What's new
By:
Pinball info

Register a free account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members through your own private inbox!

Centaur Speech Call Outs Missing- fixed the fecker

Poibug

Poibug

Registered
10 Years
Joined
Aug 10, 2011
Messages
3,307
Location
flintshire
So, I really aint great with sound / speech boards so any ideas.
I had horrible mains type hum on my Centaur ever since I got it.
Recently set it up again after a year or so and that had got even worse -but all sound/speech was ok - ie all correct just the horrible noise.
So, I finally got off my **** and recappped the vast majority of the board, with primarily, Vishay caps from CPC.

Powered up ok, and the mains type hum has now gone. But I am missing a fair few of the speech call outs.
Also they are incorrect when the flipper buttons are pressed in attract, when its says how to play the game.

Note, it the red test button is pressed, and the callouts are all there,I believe.

Does that scream out anything obvious to anyone ie the TMS5200 could be bad, or a PIA etc?
The TMS did ahve some corrosion that I cleaned off with a fibreglass pencil
It has @myPinballs aftermarket say it again as opposed to the original, but this shouldn't be an issue, as that side is all good - reverbs great, as it has for all the time since i bought it off him, and mpu has always been an Alltek

TIA

Poibug
 
Last edited:
Is it just speech calls missing /wrong or is it sounds aswell?

were they all ok before you worked on it, or did it play up before?
 
All good before
Think sound is ok.
Just speech wrong during game and attract, but all ok with red button test
 
Have you double checked your j1 connector in case you dislodged something? Especially the pins 1-4. The data ‘word’ sent from the cpu to the sound board that tells it what to play is sent via j1 and the bits for it are pins 1-4
 
Sorry to derail a bit but ... on my SI I have a original sound board and it works perfectly but I wanted to have Jim’s after market sound and echo board in it to , to make it have the echo . As soon as I fitted it in ,it would not work properly , los of sound missing speech etc etc . Knowing Jim’s work awesomeness the fault will not be on his board soI looked at the sol board for the fault ,the 2 caps on my board had not been replaced so I replaced C23 and now it works perfectly . This is a problem I had with my machine,it does sound a bit like it ,hope I have not derailed the chain of thought by anyone
 
carl lawrence said:
Sorry to derail a bit but ... on my SI I have a original sound board and it works perfectly but I wanted to have Jim’s after market sound and echo board in it to , to make it have the echo . As soon as I fitted it in ,it would not work properly , los of sound missing speech etc etc . Knowing Jim’s work awesomeness the fault will not be on his board soI looked at the sol board for the fault ,the 2 caps on my board had not been replaced so I replaced C23 and now it works perfectly . This is a problem I had with my machine,it does sound a bit like it ,hope I have not derailed the chain of thought by anyone
Click to expand...

Poibug mentions the speech is correct when he presses the test button so it seems unlikely the 12v filter cap is upsetting the newly recapped board, but its worth ruling out anyway like with the j1 connector. I think this is a fairly rare issue, though i have seen it cause an altek not to boot aswell, so always worth just changing for reliability generally.

Another thought i have this morning is to swap the 2 PIA chips over. The inputted data word is fed directly (via a 74 invertor) on centaur to the pia u11, as the EE jumpers are in place and so the sound gen chip that normally buffers them isnt used. By swapping them over we can see if anything changes. Also if there is a spare pia kicking about, swap U11 to.

could also be an issue with the invertor before the pia but that needs testing with a logic probe and isnt socketed
 
Last edited:
Don't have a cap here to change c23 on power board.

New repinned connector done for squawk J1. And header pins resoldered.
I then interchanged the 6821s, and reseated all roms and the ram.
It was the same.
I then left it on but left the game for a fee hours.
Then all was ok. Which made no sense.
Played great for a few hours.
Yet this morning, playing up again, but in a different way.
Now in attract mode, with flipper pressed, no speech is heard.
Now during gameplay some speech is heard, but not all e.g. no enjoy centaur at the start just a momentary noise.The sort of motorbike noise is no longer heard.
Plus some multiplier callouts are incorrect/duplicated.2x says 3x, 3x says 3x, 4x says 5x, 5x says 5x
But knocking downs orbs does say Activate Power Orb.
And at end of game it's incorrect as it repeats the phrase
Power,power,power orb.

I just tried new 6821s, and also a different 6809, but no different.

****ing me off


TIA
 
Poibug said:
Don't have a cap here to change c23 on power board.

New repinned connector done for squawk J1. And header pins resoldered.
I then interchanged the 6821s, and reseated all roms and the ram.
It was the same.
I then left it on but left the game for a fee hours.
Then all was ok. Which made no sense.
Played great for a few hours.
Yet this morning, playing up again, but in a different way.
Now in attract mode, with flipper pressed, no speech is heard.
Now during gameplay some speech is heard, but not all e.g. no enjoy centaur at the start just a momentary noise.The sort of motorbike noise is no longer heard.
Plus some multiplier callouts are incorrect/duplicated.2x says 3x, 3x says 3x, 4x says 5x, 5x says 5x
But knocking downs orbs does say Activate Power Orb.
And at end of game it's incorrect as it repeats the phrase
Power,power,power orb.

I just tried new 6821s, and also a different 6809, but no different.

****ing me off


TIA
Click to expand...

What you are describing isnt as random as you might think. Can you confirm that when you press the test button the speech still plays in the correct order and that you dont have repeats here.

If so then i am leaning towards the inverter chip at the start of the data word input. This is U16 i believe. It will be a 4000 series chip on it, but you can replace with a 74HCT04 or 74HC04. Ideally if you have a logic probe you would check the pins to see whether a line is stuck high or low when the cpu sends a command to play something, but as you mention similar phrases in the sequence are repeating like 2x says 3x etc, it really sounds like one of the outputs of this chip is stuck high or low

After you leave it on for awhile this chip might warm up and start working ok so its not totally dead. Also it might be failing internally in different ways each time.

So the best thing now after swapping pias is to change it to rule it out

Thats my advice anyway. :)
 
myPinballs said:
What you are describing isnt as random as you might think. Can you confirm that when you press the test button the speech still plays in the correct order and that you dont have repeats here.

If so then i am leaning towards the inverter chip at the start of the data word input. This is U16 i believe. It will be a 4000 series chip on it, but you can replace with a 74HCT04 or 74HC04. Ideally if you have a logic probe you would check the pins to see whether a line is stuck high or low when the cpu sends a command to play something, but as you mention similar phrases in the sequence are repeating like 2x says 3x etc, it really sounds like one of the outputs of this chip is stuck high or low

After you leave it on for awhile this chip might warm up and start working ok so its not totally dead. Also it might be failing internally in different ways each time.

So the best thing now after swapping pias is to change it to rule it out

Thats my advice anyway. :)
Click to expand...


Cheers guys.
Cleaned all socket chips on the board.

Jim, the vid I posted today,is of the test. I cannot remember if that is as it should be.
I may just shotgun and change the 4049, at u16as there is logic, pardon the pun,in that.
Assuming I have one here

Cheers

Poibug
 
Didn't have U16 here, so put in a chip holder and ordered some chips, hopefully get em before bank holiday
 
Interestingly enough, i just fixed a s&t board for a customer. It was a non booter, with digital section issues. However after getting it back working again it played fine for 5-10 mins then all the sounds disappeared even though the test button still worked and played the speech in order. Replacing u16 fixed this.
 
myPinballs said:
Interestingly enough, i just fixed a s&t board for a customer. It was a non booter, with digital section issues. However after getting it back working again it played fine for 5-10 mins then all the sounds disappeared even though the test button still worked and played the speech in order. Replacing u16 fixed this.
Click to expand...

Hopefully same for me.
Hope the chips come Saturday
Cheers
 
No go.
Replaced U16, but it's as it was.
I appear to only have , I think 4 flashes.
Which, if I read the flashes right , thats the speech chip.
But using the built in test that's working fine.
Any ideas guys
Cheers
 
Poibug said:
No go.
Replaced U16, but it's as it was.
I appear to only have , I think 4 flashes.
Which, if I read the flashes right , thats the speech chip.
But using the built in test that's working fine.
Any ideas guys
Cheers
Click to expand...

You only get 4 ‘official’ flashes on centaur as centaur doesn’t use the sound gen chip. EE jumpers are in place

need to think more about this problem for more ideas, hmmmm
 
Few more basic things to rule out of the top of my head

+5v gen on board is stable
swap cpu and at same time change to 6802 instead of 6808 and separate ram
Change rom set - will require rejumpering to use 2732 roms
 
You might want to replace the socket for the vocoder too if that's not been done.
 
stevebm1 said:
I had a similar issue with my flash gordon,wrong call outs and speech cutting off,that turned out to be a stuck switch,causing the bizzare behaviour
Click to expand...

now that sounds interesting, worth a quick check using the switch test

it could turn out to be something unrelated to the cap change or the sound board itself.
 
5volts tested ok both plus and minus at test points 2 and 4.
TP3 is high at about 16v.
Also replaced 6808 for a 6802 which allowed removal of the 6810, still same, and yes I changed jumper L to K.
All main caps replaced on this board.also big cap on solenoid power board.
6821s replaced for new spares, still same. Even interchange with cpu ones
 
I get some phrases
1234 says power lane
Orbs say activate power orb
Rollover lanes at top when rolled over lit switch still correctly says slow arnt you

2x still wrong saying 3x, and 4x still says 5x


I run a switch test with balls removed and there are none stuck.
Correct when ball switches checked with balls in.

Stuck for ideas now as sound test still good so can see it being roms.
My only thought,somehow is the 5200 as it was corroded, and two pins snapped off, I had to jumper to board
 
Poibug said:
I get some phrases
1234 says power lane
Orbs say activate power orb
Rollover lanes at top when rolled over lit switch still correctly says slow arnt you

2x still wrong saying 3x, and 4x still says 5x


I run a switch test with balls removed and there are none stuck.
Correct when ball switches checked with balls in.

Stuck for ideas now as sound test still good so can see it being roms.
My only thought,somehow is the 5200 as it was corroded, and two pins snapped off, I had to jumper to board
Click to expand...

I can send you a used 5200 to try if you want. I have a few old ones with pitch issues that are no use to me but work fine otherwise

i doubt it will help though because the speech chip creates what it’s told to using its lookup tables. The program of what to create is in the roms hence my change the roms suggestion
 
im surprised no one has designed a board that emulates the tms 5200 or allows for custom sounds/speech to be added to the original roms for these classic bally games
 
Had another thought.
I could use my sound tester software to work out the sound ids for 2x to 5x . If I’m right and they are in order then I should be able to tell what’s common with the address binary when 2x plays 3x etc and therefore may tell us where to look address/data wise

I think the fact 1 speech calls another is key to solving this
 
You must log in or register to reply here.
Back
Top Bottom