What's new
By:
Filters

CPS-A-01 Custom

T

Tele-Viper

Beginner
Joined
Jun 23, 2017
Messages
17
Reaction score
4
So we're all aware of the high failure rate this chip appears to have on CPS1 A boards.

Has anyone ever heard of one failing on a CPS2 A board?

I can't find any evidence and if so.. I have a CPS2 A board someone used as a parts board in the past, and a bad CPS1 A board. And am wondering if it's a 1:1 switch for the chip?
 
Yes, you can replace the A custom from a CPS1 with a A custom from a CPS2.
I don't think the failure rate is really much higher tho, remember CPS1 was released starting 1988, thus predates CPS2 (1993) by 5 years in some cases.
Only natural IMO we'll see failures on oldest chips first.
 
jassin000 said:
Yes, you can replace the A custom from a CPS1 with a A custom from a CPS2.
I don't think the failure rate is really much higher tho, remember CPS1 was released starting 1988, thus predates CPS2 (1993) by 5 years in some cases.
Only natural IMO we'll see failures on oldest chips first.
Click to expand...
Actually I kinda wonder if Capcom moved production to a different facility for the new platform. Just a minor difference like that could be the whole difference.
 
Tele-Viper said:
Actually I kinda wonder if Capcom moved production to a different facility for the new platform. Just a minor difference like that could be the whole difference.
Click to expand...
I've repaired quite a few cps1 A and C boards with customs from parts cps2 A boards. Havent had a bad pull yet from a cps2 A board. There might be something to the "later production, better quality" theory. Also worth noting is the load on these chips in the cps2 configuration was much less, so stands to reason theres more life left in them for this reason.
 
Knowing all this and going forward. It might work out best for me if I somehow socket on a known good board like a CPS2 A board and use it as a tester.
 
jassin000 said:
Yes, you can replace the A custom from a CPS1 with a A custom from a CPS2.
I don't think the failure rate is really much higher tho, remember CPS1 was released starting 1988, thus predates CPS2 (1993) by 5 years in some cases.
Only natural IMO we'll see failures on oldest chips first.
Click to expand...
CPS-A-01 failures on CPS1 have been very common for 10+ years now, given CPS2 is 5 years younger we should have already seen the same dramatic failure rate.
This isn't the case, so either:
- manufacturing process of those chips has changed between CPS1 and CPS2
- chips are less "stressed" on CPS2: lower temperature, lower load (less data written/read), etc.
 
Honestly. I wonder even more now.

Okay, have wee seen dramatic A-01 failures on Qsound boards? CPS Changer boards?


Anyways. I am going to have to do some science! \m/
 
Tele-Viper said:
Okay, have wee seen dramatic A-01 failures on Qsound boards? CPS Changer boards?
Click to expand...
Don't forget the CPS1.5's, like the Changer yes it seems to have lower failure rates.
But again I maintain some of this is due to the fact CPS2 boards were produced right up until 1999.
So if you have a rev7 CPS2 board, chances are as @Apocalypse pointed out its some 10 years newer than a early CPS1 game like Nemo or Forgotten Worlds.

Given the stress of the arcade environment (ie daily operation of 8 hours or greater often in smoke filled rooms) a decade can make a world of difference.
 
I noticed my CPS A-01 chip on my Strider board when I first got it used to get really hot. Like, it would leave a residual warm area way above normal below the surface of the PCB. I ended up throwing an appropriately sized passive heat sink on it (along with the Motorola 68k...that thing gets put to work on these!) and to this day the board still runs much “cooler” than it did before adding it & I get no graphical errors when playing.

I wonder if these A-01 chips last much longer in CPS2 boards due to the fan installed in that board’s design. Seriously...next time you power up a CPS1 board for a little while, try feeling the area on the solder side where the chip is mounted. You might be surprised how warm it gets.
 
That's some food for thought. I wonder how many of these chips have failed simply due to insufficient cooling...
 
that's a good point, especially considering the CPS Changer and CPS Q-Sound boards have fans (as well as cases to properly direct airflow) as well.
 
twistedsymphony said:
that's a good point, especially considering the CPS Changer and CPS Q-Sound boards have fans
Click to expand...
Uh huh totally... I'm swear to ya'll the failure rate being high among the CPS1 custom must come down to other factors like cooling and age.
I firmly do not believe these (CPS2) chips are somehow "tougher" because they have potentially been made at another factory.

I work in semiconductor fabs, I have for the past 16 years... We get specs from customers, if specs are followed it produces the same die.
When I'm told 550Anstrum of oxide, you better believe that is with in tolerance (+ or - 25A) if its not the chip never leaves my factory.
 
has anybody metered the voltage on the cps2 chip?
i'v seen cps1 bootlegs that had a diode in line with power to one of the customs to cause a small voltdrop.
 
jassin000 said:
if specs are followed
Click to expand...
^that's the key right there.

you're 100% right that "if specs are followed" the chip will come out exactly the same no matter where it's made.

But there are still other possible scenarios
1. maybe the original specs were wrong, causing premature failure in the early chips and they changed the specs for production in later batches?
2. maybe they used a shitty manufacturer for the early chips who DIDN'T properly follow the steps and then they moved to a better manufacturer that did properly follow them.

I mean there's nothing to say that these factors AND the fans AND the amount of stress that the chips experience are all factors that are contributing.
 
Capcom changed the CPS1 A1 customs package. I have a couple of boards that use ceramic packages and they seem to be earlier models. From an engineering point of view replacing the package by plastic makes sense as it is cheaper and has an acceptable mean time before failure. It doesn't dissipate heat as well but for its original intended purpose it's fine.
Cooling the chip is a good idea. In fact, cooling the entire board is best. There's a few components real close to the chip that get quite hot (buffer drivers) and the 68k CPU also gets hot.
So, hum... Maybe add a fan like in the CPS2?
 
I’ve gotten into the habit of taping a copper heatsink to the CPU, sound CPU, sound chip, and any custom chips on all my boards that don’t have active cooling.

I started after I noticed how warm the CPS1 custom gets. I haven’t run any scientific tests but most chips run 5-10°C cooler, (recorded with infrared thermometer) with even small heatsinks. The hotter the component, the more effective the heatsink is in my observations.
 
Has anyone ever decapped a failed CPS1 custom before and SEM'd it?
I suspect the failure would deration the copper or aluminum pathways and vias due to prolonged heat stress.
I've seen aluminum pathways burn clean thur just like a fuse would behave if overloaded.
 
Any links to the heatsyncs you used?

GoldenGlover said:
I ended up throwing an appropriately sized passive heat sink on it (along with the Motorola 68k...that thing gets put to work on these!) and to this day the board still runs much “cooler” than it did before adding it & I get no graphical errors when playing.
Click to expand...
 
You must log in or register to reply here.
Back
Top