Q&A Thread

[channelmaniac]

I have just purchased a Neo Geo Arcade and Bust A Move.  I was curious on how one may reset the high scores.

Flip dip switch 1, power it up, and go into the hardware tests. There you will find a spot where you can press multiple buttons to reset the backup RAM.

Don't forget to turn off dip switch 1 when you power cycle the cab to play the game again!

[Sprout]

Appreciate the assistance; the 82S123 adapter seems to work properly, but no luck with the 82S126; tested with various 2716 and 2732 EPROM's; 200ns for the latter - no luck.  Going to go over the adapter yet again to verify that I didn't screw something up in the process, but it just doesn't seem to like it for that particular usage, and may well be speed-related.  I'm likely going to bite the bullet and just spend the ~$150 for PROM adapter used by my burner, and pick up a few blanks.

Thanks again!

Brent

The 2716 is WAAAAAAAY slower than a PROM.  We're talking 50-70ns vs 300-600ns in speed.

However, the circuit you are using it on is pretty slow. To be sure you'd want to check it with a calibrated oscilloscope or logic analyzer.

Now... The 82S23/123 chips have a single /CE line (the / means active when at logic low) and the 82S126/129 chips have TWO /CE lines. The chip goes active when both are low.

On the 2716 you have 2 enable lines, a /CE and a /OE (chip enable and output enable) and both must be low for the chip to go active. Some data sheets call the /CE line (pin 18) "PD" and the /OE line (pin 20) "/CS" so read the data sheets carefully. In reading the Hitachi 2716 data sheet it appears that if you tie /CE low and use /OE to toggle the chip for activity that the cycle time is far less.

Now each of those chips have varying numbers of address and data lines...
2716: 11 Address lines and 8 Data lines
82S23/123: 5 Address lines and 8 Data lines
82S126/129: 8 Address lines and 4 Data lines

Extra upper Address lines must be grounded. You could use the upper address lines and connect them to switches to switch between banks for different games.

For example: 82S23 uses 5 address lines - A0 through A4. On the 2716 you could use A6 and A5 to toggle through different "banks" or versions of code.
00 - first bank
01 - second bank
10 - third bank
11 - 4th bank

Where the first digit is A6, the 2nd is A5, and 0=ground, and 1 = 5v.

If you leave the address lines floating you are asking for trouble. Same thing if you leave the chip select or chip enable lines floating. Some chips will internally have them tied high or low, but you can't trust all manufacturers to do that or to do it the same way (high or low)

Now... Looking at the schematics for Pac Man - 4A has both /CS lines grounded. The chip is always active. The chip at 7F is used the same way. It's select line is grounded too.

Enjoy!

[channelmaniac]

Appreciate the assistance; the 82S123 adapter seems to work properly, but no luck with the 82S126; tested with various 2716 and 2732 EPROM's; 200ns for the latter - no luck.  Going to go over the adapter yet again to verify that I didn't screw something up in the process, but it just doesn't seem to like it for that particular usage, and may well be speed-related.  I'm likely going to bite the bullet and just spend the ~$150 for PROM adapter used by my burner, and pick up a few blanks.

Thanks again!

Brent

The 82S126 is Open Collector output while the 2716 will be a Tri-State output. You can't always sub one type for the other in a circuit. Sometimes you have to modify things.

[Sprout]

Wanted to follow-up in case others look to do something similar...

Short answer - when wiring data and address lines, make sure you match them up - especially when the datasheets don't use the same structure...  In this case, I had D0 through D1 on one datasheet, and D1-D4 on another.  Basically, all four data lines were off by one.  The address lines *did* match, so just a matter of paying attention 

Re-wired those four lines, and it seems to be working pretty well.  Some slight graphics glitches here and there, but definitely very close to where it should be.

Regards,

Brent

Appreciate the assistance; the 82S123 adapter seems to work properly, but no luck with the 82S126; tested with various 2716 and 2732 EPROM's; 200ns for the latter - no luck.  Going to go over the adapter yet again to verify that I didn't screw something up in the process, but it just doesn't seem to like it for that particular usage, and may well be speed-related.  I'm likely going to bite the bullet and just spend the ~$150 for PROM adapter used by my burner, and pick up a few blanks.

Thanks again!

Brent

The 82S126 is Open Collector output while the 2716 will be a Tri-State output. You can't always sub one type for the other in a circuit. Sometimes you have to modify things.

Appreciate the assistance; the 82S123 adapter seems to work properly, but no luck with the 82S126; tested with various 2716 and 2732 EPROM's; 200ns for the latter - no luck.  Going to go over the adapter yet again to verify that I didn't screw something up in the process, but it just doesn't seem to like it for that particular usage, and may well be speed-related.  I'm likely going to bite the bullet and just spend the ~$150 for PROM adapter used by my burner, and pick up a few blanks.

Thanks again!

Brent

The 2716 is WAAAAAAAY slower than a PROM.  We're talking 50-70ns vs 300-600ns in speed.

However, the circuit you are using it on is pretty slow. To be sure you'd want to check it with a calibrated oscilloscope or logic analyzer.

Now... The 82S23/123 chips have a single /CE line (the / means active when at logic low) and the 82S126/129 chips have TWO /CE lines. The chip goes active when both are low.

On the 2716 you have 2 enable lines, a /CE and a /OE (chip enable and output enable) and both must be low for the chip to go active. Some data sheets call the /CE line (pin 18) "PD" and the /OE line (pin 20) "/CS" so read the data sheets carefully. In reading the Hitachi 2716 data sheet it appears that if you tie /CE low and use /OE to toggle the chip for activity that the cycle time is far less.

Now each of those chips have varying numbers of address and data lines...
2716: 11 Address lines and 8 Data lines
82S23/123: 5 Address lines and 8 Data lines
82S126/129: 8 Address lines and 4 Data lines

Extra upper Address lines must be grounded. You could use the upper address lines and connect them to switches to switch between banks for different games.

For example: 82S23 uses 5 address lines - A0 through A4. On the 2716 you could use A6 and A5 to toggle through different "banks" or versions of code.
00 - first bank
01 - second bank
10 - third bank
11 - 4th bank

Where the first digit is A6, the 2nd is A5, and 0=ground, and 1 = 5v.

If you leave the address lines floating you are asking for trouble. Same thing if you leave the chip select or chip enable lines floating. Some chips will internally have them tied high or low, but you can't trust all manufacturers to do that or to do it the same way (high or low)

Now... Looking at the schematics for Pac Man - 4A has both /CS lines grounded. The chip is always active. The chip at 7F is used the same way. It's select line is grounded too.

Enjoy!

[channelmaniac]

Congrats!

Yeah, those pesky data sheets. Some manufacturers start with D0 and others with D1 and it could bite ya!

RJ

[tjc02002]

you mentioned that "These games ARE repairable as long as the CPU is OK". Do you know how to do this, or even a guess at where to start? Thank you for all the awesome info!!

[channelmaniac]

you mentioned that "These games ARE repairable as long as the CPU is OK". Do you know how to do this, or even a guess at where to start? Thank you for all the awesome info!!

It all depends on the symptoms. If the CPU is finger blistering hot then it's internally shorted. If it's not generating the proper signals on the address/data busses then either the clock is bad or the CPU is bad. There really isn't any other chip that would do that on those boards as they are very stripped down.

Raymond

[simioni]

hi Mr Channel  im from brazil, hard to find any service repair here 

Can u help me solve this problem with my neo geo aes version 3-6 pls?

i didnt found any traces on board, ived looked for 3 days and nothing ...


heres the video: <a href="http://www.youtube.com/v/_w6o2mC4M9M&rel=0" target="_blank">http://www.youtube.com/v/_w6o2mC4M9M&rel=0</a>

and without cart, blue screen appear, sounds ok ...


cleaned all board, pins, chips and no corrosion, and after 10 min playing starts constantly reboots

Do you u think it could be the LSPC2 , capacitor or bios?
thx

[channelmaniac]

Looks like a problem with an address line on one of the graphics ROMs. Check the cartridge connection for dirty/corroded/bent pins.

Open the cart and clean the edge connections with a pink pencil eraser and clean the cartridge slot. To clean the slot, take some white posterboard and bend it in half once or twice until it's as thick as the cartridge board then put some alcohol on it. Insert/remove it repeatedly several times. That will wipe all the gunk off the cartridge connectors.

If that's not it then you'll have to use a digital multimeter to check the address pin connections from the cartridge slot connectors over to the various chips they go to.

Raymond

[channelmaniac]

And not a problem with the wrong place... I can move the posts around as needed.

[simioni]

thx for quickly reply Ray,

ya i think needed one multimeter, lets see .... cya 

[anotherit]

I just purchased my first machine and I'm having a fun time restoring it.  It is a Double Dragon machine with all the original hardware.  The current problem I am having with it is that the board only runs when I have the power supply running at +6VDC and let the machine run for a few minutes.  Under these conditions the machine starts up perfectly with absolutely no errors.  Any less and it gives me a ROM 3 Error and then has all sorts of sketchy behavior.  The most common problem with the ROM 3 Error is that it will freeze shortly into the Attract mode but the music will keep playing.  I pulled all the ICs and checked for any bent pins and I found one on the sound board, which really doesn't help my problem.  The board looks to be in excellent condition with no broken leads, traces, or leaking caps.  Other than the dust it looks brand new.

What should my next step be?  I plan on turning this machine into a street fighter machine in the future so I'm planning on replacing the power supply with something a little more trust worthy.  Should I go ahead with that upgrade and then spend the time chasing voltages around the board or could this be as simple as replacing a single bad component on the board?

[channelmaniac]

6v will fry chips.

I'd start by checking the voltage levels at the ROMs on the game board. If you are checking a 28 pin EPROM then pin 14 is ground and pin 28 is Vcc (+5v)... If it's over 5.25 then lower it. If it's close to 5v yet the power supply is cranked to 6v then you have some serious losses - either thin power cables or bad connections somewhere such as the edge connector.

As for the ROM 3 error, if the voltage is set to around 5.1v at the ROMs and you still get the error, you should try to replace the bad ROM. Cranking the voltage up will damage other chips.

RJ

[anotherit]

I set my power supply to a save level and I am reading 5.08v at pin 28 and reading .025v at pin 14 so the voltages at the chips seems to be at the correct levels.  Doing this has brought back the rom error 3 and I have made two videos showing the error:

http://m.youtube.com/index?client=mv-google&desktop_uri=%2F&gl=US&rdm=4owzg9gnw#/watch?v=Wzl5iq5IFsI

 http://m.youtube.com/index?client=mv-google&desktop_uri=%2F&gl=US&rdm=4owzg9gnw#/watch?v=MEJ0ViHmeSw

It seems the machine is locking up at a certain point everytime making me think that I do have to replace the eproms but what would cause the chips to work at the wrong voltage but not work at the right voltage?  Wouldn't it be the case that if it was just corrupted programming it would crash regardless of voltage?

Is there someone who offers a rom replacement service or would I be better off biting the bullet and buying my own uv light and programmer?

Thanks.

[channelmaniac]

I do PROM/EPROM programming as a service. Simply order the EPROMs and the programming SKUs off my website. Seth at HobbyROMs also does them.

At this point if it were my game I'd open the top on the power supply (with the game unplugged!!!!) and check the tops of the electrolytic caps in the output section to see if any have split their tops open. If they have, replace them and try it again. Poor power can cause these types of errors and it's easier/cheaper to check that first than after replacing chips.

RJ

[anotherit]

So I cracked open the Double Dragon power supply and found this ...

Needless to say I think I'm going to be ordering a new power supply tonight.

[stayouttadabunker]

I don't see anything wrong with it really... 
All the caps look happy...

[anotherit]

... I'm confused... all of that stuff around the base of the capacitors can't be a good thing... also the power supply makes random wizzes and pops which can't be good either.

[stayouttadabunker]

Oh, I thought it was the non-conductive gooey stuff they put on the bottom
to sort of glue the caps to the board.
The sounds it makes isn't good though...unless you're shooting a commercial
for a cereal that needs "Snap, Crackle, and Pop!" 

[channelmaniac]

That's glue under the caps and not the caps leaking... they look ok from a physical standpoint... they haven't split open their tops or blown out the rubber plugs in their bases. I wouldn't worry with replacing the power supply.

[cptbeef]

I just submitted a service request message through ArcadeComponents.com, so sorry for repeating some of this;

I purchased a Neo Geo cabinet recently, and it has never had working sound.  I'm guessing it will need a cap-kit installed.

I purchased a Metal Slug 4 cart since owning the cabinet, and the game freezes moments in to the game.  I have cleaned all the contacts, as well as the slots, which doesn't seem to be any help.  I have also cleared the back-up RAM, and run a RAM test by turning on all the hard dips, and the screen said "WORK RAM TEST" the entire time.  I ran the test for approximately 20 minutes.  Are you aware of any other potential causes for this? Should I also update the bios? is there a chance of bad RAM even though the system seemed to handle the RAM test fine?  To clarify, there is no error message, the screen and sprites just freeze in place.

3rd problem, and unrelated to the board itself;

The screen adjustment board has a broken nob on it, probably from someone bumping into it when they were reaching into the cabinet.  I have attached photos for reference.  Is there any hope of a replacement / fix?  I have done some searching, and haven't found anyone selling any.  Is it possible to repair this sort of nob? It is currently very delicate so I don't dare mess with it, but the screen is a bit wide and I'd like to be able to adjust it as well as contrast.

Any advice or suggestions are greatly appreciated.

[channelmaniac]

No problem repeating here... I'm behind on emails thanks to a hugely important project going on at the day job.

That broken pot will need to be replaced. You can find replacements at http://www.mouser.com but since I don't work on monitors I can't tell you exactly which part you need.

Look at the component and look for numbers printed on it. They will be something like "102" or "503"... which will tell you the value of the item in ohms. 102 is 10 followed by 2 more zeros, or 1000 ohms. 503 will be 50 followed by 3 more zeros or 50000 ohms. Once you find the value you want to look at the trim pots (potentiometers) at Mouser. Pay close attention to the data sheets for the parts as they have the sizes and pin spacings listed. This is important as you want something that will fit in the hole spacings on the board.

For the cartridge you need to clean the edge connectors with a pink pencil eraser, clean all the dust off well, reassmble the cartridge and test. Beyond that, check for broken solder joints or bad traces around the Program ROMs. If you had a UniBIOS chip in your Neo Geo you could use it to test the checksums of the ROMs to see if any are bad. Without that the only way to really tell if one of the program ROMs is bad would be to desolder each, read them, and compare their checksums with known dumps from MAME.

Raymond

[cptbeef]

Thank you for the quick response.

I have cleaned the contacts on the cart with an eraser, but no luck sadly.  I have tested ~12 other carts and all work fine in the system, though they are all older titles.  What are the chances it could have to do with the system and not the game at this point?  Just trying to see what my odds are through process of elimination.  The game's PCBs look fine, almost pristine compared to some of my older carts.

I don't know enough to be sure where the problem originates, and am greatly appreciative of your knowledge about these systems.  Thanks for taking the time to respond even while you're busy with your day job, and I look forward to hearing back from you when you're available to install a cap kit / remote battery.  Would it be realistic to send the cart with the system while you're servicing it, in order to detect the issue?  Or would this be more time consuming than it's worth? (sorry for my naivety about the repair time required, I know this can be tedious time consuming work).

[cptbeef]

Small update:

I purchased a copy of Metal Slug 3, which is having the same problem as Metal Slug 4.  A few moments into the game the system freezes, with no errors showing.

I am able to run the following games fine on the system:

Art of Fighting
Fatal Fury
Samurai Showdown 2
NAM 1975
Magical Drop 2
Puzzle de Pon
3 Count Bout
Top Players Golf
League Bowling
Baseball Stars Professional
Magician Lord
Ninja Combat
Aero Fighters 2

But for some reason Metal Slug 3 and 4 have the same symptom, which makes me suspicious of the board.  Does this info mean anything to you?

EDIT: After talking to the folks over at neo-geo.com, it sounds like my board is having a bank error; http://www.neo-geo.com/forums/showthread.php?234072-Metal-Slug-4-freezes-without-showing-errors

In the example given, the person did see an error screen.  My only concern is that my screen stays frozen.  Otherwise it sounds like a very similar problem.

[channelmaniac]

Hmmm...

The only thing that comes to mind is possibly a broken or gouged trace on the board. Look VERY closely at the board for any kind of liquid spill damage or corrosion and check for gouged traces.