TECH: Black Knight/Sys 7 sound
bobnatlanta
I have done the following, to no avail:
Replaced all molex connectors on sound board
Replaced IC36 on CPU (see my rant on replacing PIAs late last week)
Tested connectivity on the ribbon cable connecting the speech and
sound boards
Here's what is happening:
Sounds all work in test
In game play, there is no sound at game start. Feature sounds
eventually start, usually on the timed drops first. Other sounds kick
in periodically, but are not timed properly. For example, if I lock a
ball, the lock sound is delayed. Once the ball drains, bonus count
down sound is very high pitched. I know this should happen as the
bonus gets very high, but not when bonus doesn't get over a few
thousand. When the final ball drains, the background sound kicks in
as though bonus is maxed out, and repeats until the game is shut
down. If I start a new game at this point, the background sound
conintues in this mode.
In short...WTF?
seymour-shabow
How about the connectivity between the sound connector on the mpu and
the sound board? Sounds like it's missing something or a signal's not
making it. Even with the dedicated PIA for the sounds on system 7, the
board is still activated the same way, the pins get grounded in varying
combinations to produce all the different sounds. Sounds like one or
more isn't making it. You should be able to ground each pin on the
connector and get a sound (except for +5 and ground, which may or may
not be in that signal connector, check schematic)
-scott CARGPB#29
BarryMinnesota13
Sound board I would
Check the 10K SIP resister at IC37. If one of legs is failed you
could have unusual sound results.
Barry
bobnatlanta
> bobnatlantawrote:
Thanks! And here's my update: I grounded each of the pins on the
connector and got nothing...a bit alarming. I noticed that the
connector on the upper right corner of the sound board was not making
good contact - the female pins were sticking out a bit. So I
corrected this and tried again. This time, I got sound when grounding
pin 2 only. All others, nada. I checked the pins with a voltmeter,
and got 5VDC on 2-8, and a wierd reading on 9.
Sound test still works fine with the connector plugged in or not.
So...I'm still ocmpletley stumped.
All thoughts/suggestions appreciated!!
bobnatlanta
AND - connectivity between CPU and sound board is fine.
firepower
>
> - Show quoted text -
1. Your sound board jumpers are all set up correctly? For BK W2, W5,
W7, W9, W10, W15 should be pesent and all others removed.
2. The self test doesn't use the input section on the sound board, so
having a reliable self test just means the card can produce the sounds
and speech. It does mean the CPU and ROMs (including speech if you
hear that) are all good, it has a working PSU that can generate the
+5v logic supply.
You should test first with the CPU board cable disconnected. You
should be able to ground pins: 10J3-2 through 10J3-5 on the sound
board header pins and get a sound produced. If you don't hear sounds
on these pins, you won't get reliable sound effects in the game.
Pins other than 2-5 on the sound board are not sound selects, so you
can ignore them. With the CPU board connected these pins would still
run high (+5v or so) unless being grounded under game control. There
are 4.7K pull-up resistors connecting +5 only to sound board header
pins 2-8. So those readings look right.
3. Sounds cutting in and out might be a CAP problem on the sound
board, I have seen this. Get a replacement kit and do them all at the
same time, I think Ed at GPE has them you can google this group to get
more details.
4. Also there is a PIA (means Pain In the A**) on the sound board too,
I know you don't want to hear that. If it's socketed, you could
change it as a test. The WMS sound board is like a mini cpu card with
a CPU and PIA and it's own PSU. But before you replace that you need
to be sure it's the probelm. There is a Leon test chip for the sound
board which will pulse the PIA and let you test it with an LED on the
bench. See: http://www.flipper-pinball-fan.be
Best wishes
-Richard
bobnatlanta
Situation has not changed.
Test works fine
No sounds at game start, though some kick in during game play. Once
sounds kick in, they are often incorrect or late, and once a ball
drains, the backgrourd sound doesn't stop...Just bizarre.
Game was working fine one, sounds screwed up the next.
I've replaced the connectors and caps, IC36 on CPU, and verified
connectivity of the ribbon cable. I am, once again, completley
stumped.
Any other thoughts appreciated, and please, dumb them down as much as
possible.
Cheers!
Bob
Firehousepinball
How are the sockets/roms? Missing sounds can be due to the masked roms
developing bitrot and breaking a leg. Also they used scanbe sockets in
sounds boards too.. doesn't hurt to replace the sockets.
railbender
I don't have a system 7 but if this was a system 6, I would say make
sure to run the solenoid test to see if one of the sound "solenoids"
is out.
My Firepower had similar problems, wrong background sound when I
started a game, and the wrong sound played when I accomplished major
objectives.
It turns out that the sounds on system 6 are created through multiple
"solenoids" being fired in different combinations. So I had one tip
transistor out (13 I think) that was the reason for my problems. I
found it by running the solenoid test and after the first bunch of
solenoids fired you started getting sounds. I did not get a sound for
one of the solenoids (you can see the solenoid number in the credit
display) and when I replaced the driver transistor and the pre-driver,
everything worked.
Since I don't know if system 7 works the same, this may be way off
base, but the solenoid test takes no time and you will know quickly
Good luck,
Kenny
seymour-shabow
It doesn't - uses a dedicated PIA on the MPU that system 6 and below
doesn't have.
The principle is the same though, sounds like the PIA's lines aren't
making it to the sound board, so same repair methodology applies, during
sound test you should get a ground on each of the lines. If there's a
sound missing in sound test figure out if it's connectors, the cabling,
or the PIA.
-scott CARGPB#29
Retrogameconnection
My guess would be to look at the PIA on the CPU board for sounds....
firepower
> doesn't have.
>
> The principle is the same though, sounds like the PIA's lines aren't
> making it to the sound board, so same repair methodology applies, during
> sound test you should get a ground on each of the lines. If there's a
> sound missing in sound test figure out if it's connectors, the cabling,
> or the PIA.
>
>
> - Show quoted text -
I would go "bare bones" -add a Jumper W1 on the sound card and set the
two DIP switces: SW1 and SW2 to OFF and run without a speech board
until the sound is working 100%. The speech cards don't go wrong
that often, but it does extend the bus on the sound card, so may be
causing your problems. If the sounds are perfect after that (but no
speech, obviously) -you know you need to change sockets or have a ROM
(use new 2532s) or other issue on the speech card.
Next, if you don't want to go on troubleshooting the sound card / CPU
board, I would buy a known working sys 6-7 sound card,or borrow one to
test with. www.pinballpcb.com does a nice repro - and I think the
price is right for a brand new board at $90. Tell James Kohout what
game it is for and you will get a working board with the right sound
ROMs. Can't go wrong, the only issue is you still use your own speech
card, but you would have ruled that out as the issue above.
It will probably fix the problem, and that way you will know it is on
the CPU or cablng if it doesn't. You can always sell your sound card
with the CAPs upgraded to get some money back. Someone will use the
Leon test ROM and get it working, they aren't too bad to fix once you
replace sockets on the bench. See my site: http://www.firepower.2ya.com/FP-Parts-Snd.html
although that is more geared to Sys 6 sound problems. But the sound
and speech boards used in both games are identical. And of course you
should use the http://www.marvin3m.com/fix site (Clay's).
Best of luck.
-Richard
Chad
Very helpful thread... thanks guys... I'm working on a Black Knight
sound issue right now too. No speech, sounds work if you ground the
right sound card header pins but only one sound plays in sound test.
Ed's cap kit is installed, new sockets on the speech card, all new
headers on the sound card (still need to replace or repin connectors).
If you manually ground the header pins is that bypassing the sound
card PIA?
firepower
NO.
The sound card is still using the input/output onthe PIA to trigger
sounds from the ROMs and send to the sound generator chips.
You are bypassing the PIA on the CPU board obviously, and the CPU to
sounds board wiring. Better to ground pins on 1J8 on the CPU board.
That tests your cabling as well. It's good to get the sound card
working from its own self test and then move outwards to grounding
pins and running a sound test from the Dignostics via the game's
solenoid tests.
See the logic diagram here:
http://www.blackknightpinball.co.uk/BK-Logic/Sys67-Sound-Logic.pdf
I also just finished drawing the BK (WMS sys 7) CPU Assembly Drawing
PDF:
http://www.blackknightpinball.co.uk/BK-Logic/sys7-cpu1.pdf
Looking for anyone finding errors / corrections I need to make. I
showed the optional LEDs and components that are sometimes present on
older sys 7 CPU boards.
Cheers
-Richard
Chad
> On Jun 24, 4:37 pm, Chad <chad.to...@gmail.com> wrote:
> > If you manually ground the header pins is that bypassing the sound
> > card PIA?
>
> NO.
> The sound card is still using the input/output onthe PIA to trigger
> sounds from the ROMs and send to the sound generator chips.
Cool, thanks. I guess the PIA is the place to look in this case,
since manually grounding the header pins *is* bypassing the PIA, yes?
firepower
It's bypassing the CPU's PIA not the sound card's PIA. The sound card
PIA at IC10 is still being used.
I forgot to say the sound select input pins on the system 8 CPU board
are J8-8 to J8-12:
1J8 Pin 1 Comma 3 & 4 Brown/White
1J8 Pin 2 Comma 1 & 2 Violet
1J8 Pin 3 Key (*) N/C
1J8 Pin 4-7 No Connection White
1J8 Pin 8 Sound Select 4 Yellow
1J8 Pin 9 Sound Select 3 Green
1J8 Pin 10 Sound Select 2 Blue
1J8 Pin 11 Sound Select 1 Red
1J8 Pin 12 Sound Select 0 Red / Yellow
Hard fought information. Not available everywhere.
-Richard
Chad
> On Jun 24, 5:47 pm, Chad <chad.to...@gmail.com> wrote:
>
> > On Jun 24, 12:40 pm, firepower <harve...@tiscali.co.uk> wrote:
>
> > > On Jun 24, 4:37 pm, Chad <chad.to...@gmail.com> wrote:
> > > > If you manually ground the header pins is that bypassing the sound
> > > > card PIA?
>
> > > NO.
> > > The sound card is still using the input/output onthe PIA to trigger
> > > sounds from the ROMs and send to the sound generator chips.
>
> > Cool, thanks. I guess the PIA is the place to look in this case,
> > since manually grounding the header pins *is* bypassing the PIA, yes?
>
> It's bypassing the CPU's PIA not the sound card's PIA. The sound card
> PIA at IC10 is still being used.
Hrm... still learning the circuit here. So grounding a pin
individually still involves the sound card PIA by pulling down voltage
on a PIA input pin? What I don't get is how the sound card test cycle
is involving the CPU's PIA - if it isn't then the sound card test
cycle is no different than grounding pins individually.
firepower
>
> - Show quoted text -
Well it is kind of different, here's what I think:
The diagnostic on the sound board is held in ROM. Whern you press the
diag button, the switch triggers a NMI (non-maskable Interrupt) on the
CPU and it executes instructions from the ROM that sends data over the
Data BUS (which the CPU and PIA share). This then tells the PIA to
set up to output sounds on PA0-PA7 which than go through a 1408 D/A
converter at IC15. Except for a little amplifiction, that's basically
it.
So the sound test DOES use the PIA in the same way that grounding pins
does on the output side, it just doesn't use PB0-PB7 on the input side
of the PIA.
If you get all the sounds grounding pins on the sound board, but only
one sound on the sound test, it points to ROMs / ROM socket / jumper
settings.
Also may have ONE of your sound select pins shorted to ground
permanently during gameplay. Run the game up without starting any
game sounds and measure voltages at the sound select pins. The should
be logic high, +5v (or so) supplied through that pesky SR1). None of
them should be at ground (0v with respect to the ground plane), as
this would stop other sounds from working. If you get different
results with the CPU (or driver board on sys 6) connected to the sound
board, it would point to a PIA (or other fault) on the sys 7 CPU
board, or a solenoid transistor #9-13 or PIA or other IC chip on the
Driver board for Sys 6.
Sounds to me like your PIA should be working. It would help to have
Leon's sound board test chip and a logic probe or Led on 2 jumper
wires to rule out the sound board PIA first and then move to the next
components from there. Google "Leon +pinball" and you will find his
site.
If it was the other way around, let's say the sound test was perfect.
But grounding pins 2,3,4,5,7,8 produced NO sounds.
Then you could suspect first the input section of the PIA (unless that
tested good with the Leon ROM),
the resistor array at SR1 (4.7K ohm 10pin SIP which can go faulty,
and can just be replaced by resistors),
the resitors R2-R10 (100 ohm),
and especially IC5 (4050 Buffer) .
There are also a few 470pF Caps there, if ANY of these are shorted
then the sound lines are selected all the time which means lots of
other sounds can't fire.
All of this can be checked with Multimeter/logic probe. Caps should
show as OPENs on ohms, resistors need to test within about 10% of
value to be connsidered good. So I take 100 ohms to read around
95-105 ohms, give or take a few.
Long live BK - a great Steve Ritchie game.
Cheers
-Richard
Chad
Wow, great effin response... :)
I'll check voltages on the sound select header pins tonight. Sound
board sockets are newly replaced with SIPs - though this problem
existed before I did that and I did check for shorts before I put the
ROMs back in. I don't know for sure the ROMs themselves are good so
that is worth considering. I will also double check jumper settings
but I'm pretty sure from memory that they are right.
Would the individually grounded sounds work at all if the ROMs were
bad? I'm actually not all that familiar with what the RIGHT sounds
are - I'm just going by the fact that grounding pins 2-5 got me sounds
on each.
firepower
Thanks fior the kind words, I'm trying to help.
Grounding and getting sounds on all 5 pins would suggest your sound
ROM and PIA are both working. Be sure you have the right ROM for the
game, If so turn the game on to attract mode and try grounding pins
on the *CPU connector* pins 1J8-1J12. You should also get 5 sounds
selected from there.
If that works and you do have +5v (or so) there on CPU Pins 1J8-1J12,
I would start to suspect the PIA at IC36 on the CPU or something else
on the CPU. Barry mentioned IC37 previously on this thread. This is
a DIP Resistor Pack that does a "pullup" on the CPU's sound PIA pins
2-8 (PA0-PA6). So you woudl see +5v on pins 4-11 of IC37 when sounds
aren't being "selected" and 0v when they are. Also you need to see
+5v at pin 16 of IC37 Always. WIth power off you should see a resitor
between pins 16 (common) and all pins 4 -11 of the IC with a
multimeter. I think it's 4.7K ohms - but another source says it's a
10K ohm resistor pack. But that exact value doesn't matter much - you
just need to get a value between 5K to 10K ohms on each of those pins
with respect to pin 16..
All the Best. -Richard
Chad
Thanks for your time, so appreciated. I'll start checking on all of
that tonight.
And apologies to bobnatlanta, didn't intend to hijack your thread if I
did, just figured we're having problems in the same area so more
detail is better detail. :)
firepower
It's all good. I was careful to cover off his problem too!
"If it was the other way around, let's say the sound test was
perfect.
But grounding pins 2,3,4,5,7,8 produced NO sounds."
Then you could suspect first the input section of the PIA (unless
that
tested good with the Leon ROM),
the resistor array at SR1 (4.7K ohm 10pin SIP which can go faulty,
and can just be replaced by resistors),
the resitors R2-R10 (100 ohm),
and especially IC5 (4050 Buffer) .
There are also a few 470pF Caps there, if ANY of these are shorted
then the sound lines are selected all the time which means lots of
other sounds can't fire.
All of this can be checked with Multimeter/logic probe. Caps should
show as OPENs on ohms, resistors need to test within about 10% of
value to be connsidered good. So I take 100 ohms to read around
95-105 ohms, give or take a few.
-Richard
Chad
Quick note as much to myself as to everyone else:
-grounding header pins 2-5 on the sound board with CPU not connected:
sounds
-grounding pins through the back of the IDC connector on the CPU
board: no sounds
-grounding pins through the back of the IDC connector on the sound
board with CPU connected: no sounds
-voltages on sound board pins 2-5 with CPU connected: p2-almost
nothing, p3-almost nothing, p4-4.9v, p5-4.9v
Definitely looks to be an issue on the CPU board. I did just replace
the female side of the interboard connector, still have to replace the
male side and the critical headers on the CPU board along with the
sockets. I'm bullet proofing these boards so maybe this issue will
get cleared up in the process. While I'm replacing the other sockets
I can socket the two PIAs for swapping to see if the sounds start
working with the other one in the sound PIA's spot.
Chad
> Definitely looks to be an issue on the CPU board. I did just replace
> the female side of the interboard connector, still have to replace the
> male side and the critical headers on the CPU board along with the
> sockets. I'm bullet proofing these boards so maybe this issue will
> get cleared up in the process. While I'm replacing the other sockets
> I can socket the two PIAs for swapping to see if the sounds start
> working with the other one in the sound PIA's spot.
Quick update:
Replaced CPU side of the 40 pin, all relevant sockets, and the sounds
header. Sound is still acting the same.
Tried to pull the appropriate PIA from the CPU but couldn't manage it
without damaging either the chip or the pads. I'm a pretty good
solderer. Is there a technique or is that just really difficult? No
matter what I did I just couldn't seem to get all of those pads clear
enough to attempt removing the chip without pulling hard enough to
lift pads. I had been hoping to socket that PIA so I could swap in
another one without trashing this one just in case it's still good.
At this point should I try the Leon test rom or just cut the PIA out
and try a new one?
seymour-shabow
> Tried to pull the appropriate PIA from the CPU but couldn't manage it
> without damaging either the chip or the pads. I'm a pretty good
> solderer. Is there a technique or is that just really difficult?
It's difficult, but try adding some solder to the stubborn pins. Do you
have a desoldering gun? Makes chip removal much easier, but unless it's
a hard to find chip I usually cut them out, remove the pins with my
soldering iron, then use the desoldering iron to clean the pads.
> No
> matter what I did I just couldn't seem to get all of those pads clear
> enough to attempt removing the chip without pulling hard enough to
> lift pads. I had been hoping to socket that PIA so I could swap in
> another one without trashing this one just in case it's still good.
>
> At this point should I try the Leon test rom or just cut the PIA out
> and try a new one?
Cut the PIA out if it's already damaged, and put a socket in.
-scott CARGPB#29
Chad
> It's difficult, but try adding some solder to the stubborn pins. Do you
> have a desoldering gun? Makes chip removal much easier, but unless it's
> a hard to find chip I usually cut them out, remove the pins with my
> soldering iron, then use the desoldering iron to clean the pads.
No desoldering gun, just a temp controlled station and soldapult. I
was trying adding a little solder and would get them pretty damn clear
but then I'd pull on the chip a little and it's not moving. So I'd
try them again and still no luck... then it would move a little but
I'd be unable to tell which pin was still not clear, and with so many
pins, it just wasn't happening without damaging the chip or the pads.
I know I can solder at that size, too, since I did just replace every
socket (including the CPU) on that board.
> Cut the PIA out if it's already damaged, and put a socket in.
Well that's the thing. I don't know for sure that it is. Isn't this
an OOP chip? I'd hate to destroy a working one and be wrong if
they're not making these anymore.
firepower
I usually get as mutch the solder off the pins with a temp controlled
station and soldapult. Then I take a smal flat blade screwdriver and
"snap" the pin to the other side of the hole. This usually frees the
pin, you can tell by giving it a wiggle side-to-side, If not, add
back new solder and clear it again.
That's the trick that works for me. When I know all the legs are
clear I put a piece of eraser on the board and pry up first one side
of the chip and then the other.
If that doesn't work, I sacrifice the chip to avois damaging the
board. As Scott said, cut the legs at the top nearest to the chip
top (some say dremel. I've tried that and also use very sharp and
small snips) and then desolder them one by one.
They don't make PIAs but there are lots of types that are available at
a cost, B (2MHz) is faster that A (1 Mz) both work for this era WMS
pinball:
6821 PIA = xx6821, xx68A21, xx68B21,
xx can be MC (Motorola) or HD (Hitachi Data).
also G65SC21P (CMD).
-Richard
Chad
Just for giggles, before I cut out the PIA, I'd like to verify the pot
on the speech board isn't the issue. It says it's a 5k pot but what
should the actual range of values be?
bobnatlanta
sound test works, sounds in game, when they happen, are incorrect.
After about thirty seconds of play, sometmies the timed drop target
sound will kick in when a target is struck, but that is the only thing
remotely correct about the sound. Bonus countdown sound sometimes
kicks in, but it is not the correct pitch, and once it starts, it
doesn't stop. No other sounds or speech as far as I know during game
play.
IC36 replaced and socketed (working static safe)
sound board connectors replaced
ribbon cable continuity verified
caps replaced
ROMS socketed
Sound test works fine
No sound on solenoid test
No sound when pins are grounded
So, I'm now focused on this part of the responses thus far:
>If it was the other way around, let's say the sound test was perfect.
>But grounding pins 2,3,4,5,7,8 produced NO sounds.
>Then you could suspect first the input section of the PIA (unless that
>tested good with the Leon ROM),
Is this the sound card PIA or the CPU PIA?
>the resistor array at SR1 (4.7K ohm 10pin SIP which can go faulty,
>and can just be replaced by resistors)
How exactly, do I do this?
>the resitors R2-R10 (100 ohm),
>and especially IC5 (4050 Buffer) .
OK, this part I get.
>There are also a few 470pF Caps there, if ANY of these are shorted
>then the sound lines are selected all the time which means lots of
>other sounds can't fire.
Don't think this is the problem, all new caps on the sound board.
THANKS for all the help! Over the weekend, I was really close to
trading this thing for an EM project...
Cheers!!
Bob
bobnatlanta
As I study this and the schematic, I realized that all references here
are to components on the sound card, so I assume the PIA reference is
for the sound card as well. As for replacing SR1 with resistors, I
think I get that too, a series of 5 resistors totalling 4.7K ohms
+/-10%, correct? So, 5 1K ohms should do the trick, but what wattage?
Thanks again!
Bob
seymour-shabow
No, there's a PIA on the cpu board that controls the sound. There is
also one on the sound board, but you know that one works because the
built in sound card program fires the sounds correctly (the cpu on the
sound board is controlling the test program, not the main cpu board's cpu)
If you've tried grounding the input pins to the sound board and get
nothing, the input side of the sound board has an issue - if you do get
sounds, the connection between the mpu and the sound board or the
pia/related circuits on the mpu board are the issue.
-scott CARGPB#29
bobnatlanta
>
> - Show quoted text -
So, the input side of the sound board is the issue...what are the next
steps?
Chester_pete
>
> - Show quoted text -
Replace the 6810 - especially if you find the input gates are all
switching OK.
bobnatlanta
> On Sep 11, 6:47 pm,bobnatlanta<rwke...@yahoo.com> wrote:
>
>
>
>
>
> > On Sep 11, 12:20 pm, seymour-shabow <seymour.sha...@gmail.com> wrote:
>
> > >bobnatlantawrote:
>
> - Show quoted text -
You guys are giving me WAY too much credit. Replace 6810, that I can
understand. Inputs gates switching OK...what does that mean?
I'm a monkey see, monkey do electronics guy. :)
Thanks!!
Bob
bobnatlanta
OK, 6810 has been replaced. I socketed it, and worked carefully
static free. No change. Sound test is fine, only a few sounds work
in game play, and they are incorrect when they do work, no speech.
Once again, I throw myself on the mercy and wisdom of RGP. This is
killing me. One of my all time favorites, and I nearly traded this
thing for an EM last week...
Thanks!!
Bob
Chad
Heh! And I'm sitting here working on mine too...
I followed firepower's advice and went barebones. I added in W1 and
set both dips to off. Now the sound test DOES run continuously -
except it only plays one sound! The one sound just loops over and
over. Could that be a ROM issue?
Chad
Turns out it was a ROM issue! New sound ROM, and after all that
troubleshooting, the problem went right away. Now it speaks too!
Chad
Guess I spoke too soon... sound board sound test cycle works properly
as far as I can tell. Speech works. Sounds 1-4 fire in sound test.
No sounds during gameplay other than speech.
Grrrr.
FWIW, it's also not registering ball endings despite all switches in
the trough working in switch test.
New 40 pin, new sockets, headers reflowed, speech board capped, pretty
much every update/upgrade Clay suggests has been done. This one is
making me crazy.