happen when you finish up a bank of 3 standups, or also when you
complete all 6. You have lighting effects in those arrows, the kick-
save light turns on, additional sounds and speech are played (which go
through the solenoid circuits in the driver board), and lights turn on
in the pop bumpers. So it's always possible that something
downstream is happening, if it's because a bank is completed.
-Hans
firepower
it is a "switch fault", though.
Do you mean you use a *2732* "combo ROM" in the System 6 board, right?
http://www.pinrepair.com/sys37/index2.htm#dips
I would have said 2532s can't work right on Sys 6 MPUs without a couple
of serious address line hacks or an adaptor. (They can be used on the
System 7 boards, though). But they are close enough in pinout to cause a
problem (need a courier 10 or fixed font to view the Ascii drawing):
__ __ __ __
A7| U |+5V A7| U |+5V
A6| |A8 A6| |A8
A5| |A9 A5| |A9
A4| |A11 A4| |+5V
A3| 2 |!CS A3| 2 |!CS
A2| 7 |A10 A2| 5 |A10
A1| 3 |!CE A1| 3 |A11
A0| 2 |D7 A0| 2 |D7
D0| |D6 D0| |D6
D1| |D5 D1| |D5
D2| |D4 D2| |D4
GND|_____|D3 GND|_____|D3
Note the similarities between the pinouts. Move A11 over to the proper
pin, and bring a GND signal to the 2732's !CE pin to keep it enabled.
Oh and 2532 is the right replacement Speech Board Eproms, without any
changes. Let's hope that 2532 was a typo- I've never tried that setup,
shouldn't work like that at all, but might run enough to fool you.
If you swapped the CPUs it shouldn't be the RAMs, then. Are you
changing to a board with new Eproms or just swapping the same ones
between boards and taking the problem with you?
You could try removing the playfield lamp matrix connectors from the
driverboard as a step. J4-J7 off, try it. Then if it goes away, put
back J4 (Lamp Power) and J6 (Lamp Ground) one at a time and see if it
returns. Still nothing? Then add back J5 and J7, that way you can
eliminate or nail if the problem is Lamp PWR, GND or Matrix related. If
not, move on. I don't think just removing fuses and keeping things
wired to the driverboard is a good test. As you say Lamp GND is still
there to short to a voltage and cause mayhem.
Have you checked for blanking staying high on pin 37?
But I don't see how that problem can move to another complete Driver &
MPU pair! pinrepair.com covers how to put an LED (and a resistor) on
the driver board to monitor blanking. I forget where on the site, but I
always have a driver set up like that I can swap in - or use it to test
a new MPU board on the bench. Always check your blanking is working
right before installing a newly repaired board in a game! Learned that
lesson, the hard way. Reminds me, I need some more 2.5A SB fuses...
http://www.firepower.2ya.com/img/driver-mos-2.jpg
:o))
Cheers
-Richard
firepower
But the important point is that it is pulsind sound select solenoids...
according to my notes, it would need to fire 10,11 and 12 all together
to make that sound. Hmmm.... something else I hadn't considered.
If you measure potential at connector J9 for Sound Solenoids 9-13, you
will find they run at a +DC voltage coming from the soundboard, what is
it. Like 4-5v each?
Try removing just the sound board connector to the driverboard at J9 as
a test. Could be those sound select solenoids firing that locks the
game. Does it get through the solenoid diagnostic test without locking
up? That will fire the 5 sound selects at 9-13 separately.
I hope you can track it down, or at least localise it to a single place.
-Richard
Andy
IIRC hitting three standups lights the kickback.
Another area to look at, as if you don't have enough already:-)
Andy
firepower
You can change feature #32 to 00 and then it only lights from making the
"power" targets on the RHS. Worth a try. I was going to mention that
adjustment, but decided my post was getting too long already.
-Richard
Ron Lyons
wiring at the driver board, and the solenoid wiring, and it didn't do
it. Plugged the solenoid wiring back in, and it also didn't do it.
I'm going to try plugging the light wiring back in one at a time until
we get it. If I go into test mode all the lamps work properly but I
guess hitting two switches while a lamp is on must be doing it,.....
be back in a bit.
Ron
Ron Lyons
in. If you leave that one unplugged and everything else plugged in,
everything's fine. I have NO CLUE why 5v goes in at that connector,
what's it running folks? Is that maybe for the flashers? With that
one unplugged, everything seems to work pretty good.
Is that how the board gets it's logic power? Hell it's working
without it, through the interconnect I assume.
Ron
Hans
You nailed it, it's the logic power (gray) and ground (black) for the
driver board. Runs from there to 3J6 on the power supply.
-Hans
Ron Lyons
Upon inspection, it specifically provides power for the solenoid
7408's and 7402's, I checked them all with the diode test on
pinrepair.com.. and they all tested out fine.
I think I'm going to go buy a gascan.
For some of the other things mentioned, I double checked and yes it's
a 2732, sorry about that.
It does run through solenoid test without locking up. Also if I
remove all the solenoid connectors at the driver board (not just the
sound board one) it still does the lock up thing.
...Still digging...
AVP Pinball Division
>Firepower headaches<
Ok, swapping CPU and driver boards doesn't cause the issue. That leaves
harness and power supply. On the power supply check ALL of the fuse holders
for weak prongs or corrosion. Also be sure to reflow the solder on the
headers on the power supply board. Inspect the terminals in the harness by
removing them one at a time. Check for corrosion or terminals that don't
keep tension. If you lose 5 volts to the driver board the CPU will lock as
it loses the blanking signal.
--
Pistol Pete
AVP Pinball Division
Towson, MD
410-583-9200
www.AVPpinball.com
ser...@AVPpinball.com
Ron Lyons
it does it, and also no lights blink or come on the cpu when it does
it.
So to summarize, if the solenoid 7408's and 7402's are getting their
power from the driver board connector (j8), it locks up when you start
hitting the stationary targets in the middle of the game, after at
least 1 bank has been completed. Doesn't always do it, but you can do
it manually with your hand about as soon as you start triggering
multiple switches. If you remove connector J8 and start the game, it
won't do it.
Ron
firepower
the 40-way... as I recall pins 1+2 carry +5v and pins 39+40 are Gnd. I
never hook up J8 on the bench, because it helps to check that the 40-way
seems OK.
The power from J8 goes everywhere on the driverboard. It powers all the
ICs including the PIAs.
Yep, just checked that on a schematic. And it's correct, so you can
work without that connector at J8 while you figure out what's wrong.
From the schematic, the only suspects are filter caps just above J8:
C15 (75uF @35v) electrolytic cap so connect it the right way round.
C36 (.1 uF @ 50v) this ceramic cap has no polarity.
Looking at two driverboards that are laying around my bench awaiting
repair, they have the same exact parts:
C15 100uF @ 10v (TI 85 degrees C)
C36 104 M5J* 0.1uF
* a green ceramic bypass cap, I think the M5J is the dielectric code or
a tolerance and working voltage. My notes say "100nF (same as 0.1uF)
+80% -20% @ 50v
J8 +5v ---*--------*------ +5 (filtered)
| |
--- ---
/-\ ---
| |
| |
| |
J8 Gnd ---*--------*------ +5v return (filtered)
100uF 0.1uF
elec. cer.
Not a clue why they're rated at such high voltages (35v and 50v on the
schematics. Sense tells me the voltage can be only +5v or so, why
wouldn't a 10v spec be fine? Maybe it has to do with their function.
The logic power is already regulated and out of an LM323K (or similar
VR) and also a big filter cap on the PSU. So the +5v should be already
clean... maybe with all the solenoids firing and switching going on,
it's good to have them there (as noise filters). The CPU has them too
at C23 and C24, same values, so probably to filter transients.
I would be inclined to desolder one leg of C15 and without the solenoid
fuse installed power everything up and see if the problem is gone when
you run the game and close switches. If not put it back and try C36.
Betcha it's one of them- and probably the big'un.
-Richard
> ....Still digging...
>
firepower
it. The driverboard *is* grounding the 28-35v solenoid voltages and the
flipper ground returns. So you would have a huge "noise spike"
potential even with snubber diodes on the coils supressing the EMF.
That's why those caps need to work, but why removing J8 fixes your
problem, maybe the voltage is reaching the driver through the new 40-way
connector and the caps on the CPU board are good?
Are the two voltage connectors at different potentials? They shouldn't
be. Just for fun, take the power inputs off the CPU and Driver and with
a meter on DC measure ground to ground +5v to +5v. Both should be zero.
CPU 1-J2 Driver 2-J8
7 * Key
6=| |=9
5=|= +5v =|=8...
4=| |=6
5 * Key
3=| |=3...
2=|= GND |=2
1-| |=1
-Richard
Ron Lyons
It doesn't do it if the power connector to the driver board is removed
(and the game plays with no solenoids since the 7408's and 7402's
aren't getting their 5v.
Sooo, I tested the solenoid PIA with the game in attract and it tested
out fine.
I tested the 7402's and 7408's pin by pin as instructed by
pinrepair.com and couldn't find any issues. All the solenoid drivers
and predrivers test out fine... I replaced the 7408 that drives the
two drop target reset solenoids just for giggles, and that didn't
change anything either.
I'll keep messing with it. 13 hours so far on this 1 issue.
pinballpsycho
wrote:
Wouldn't the reset circuit also kill the blanking circuit? I mean...
while he's checking the logic power, that unregulated +12V (11.6)
power might also come into play because if it goes away for a moment,
the reset circuit should shut the machine off. Hard to see how that
relates to a reset after 3 targets close simultaneously, though.
Ron Lyons
wrote:
> See my reply. I did sort of "figure it out" more when I thought about
> it. The driverboard *is* grounding the 28-35v solenoid voltages and the
> flipper ground returns. So you would have a huge "noise spike"
> potential even with snubber diodes on the coils supressing the EMF.
>
> That's why those caps need to work, but why removing J8 fixes your
> problem, maybe the voltage is reaching the driver through the new 40-way
> connector and the caps on the CPU board are good?
>
> Are the two voltage connectors at different potentials? They shouldn't
> be. Just for fun, take the power inputs off the CPU and Driver and with
> a meter on DC measure ground to ground +5v to +5v. Both should be zero.
>
> CPU 1-J2 Driver 2-J8
> 7 * Key
> 6=| |=9
> 5=|= +5v =|=8...
> 4=| |=6
> 5 * Key
> 3=| |=3...
> 2=|= GND |=2
> 1-| |=1
>
> -Richard
>
>
Richard, I think we're getting down to it. I'll work on it again
tommorow. Before I left (and before I saw your reply) I did go ahead
and replace the large electrolytic cap just above that connector, and
I had the same problem as before. I visually followed the traces from
2J8, and couldn't find where they tie in with the other 5v
connection. They seemed to be isolated to only the 8 ic's in the
solenoid section, but I may have missed something. Can anybody double
check that for me? It'd help isolate it to the solenoids if that's
the case.
I'll also measure the potential difference like you mentioned, and
check to see if I can swap out that small capacitor as well.
Because it's relevant, just thought I'd mention that I've replaced the
filter cap on the power supply, resoldered the header that all of
those voltages leave through, replaced the female connector and all
the pins. On the main cpu, I've resoldered the header, and replaced
the female connector and all the pins. Also I've replaced the 40 pin
interconnector.
Thanks for everybody's help so far.
Ron
firepower
problem", in that the matrix is not the cause of the problem.
And the traces from Pins 1&2 (and pins 39&40) of the "40 way" do reach
around to connector J8. Start from the other way, from the the back of
the board. Find pins 1&2 of the "40 way"... that's +5v.
If you look on the solder side, start at pin 1 and it traces around past
J2 (switch matrix) going just inside the ground trace. Then at the back
of IC14, +5v power hits pin 14 - flip it over and there's a bypass cap
for the IC. On the front, you follow that trace further around the
board clockwise (still just inside the ground trace on the putside of
the board with the screw holes in it)...
Bingo! It runs right under the 4 power pins at J8 (6-9) on it's way to
the 100uF 10v cap at C15.
Ground is easy. Under J8 there's 4 holes drilled in the ground trace.
I'm sure those are "through holes" that connect the front and back gnd
traces together, just at J8. Makes sense.
So no - it doesn't just drive some of the chips or part of the
driverboard. Definitely *all* the +5v lines on both boards are
connected. And of course all the GNDs on both boards are connected.
This is true even out of the game if the boards are mated by the 40-way.
In the backbox the GNDs are also connected together by the screws &
metal rails. It's very important to use all 6 screws on the boards, and
all 6 on the sound and speech boards for good ground bonding.
Did you do the female connectors and male header pins for the Gnd, +5v
and +12v on the PSU board? that could be the culprit as someone else
has said. You could be chasing your tail if the same thing happens with
other CPU/Driver combinations.
Regards
-Richard
On 09/03/2011 03:14, Ron Lyons wrote:
> On Mar 8, 6:50 pm, firepower<harvey_r@_ANTI_SPAM_tiscali.co.uk>
> wrote:
>> See my reply. I did sort of "figure it out" more when I thought about
>> it. The driverboard *is* grounding the 28-35v solenoid voltages and the
>> flipper ground returns. So you would have a huge "noise spike"
>> potential even with snubber diodes on the coils supressing the EMF.
>>
>> That's why those caps need to work, but why removing J8 fixes your
>> problem, maybe the voltage is reaching the driver through the new 40-way
>> connector and the caps on the CPU board are good?
>>
>> Are the two voltage connectors at different potentials? They shouldn't
>> be. Just for fun, take the power inputs off the CPU and Driver and with
>> a meter on DC measure ground to ground +5v to +5v. Both should be zero..
Ron Lyons
wrote:
Richard;
I've been messing with this machine off and on the last few days,
it's still acting up. I'm going to focus on the power supply (again)
next to eliminate that. I've already replaced the male and female
pins on the connector, and the two rectifying diodes, but I think I'm
going to order another 383 power regulator and swap that out.
I've been playing it and shopped out the playfield. The J8 connector
on the driver board was just a wild goose chase, eventually It locked
up with that connector removed as well. I've since figured out it'll
lock up with ALL of the light connectors removed. So it'll lock up
with everything removed but the switch matrix, and I've swapped the
driver board and cpu board a few times each which I hope at least
reasonably eliminates them.
I've also noticed sometimes it'll lock up when just one of the targets
is hit. Like I mentioned before too, it also exibits this behavior in
switch test mode! If you put it in switch test mode and start hitting
the standups, eventually the game will lock up. It seems to do it
faster if you hit a couple at a time, but if you hit just one
sometimes it'll lock up as well. I'm hoping the power supply is weak
(although it tests fine) and maybe it's spiking the power and dropping
it low or something.
I'll rebuild the power supply and if that doesn't fix it, I'll buy
another power supply to make sure. Then I'll buy an axe!
Ron
Hans
Those standup switches. Don't they have a metal backing plate behind
them?
And isn't there fish paper between the switch leafs and that backing
plate?
Do they all have the fish paper? And is it in good shape?
-Hans
Ron Lyons
Hi Hans;
They have nooooo fish paper, and they do have metal backing plates.
The top of the targets I THINK can touch the metal plate. Hmmm. I'll
inspect that closer when I get there to see if everything looks legit.
Thanks,
Ron
Ron Lyons
think they're designed to.. they have a metal backing plate, but the
backing plate isn't grounded or anything. The front of the target
touches it, but since it isn't connected to anything else it doesn't
matter. Even if the two blades touch the plate when hit, it doesn't
seem to matter since they're connecting at the points anyways.
WITH THAT SAID, those 6 targets are the only ones that make it reset
(they're all in the same column).. and the other three on the right
are designed differently. They don't reset the game. Hmm.
Ron
P.S., just to reiterate, this doesn't do it every time you hit the
target. It's after they've been hit many times... so it might take 40
hits to do it, but eventually you hit the target and the game dies.
You may play 1 bad game fine, but once you get a decent game going, it
does it. Sometimes though, it does it about the third time the taget
gets hit. It's rediculous.
Ron Lyons
doing it. This has got to be a switch problem, I can unplug
everything except the switches and it still does it. The switch
matrix tests out alright with Leon's test chip, and it does it with a
couple different driver boards, but there's nothing else it can be.
- New diodes, 5v regulator, filter capacitor, header pins on power
supply
- New female pins on both ends of power harness between the power
supply, driver board, mpu.
- Replaced 40 pin interconnect, everything tests fine with Leon's test
eprom
- Replaced eprom sockets, and used combo eprom
- Game locks up even with everything except the switches disconnected
from Driver board
- All switches test 'open' in switch test, and only show 1 number when
pressed in switch test
- All diodes test o.k. on all switches in the game
- Game also does it with all the fuses except +5 removed
Only problem is when you're playing the game, and you hit the center 6
targets, the game will lock up sporadically, just after hitting a
target. Every target seems to be able to do it, and all the targets
are in the same column. It also does it in switch test mode. It
seems to do it more if you hit a couple targets at once, or if you hit
one repeatedly.
Driver board hasn't had the resistor mod done to it in the switch
area, but I'm thinking that may actually make it worse, I don't know.
Also It's doing it on 3 different driver boards, that all test fine
with Leon's test eprom.
Ron
pinballpsycho
Chripes. Now I think this is driving me crazy. No wonder you're
going nuts.
You're sure there are no cut-through's in the wiring harness under the
playfield, right? I had one where the bundle was run across a lamp
fixture and I had to get a lamp down there good and close close
because it was a white wire, but once I got it in there I could see
the shiner; bare wire laying on a grounded fixture from years of
rocking.
This one is a bear man, but once you finally do find it you'll never
forget to look for it again...
BlueMalibu
Do you know anyone local with a Firepower? If so, maybe you could
swap in their PF. This would narrow it down.
Parker
Ron Lyons
Psyco, I've looked and looked and can't find any place where the wires
are rubbing. The game is actually built pretty nice, the switch
wiring hangs loose at the back of the playfield with plenty of slack
so it doesn't rub on anything, and the switch wires have their own
connector so that bundle is loose from the other wires until it gets
to the back of the playfield. From there on everything looks really
clean.... if I check continuity everything checks out good and isn't
shorted to any of the other switch wiring, and also in test mode all
the switches show up properly (none stuck, only 1 number shows up when
you hit a switch).
I used a meter and checked every switch diode in the game tonight, no
problems found :( none were touching the wrong blade or anything
either, and all the wiring was on the right blades, all the diodes
were turned the right way.
I guess I'm going to take every standup target's wiring off and add
them back one at a time and see if that tells me anything.
Ron
pinballpsycho
That sounds like a plan. Back everything out and put it back one at a
time to see when it bombs again. Maybe even leave the column leads
connected and just remove the rows from the switches. You'll have to
make sure not to break the daisy chains, though, or you'll get
unexpected results.
Ron Lyons
> Hi folks;
> Help! I've got a Firepower I just picked up the other day that I'm
> bringing back to life, and It's trying to kill me.
>
> Anyways, theswitchmatrixhas an issue that's making me pull my hair
> out. I've swapped out cpu boards and driver boards several times, and
> the problem has presented itself on 2 or 3 different boards. I'm
> fairly certain that the problem is not on the driver board, because
> I've swapped it out with 3 different ones, they all do this.
>
> My problem is, when you hit more than 1 of the standup targets in the
> middle of the playfield (1-6) sometimes, you lock the game up. No
> other switches do this. You can hit all the switches you want until
> you're blue in the face, and it won't do it... but when you start
> hitting multiple standup targets in those two banks, it'll work fine
> for a few seconds, then the whole machine goes dead (except for
> G.I.).
>
> I can't find any rhyme or reason to it, all of those switches are in
> the same column, and only share that column with each other.
>
> I was sure I figured out a way to find it... so I took the solenoid
> fuse out. It still does it. I took the G.I. fuse out. It still does
> it. I took the CPU controlled lamp fuse out, and it still does it. I
> took the flipper fuse out, and it still does it. I unplugged the
> sound board... and it still does it. I took the display fuse out...
> and it still does it.
>
> the G.I.voltage, or the solenoidvoltage. I can't find anything
> shorted out under the playfield, and I believe I've tested all of the
> diodes on everyswitchin the game!
>
> If you go intoswitchtest, it does it there as well. EVERYSWITCH
> works, if you hit them one at a time. If you hit two at a time,
> sometimes that works too... and then sometimes it locks up the
> machine. InSwitchtest, none of the cpu controlled lights are on,
> either, so it's not a "when this light is on and thisswitchis
> pressed" type of thing.
>
> Is this likely a bad diode? Why is it doing this, can somebody
> explain to me what I should be looking for? I know the driver board
> strobes the switches, is that 5 volts (or 18 strobed down to 5) and is
> it resetting because 5volts is hitting 5volts or something?
>
> Any ideas are greatly appreciated! I'd like to get this sucker up and
> running.
Guys, I finally figured this out. FINALLY. It was really simple.
The switches just couldn't have anything wrong with them, they're too
simple and looked fine. Also I took the wiring harness apart, and
nothing was shorted together. It did it in the same place... when you
hit a few targets at the same time, on those center 6... but all the
switch matrix stuff tested out fine, and the switches tested fine. I
tried two different cpu's, and 3 or 4 different driver boards... the
only thing that was the same, after I started thinking about it, was
the '4 chip' combo rom chip I burnt onto a 2732. The cpu didn't work
when I first got it, so I burnt new flipper roms. I also burnt the 4
chip combo rom... when I was testing the other cpu, I was able to swap
the flipper roms with new ones I burnt, but I never reburnt the 2732!
It tested fine in my eprom burner, and I had already replaced the
socket (on one board). For giggles, I burnt a new one... and problem
solved, it doesn't do it anymore.
So I had a testing fine rom chip that really wasn't fine. Now the
game plays great. FUN game, too, I'd never played one before this.
WHOO!!! Saved another one.
Ron
Hans
>
> - Show quoted text -
Wow, that's definitely an oddball one. It was passing the self-tests
and everything?
-Hans
Ron Lyons
> Wow, that's definitely an oddball one. It was passing the self-tests
> and everything?
>
>
Hans it was playing 99%, and passing all tests.. .only issue was from
time to time, when you hit the center 6 switches or two of them at a
time, the game would die (the blaking signal would go low)
Turn it off and back on, it'd play fine. Then it'd do it again maybe
the first time you hit one of those 6 switches, maybe the 100th.
Crazy!
Ron
Gott Lieb?
Thanks for the follow up, Ron!
Jim
Mark Smith
end to tough dogs and are just left hanging, wondering.
Mark
"Ron Lyons" <rly...@carolina.rr.com> wrote in message
news:b9375f15-7134-496d...@q40g2000prh.googlegroups.com...
> Hi folks;
> Help! I've got a Firepower I just picked up the other day that I'm
> bringing back to life, and It's trying to kill me.
>
> Anyways, the switch matrix has an issue that's making me pull my hair
> out. I've swapped out cpu boards and driver boards several times, and
> the problem has presented itself on 2 or 3 different boards. I'm
> fairly certain that the problem is not on the driver board, because
> I've swapped it out with 3 different ones, they all do this.
>
> My problem is, when you hit more than 1 of the standup targets in the
> middle of the playfield (1-6) sometimes, you lock the game up. No
> other switches do this. You can hit all the switches you want until
> you're blue in the face, and it won't do it... but when you start
> hitting multiple standup targets in those two banks, it'll work fine
> for a few seconds, then the whole machine goes dead (except for
> G.I.).
>
> I can't find any rhyme or reason to it, all of those switches are in
> the same column, and only share that column with each other.
>
> I was sure I figured out a way to find it... so I took the solenoid
> fuse out. It still does it. I took the G.I. fuse out. It still does
> it. I took the CPU controlled lamp fuse out, and it still does it. I
> took the flipper fuse out, and it still does it. I unplugged the
> sound board... and it still does it. I took the display fuse out...
> and it still does it.
>
> So it's not doing it because it's shorting power from the lamp matrix,
> the G.I. voltage, or the solenoid voltage. I can't find anything
> shorted out under the playfield, and I believe I've tested all of the
> diodes on every switch in the game!
>
> If you go into switch test, it does it there as well. EVERY SWITCH
> works, if you hit them one at a time. If you hit two at a time,
> sometimes that works too... and then sometimes it locks up the
> machine. In Switch test, none of the cpu controlled lights are on,
firepower
unexpected. Usually they fail checksum or won't boot the game at all.
Glad you fixed it!
-Richard
pinballpsycho
Nice to see this one put to bed... congrats on the fix man. I'm also
glad to see that you took a moment to post a resolution for us, so
kudos to you. I'll bet you never fail to consider re-burns again no
matter how obscure the chances!
Congrats.
Jim
Ron Lyons
>
> - Show quoted text -
Yup, just goes to show you... don't test your Eproms, even if they
test fine in the burner. As a matter of fact... this one STILL tests
fine.
Ron