What's new

mR_CaESaR

Champion
Joined
Jul 10, 2018
Messages
1,287
Reaction score
1,804
Location
Sydney, Australia
I purchased a CPS3 mobo only from YAJ a little while back in hope that it would work, but unfortunately my YAJ luck has run out (previously most "junk" stuff I've purchased has been perfect, but sadly this time it's not so much the case :()

When the CPS3 arrived, it looked immaculate!

O703HI9l.jpg


I loaded up my simms and DS cart and proceeded to write 3s into the SIMMS. 45 mins or so later, it wrote successfully and here I am thinking we have a winner...

As the Capcom Logo came on and the attract mode started to play, there was no sound. I thought it may have been because the audio setting was configured for external, so I checked that and it was configured for JAMMA, but I then swapped it over to external and plugged in to an external amp still no sound. I played around with the volume pot but there was no crackling or anything like that, cleaned the pot with IPA and still no change. I thought, let me change the volume pot as those seem to be one of the troublesome parts on this system.

After I took the cart off, the whole cart socket also came without, the CPS3 is now without a socket and sound. Great!

2c6T8Cml.jpg


Cleaning this out proved to be quite a time consuming exercise as every time I used the desoldering gun, it didn't work like it does if the pin was properly secured with the cart, so there was a lot of soldering with tweezers and then wicking the area. We got there in the end

J4C5G2Pl.jpg


However, it wasn't without casualties. I didn't realise until I was cleaning that I had accidentally wicked off a resistor.

hkkaslSl.jpg


No love lost as I was able to get it off another board that's broken.

Cart arrived after a few days from ordering at digikey and within minutes, a new cart with a might firmer grip on the cart was installed.

qOL9j0rl.jpg


Information of the CPS3 was used from this page https://github.com/dankan1890/mewui/blob/master/src/mame/drivers/cps3.cpp

kFoIlfEl.jpg


With advice/help from @xodaraP and @cruzlink2, I've managed to do the following:
  • Changed cart (I ordered these ones as they had a metal brackets and not plastic like the original ones - the plastic was the reason the originals broke for me. The plastic mounts can be found on this link)
  • Changed the Sound Amp, Toshiba TA8201AK IC; no change
  • Confirmed the sound amp was receiving voltage; confirmed with 12v with red probe on pin 6 and black probe on pin 4
ut9INV0l.jpg

  • Changed the volume pot; no change
  • Checked to see if it was isolated to jamma edge, changed to external in the settings and sent to external speakers; no change.
  • Checked to see if there's continuity between GND and 12v; no continuity
  • Replaced the LM833N chips; no change
  • Checked the DAC for data, ws, bck and confirmed there's continuity on the resistor array RN3 (pin 6 = dat, pin 5 = ws, pin 4 = bck)
  • Checked the voltages on the DAC based on the following schematic
9hGoA5yl.jpg

  1. Pin 1 = 4.95v
  2. Pin 10 = 4.95v
  3. Pin 22 = 2.85v
  4. Pin 23 = 4.95v
  • Removed the DAC and confirmed every pin had continuity to the next point
078tK6Cl.jpg

  • Replaced the DAC; no change
Everything that's been replaced lol

8Sdv0T9l.jpg


At this point, I am completely out of ideas and I believe I've done pretty much everything I'm capable of doing. My knowledge of PCB repair or the CPS3 is very limited, I don't have a logic probe, oscilloscope to further do any more troubleshooting, so I'm really now at the hope someone else has experienced a very similar issue and was able to fix by changing a part. TIA
 
Last edited:
Your mechanical rework skills look like they have reached the peak of perfection. To be able to work and rework something like this is about as good as it’s possible to get.

The next step on the journey is to get a logic probe and then an oscilloscope, I’m in the same place, just learning how to use them now, and they are like a magic wand. I feel like Harry Potter every time, so satisfying. Had a Mvs-4 that had me stumped for months, less than 5 mins with a logic probe to find the fault (dead 68k which NEVER dies)

You’re in a really good place for this because it sounds like you already have a working CPS3 setup. With a logic probe you could compare the two and find the fault in 5 minutes too.
 
Aprreciate the feedback mate, but I'm definitely beyond my depth. I wish I could say it's the peak of perfection, but I do believe I can get better.

The more I work with this stuff, the more I realise I need a microscope to work on these really fine pins along with a new solder tip..my bevel tip is too big and I need to look for a bent tip.

That being said, this whole logic probe and oscilloscope for troubleshooting sounds awfully tempting. Finding the issue in 5 mins sounds a lot better than blindly replacing parts that "I think" might be the cause.

Edit: BTW.. I've read your cps3 repair thread so many times as a reference for repair work
 
Last edited:
Edit: BTW.. I've read your cps3 repair thread so many times as a reference for repair work
Haha, I had some fun with that one. I’m now sitting on about five grand of CPS3 hardware, should probably sell/trade some of them one day… but let’s be honest, I’ll just sit on them until I die… anyhoo

This was the thread I was talking about for the logic probe and the 68k.
https://www.neo-geo.com/forums/inde...hdog-even-with-neogeo-diagnostic-bios.267802/

Not sure what kind of gear you currently have, but I strongly suggest a 2nd hand binocular analogue microscope, not a digital one. Many digital ones have a lot of lag, and being able to see in 3D makes a huge difference. Only need 10x/20x for this kind of work - if you time it right there are tonnes on the market cheap when the vet/biology/medical students finish for the semester.

This is the logic probe I have, it’s not good, but for 30 year old hardware it’s fine. Technically an oscilloscope removes the need for a logic probe, but the probe is dirt cheap and super easy to use
https://www.amazon.com/Cldkem-Frequ...661636494&sprefix=logic+probe,aps,1025&sr=8-9
 
Last edited:
Haha, I had some fun with that one. I’m now sitting on about five grand of CPS3 hardware, should probably sell/trade some of them one day… but let’s be honest, I’ll just sit on them until I die… anyhoo

Yeah that's all I really want.. a second set to have as a spare in case my current one dies. They're finicky AF

This was the thread I was talking about for the logic probe and the 68k.
https://www.neo-geo.com/forums/inde...hdog-even-with-neogeo-diagnostic-bios.267802/

Nice! I'll have a read through that too.

Not sure what kind of gear you currently have, but I strongly suggest a 2nd hand binocular analogue microscope, not a digital one. Many digital ones have a lot of lag, and being able to see in 3D makes a huge difference. Only need 10x/20x for this kind of work - if you time it right there are tonnes on the market cheap when the vet/biology/medical students finish for the semester.

I have my phone camera to zoom in hahah. I know, it's terrible! :D I watch Northridge Fix all the time on youtube and his setup is something that would be nice to have but it's all very $$$ and I can't see myself being a micro soldering person all the time. Hell I can't even fix this cps3. That said, I'll have to check out the AU market and see what I can find.

This is the logic probe I have, it’s not good, but for 30 year old hardware it’s fine. Technically an oscilloscope removes the need for a logic probe, but the probe is dirt cheap and super easy to use
https://www.amazon.com/Cldkem-Frequ...661636494&sprefix=logic+probe,aps,1025&sr=8-9

I can work with a that logic probe price wise... I just need to learn how to use the damn thing haha
 
A few more things you can check:

1) Use a logic probe to make sure Pin 12 on the TDA1306T DAC is receiving a pulsing clock signal. If you have a scope, it should measure a 14.32 mhz pulse.

2) The DL-3329 SSU SD04-1536 custom chip handles all the audio. Reflow it and check for continuity with every chip leg against the corresponding pad.
 
Just tried reflowing and continuity checks on DL-3329 SSU SD04-1536, unfortunately still no sound :(
 
this is the part I can't check - I don't have or know how to use a logic probe unfortunately :(
 
A logic probe is around $25 from Amazon and I made a tutorial video on how to use one. My video covers using one to debug EPROM activity, but in your case you just need to probe the clock pin of the DAC and see if the probe pulses.

View: https://youtu.be/2PAtTIAijeA
Some awesome advice here, logic probe costs next to nothing and makes the impossible simple.

I had a neogeo with a bad 68k (the 68k never ever ever fails, but this one had). Months of head scratching, 10 mins of watching @ShootTheCore video and a $15 aliexpress logic probe = problem solved.
It really is an amazing tool for next to no money. With fully working CPS3 fetching crazy money now (over a thousand bucks??) its well worth the effort to learn to use one I promise
 
yeah I'll definitely be getting one real soon and hopefully find out where the issue lies
 
Got my logic probe and tested out pin 12 on the DAC

On my 100% working cps3, the logic probe is set to CMOS, when the probe touches Pin 12 on the DAC, it flashes and states LO

View: https://youtu.be/jAf6hMKerzA


On my no audio cps3, when the probe touches Pin 12 on the DAC, it doesn't pulse but states LO

View: https://youtu.be/gRWscb19Gpc


Does this mean the DAC is faulty? I've already tested continuity from pin 12 to its next spot that I could tell and there's continuity there. Should I just purchase a new DAC instead of using the another DAC from a different board which I knew had sound working (albeit crackling) but wasn't 100% working due to graphics glitching.
 
You'll want to set the logic probe switch to TTL instead of CMOS, but I expect the outcome will be the same in this case.

The pulsing on the working board means that the DAC is getting the expected clock signal it needs to function. The lack of pulsing on the quiet board means that it isn't receiving a clock signal. Now we know why your board is quiet - without a clock signal the DAC won't function. Without a DAC, you get no audio.

The next step is to figure out why the DAC isn't getting a clock pulse. Inspect the trace that feeds that pin and make sure it isn't damaged. Also follow the clock trace to its destination and tell me what chip is feeding it - my CPS3 board is in a case so I can't easily find out for myself right now.
 
When I changed to TTL, pin 12 on the quiet one just has a constant beep on "LO", pin 12 on the working one just flashed but no light at all or sound on HI or LO

LB3FGzT.jpg


Went through the traces and found the following

- pin 12 goes to one of the points on RN3 (orange). Both points had continuity
- point on RN3 went to pin 9 (red) on 6D which had cu383 on the chip - I'm assuming some sort of bugger chip?

There are no visible damages that I can tell on the board, especially in the area where the amp, DAC, this is possible the cleanest cps3 I've seen.

Edit did some more testing:

- Working board, DAC pin 12 = pulse, RN3 orange side = pulse, red side = pulse, 6D buffer Bus Exchange Switch pin 9 = pulse, pin 10 = LO, pin 11, = LO, pin 12 = LO
- Quiet board, DAC pin 12 = LO, RN3 orange side = LO, red side = LO, 6D buffer Bus Exchange Switch pin 9 = LO, pin 10 = nothing, pin 11 = LO, pin 12 = LO

Is it possible that the buffer chip Bus Exchange Switch has died? I suppose the only real way I can find out is if I get a buffer chip Bus Exchange Switch out of a donor board and test it out myself.

edit 2: data sheet information about PI5C3383

1665526347777.png
 
Last edited:
Yep, I’m suspicious of the 6D chip as well - RN3 is just a resistor array and you’re not reading a clock pulse going into it anyway.

The 6D chip should be a standard part that you can order from a supplier - you shouldn’t have to source one from a donor. Can you post a part number and/or a closeup pic of it? I can’t tell what it is on your pic other than that it’s a TI part.
 
Thanks.

I'll probably suss out the donor board first as I don't want to make a 60 dollar order just yet LOL

Edit: well looks like it's only TTSOP24 that I'll be ordering, when filtering, these 4 are the results... save to say, I won't be buying a real of 2000 for the SSOP-24 ones :D

1665147744635.png
 
Last edited:
Ok after speaking to Julius today I couldn't help but take a look at this as I was curious.

Checking the data sheet for PI5C3383 IC

Pin 24 and 12 are your supply (5V)
Pin 1 is 'bus enable' which is active low
Pin 13 is 'bus exchange' which switches a pair of inputs/outputs depending whether it is high/low (as per truth table in datasheet)

The DAC clock pin is connected to Pin 9 which is output D1, the input to D1 is either Pin 7 (A1) or Pin 8 (B1). So providing Pin1 remains low the changing of state for Pin 13 will switch the input to D1.

Basically:
Pin 1 Low, Pin 13 Low = B1 > D1
Pin 1 Low, Pin 13 High = A1 > D1

Tested on a working board both pins 1 & 13 were low so B1 > D1 and as such pin 8 was active (it's connected to pin 9 in this state so to be expected). I had a quick look with a scope the frequency of the pulses was just over 14Mhz, I didn't do much else at this point as its pretty late now. For reference Pin 7 (A1) was also active at a different frequency but I didn't pay much more attention to that as the IC was in a steady state and so it wasn't switching A1 > D1.

I would of thought if this chip was dead other things would be impacted as other input/outputs are connected. Regardless easy checks you can do with what you have on hand are the supply voltage across pins 24 & 12, the logic of pins 1 & 13 and for activity on Pin 8 (I suspect there is none) and the fault is going to be upstream. Essentially Pin 8 active? then look at the IC, Pin 8 inactive find what's connected to it.

If you want I have one of those kit scopes "DSO138" you can have if you pay for Auspost.

Edit: Just looking at the mame source as unable to turn brain off
TDA1306T - Philips TDA1306T Noise Shaping Filter DAC (SOIC24). The clock (on pin 12) measures
14.3181667MHz (42.9545/3)

Well that may be your smoking gun, seems like pin 8 probably goes to that 42.9545 crystal oscillator
 
Last edited:
Back
Top