F-14 CPU in High Speed, sound problems

Trying to fix up a F-14 CPU to go in a High Speed. Clay's guide indicates it should work, the original board took a pretty heavy alkaline bath and just isn't economically repairable.

It came in as a bit of a project, and I've fixed everything but this sound issue. Added the missing 4 pin connector at 1J15.

The high speed background sound board is 100% tested good with another working HS CPU. As is the ribbon cable, and the little 4 pin connector that jumps the 2 boards together.

In sound test, the game makes no sounds. Even at max volume, I don't hear any faint sounds.
In music test, the game makes all the music just fine.

ROMs are good. I tried 2 sets, and tested them both in the other HS MPU.

I tried changing out the procs for the hell of it, they work fine. U24 is definitely running, and I see activity on all the address and data lines.

Lots of socketed ICs on this board. That's good for troubleshooting, yet also bad for troubleshooting, because you have to question the quality of previous rework.

I've buzzed out the sockets and replaced a few that looked crappy.

Both 1458 op amps were socketed, so I tried 2 good ones, and it made no difference.

Putting my finger on U1 makes the hum I'd expect it to make, I believe the amp is working fine.

I put in Leon's sound test ROM. It makes the heartbeat sound as expected, and the PIA outputs dance in rhythm as expected, but the sound test button does not cause it to reset. This points at the RAM. I replaced that socket, as it was a big dodgy, and buzzed it out. Tested the 5116 in the socket, and the neoloch tester says it's fine.

So, I'm trying to figure out where to go next.

U6, U7, U8 and U9 are all socketed. I have tested U6-U8 with my IC tester, and tried a different PIA in U9. I also buzzed out all 4 sockets and everything is well.

I took a break for awhile, will get back to it later. Any ideas?

#2 9 years ago

DAV

wayout440

Pinside member

I'd start at U3 - the CVSD IC. Should be getting digital in and clock and spitting out analog audio (consisting of sound effects & voices) over to the sound/speech mixer. I have not looked at this circuit in a while in detail, so this is just a suggestion. U4 is the first output after the CVSD and there is also a DC blocking cap at C9, so you might take a scope and look for audio signals at one side of C9 and at U4 output pin 1. I think you should see some signals at these points when sounds are being generated.

cvsd.jpg

#3 9 years ago

JWJ

johnwartjr

So, I dusted my scope off tonight.

I've got a Wittig Oszifox, and I'll be the first to admit that I don't have much experience with it.

I hooked it up to +12 and ground, and fired the game up. Put it in sound test.

Probed a few address and data lines on various PIAs and processors, and confirmed the line in the display of the scope moved occasionally.

Probed C9 as you suggested - on both sides. Nothing, just a flat line.

At this point, I checked all the pins of U3. Completely dead. Both with my logic probe and the scope.

Pin 9, clock, is not moving. I got the schematic out, and it connects to U9-39, CA2, which is not moving either.

#4 9 years ago

DAV

wayout440

Pinside member

OK, John here's my thoughts. On F-14, I double checked this with my scope. As expected, I see some audio signals on C9, and moving over to U3 checked clock pin 9 and digital input pin 12. There I see varying pulse widths depending on which sound effect is being produced. (note that on some sounds it goes away completely)

I can only *assume* that this is the same on High Speed, but I don't have that game anymore to confirm. So I am only say, about 90% certain that there should be something going on there. Maybe someone with high speed could refine this.

We could assume that U3 is completely dead - dragging down its signals. A check that could be done: isolate U9-39 by lifting that pin (easier to do in socket) or make a cut in its trace. Then without U3 affecting it, you could see if data is being spit out of U9. Another option would be to shotgun replace U3 and hope for the best. I would certainly get the symptoms you describe on F-14 if U3 was dead.

These are my best guesses at this time, sorry if it is of limited help. Hopefully, this is on to something.

#5 9 years ago

JWJ

johnwartjr

Since I had another CVSD, I carefully removed the original, put in a socket, and installed my new CVSD. No difference.

How about the DAC? Is that a common problem? Ed Cheung explained how to test one in a RGP post - pull the sound ROMs, and ground pins 4-12, and each pin it should get a louder click - and a couple of the pins are dead when grounded, and the rest are about the same volume.

I don't have a DAC on hand, but I've got a couple parts boards I could pull one from - but I hate to shotgun anything...

#6 9 years ago

DAV

wayout440

Pinside member

Unfortunately I personally haven't played around with the DAC...so I can't provide suggestions there. I've read online that the DAC produces some of the sampled sounds that are stored on the ROMs, and the CVSD handles *all* speech.

I just wish someone with an HS would be able to poke around and provide some reference for you.

#7 9 years ago

JWJ

johnwartjr

Yeah, and I shipped out the repaired and working HS board I had to its owner a couple days ago

It should be similar for other games as well, so I imagine your experience is accurate.

I need to get my S11 fixture board going, it's rough, but it's repairable. Perfect for a test fixture!

#8 9 years ago

JEF

helicopter

Pinside member

St Marys, OH

10y 36,900 242 2 10

Hey John, I can ship my HS CPU back to you on Monday if you really need it. I'll be out of town most of next week anyway. It works perfectly now. Or, I can probe anywhere you want me to for pulse confirmation. Sorry, I don't have access to a scope anymore, just a logic probe.

#9 9 years ago

JWJ

johnwartjr

I made a little more progress, I think.

Noticed that sw1 and sw2 aren't registering.

Mechanically, they work. With a meter across the NC terminals, the switches do open up when you press them - but the switches never register.

#10 9 years ago

DAV

wayout440

Pinside member

OP
Quoted from johnwartjr:
I made a little more progress, I think.
Noticed that sw1 and sw2 aren't registering.
Mechanically, they work. With a meter across the NC terminals, the switches do open up when you press them - but the switches never register.

Looking at the schematic, the test switches go to NMI (non-maskable interrupts) active low. Looking at the circuit I would expect these switches to be NO (normally open) and then close when momentarily pressed - thus starting the self-test routines. Maybe I am missing something. These are also going to two different processors. Curious.

#11 9 years ago

DAV

wayout440

Pinside member

Quoted from wayout440:
Looking at the schematic, the test switches go to NMI (non-maskable interrupts) active low. Looking at the circuit I would expect these switches to be NO (normally open) and then close when momentarily pressed - thus starting the self-test routines. Maybe I am missing something. These are also going to two different processors. Curious.

<edit>
Since I had a moment I went back to check. The switches are double pole-double throw, so momentarily pressing them in pulls NMI low to ground (the side of the switch that closes is the right side sets of contacts)

Pushing in the top switch on a working game give to continuous alternating tones (I assume one of those tones indicates the CVSD is working, and the other indicates the DAC is working - but I don't know which tone is which)

Pushing in the bottom switch basically starts the self test, on a working game you get the "all-is-good" bong tone.

If the switch is not functioning, I assume you could also momentarily short the non-maskable interrupt pin to ground with a clip lead on each cpu to simulate the switch action.

#12 9 years ago

JWJ

johnwartjr

Well, I needed a break.

Saturday night, took the family down to the Louisville Arcade Expo for a few hours. Didn't go in the shop all night.

Tonight, watched a movie with my wife, and then went to the shop, and finished up my High Speed fixture 'reference' board - a board that came with my test fixture that was beat to death

hs_before.jpg

Gave it a good cleaning yesterday, let it dry out real well, and populated the missing bits tonight,as well as replacing the resistor spaghetti in the special solenoid section.

IMG_5801.JPG

So, tomorrow night, I'll get back to the F-14 CPU and see if I can figure this out!

#13 9 years ago

JWJ

johnwartjr

https://www.youtube.com/watch?feature=player_detailpage&v=GfPg5LjGYz8

Fixed!

SR8, a 1k ohm bussed resistor, was broken between pins 5 and 6.

In typical Sys11 fashion, you couldn't tell by looking at it.

After replacing SR8, the sound started working.

#14 9 years ago

DAV

wayout440

Pinside member

OP
Quoted from johnwartjr:
Fixed!
SR8, a 1k ohm bussed resistor, was broken between pins 5 and 6.
In typical Sys11 fashion, you couldn't tell by looking at it.
After replacing SR8, the sound started working.
https://www.youtube.com/watch?feature=player_detailpage&v=GfPg5LjGYz8

Great job, glad you found it!

#15 9 years ago

KLB

kbliznick

Firebird Pinball

Phoenix, AZ

11y 63,900 2,438 5 4

OP
Quoted from johnwartjr:
Fixed!
SR8, a 1k ohm bussed resistor, was broken between pins 5 and 6.
In typical Sys11 fashion, you couldn't tell by looking at it.
After replacing SR8, the sound started working.
https://www.youtube.com/watch?feature=player_detailpage&v=GfPg5LjGYz8

Glad I didn't notice your facebook message until now. Would never have gotten around to tellin gyou to check the resistor network.

After 10 years of doing a ton on work on system 11 MPU's I only recently ran into bad resistor networks on system 11's (beside the battery corroded ones). Twice actually. most recently on the lamp matrix and it took me way to long to figure it out.

#16 9 years ago

JWJ

johnwartjr

I've started having problems especially with the resistor network above the lamp column resistors. Think it's getting cooked.

#17 9 years ago

ROB

Borygard

LockWhenLit.com

St Louis, MO

11y 90,050 2,915

OP
Quoted from johnwartjr:
Since I had another CVSD, I carefully removed the original, put in a socket, and installed my new CVSD. No difference...

Glad you got it fixed John.

In my experience it's not a good idea to socket analog parts, particularly the CVSD. I've had to remove sockets and reinstall the chip directly into the board on quite a few WPC sound boards over the years because of issues. Just an FYI.

--
Rob Anthony
Pinball Classics
http://LockWhenLit.com
Quality Board Work - In Home Service
borygard at gmail dot com

#18 9 years ago

KLB

kbliznick

Firebird Pinball

Phoenix, AZ

11y 63,900 2,438 5 4

Quoted from Borygard:
Glad you got it fixed John.
In my experience it's not a good idea to socket analog parts, particularly the CVSD. I've had to remove sockets and reinstall the chip directly into the board on quite a few WPC sound boards over the years because of issues. Just an FYI.

I've noticed that the Harris chips do not like to sit very well in machine pin sockets.

#19 9 years ago

DAV

wayout440

Pinside member

We tend to socket only programmable firmware ICs in our commercial products. In fact, we have product in an early production run that has a flakey firmware PLCC package socket - we remove the socket and solder the IC directly to the board for reliable performance.

#20 9 years ago

JWJ

johnwartjr