MVS MV-2F input problem

dbstallman

Member
Joined
Jan 9, 2006
Messages
738
Reaction score
12
Location
Rock Hill, Missouri
I have a 2 slot board that seems to have developed a wierd control problem.

The joystick inputs on the character select screen of *some* games either won't let you move the selector off the first character or on other games will zip around so fast that it is very dificult to stop on the correct one.

For example : Agressors of Dark Kombat or SVC Choas will not let you move the selector on the character screen, KOF 2001 will move very quickly around and it is virtually random as to what characters you will select.

This doesn't happen on all of my games, though. KOF 2003, Fatal Fury, Metal Slug, and many others work perfectly. In fact it apears to happen on about 40% of the 50 carts I have. On all games, after the character select, the gameplay isn't affected.

I also have a 4 slot board that is exhibiting the same symptoms, but only with certain combinations of cartridges and slots, ie: SVC Choas won't let you slect if it is in the first or second slot and there is a smaller size cart in slots three or four. If all the carts are +300 meg, it doesn't matter what slot SVC Chaos is in.

For the 2 slot board, it doesn't matter which slot the problem game is in or if there is another cart in the other slot.

Thanks for your help.
 
Clear/reset the Backup RAM.

I had tried that earlier, but just now gave it another shot.

I didn't make any difference.

It almost seems like there is some kind of repeat timer like on a PC keyboard is malfunctioning. I just tested Metal Slug 4 and if you hold the joystick on the character select it moves one spot over and then after about 1 1/2 seconds it moves quickly from one character to the next (normal behavior), on other games it seems like the repeat function is going really fast right away and on others it's like it never starts.
 
I've never seen an MVS board in person so I can't really give you board locations, but it really sounds to me like a bad pull-up resistor. Measure continuity from each input on the edge connector, to ground, again on the edge connector. If you get a drastically different reading on one or more of them, your problematic component is directly connected to the offending pins.
 
I have traced out the inputs, and the only component I can find that has only the four directional inputs for P1 is this resistor array :



The controls go to pins 2, 4, 6, and 8. They leave from pins 3, 5, 7, and 9 and go directly into the NEO-C1 chip above it. I have tested the resistance from pin one to each of the others and the lines measure 1k ohm on the signal in and measure 1.33k ohm on the signal out. This is the same as the other four arrays that are right next to it for the rest of the inputs.

How do I determine the values for this array to get it replaced?

The logic of this problem is still really bugging me...

The motherboard should be handling the input values so wouldn't it just tell the cartridge that the joystick controll is pushed in a particular direction? Why does it only happen on some games and on those games why just the character select screen? The gameplay on all my carts work correctly and the games that work correctly on the character select screen never have any issues. If it was a failing component, wouldn't it be intermittent or not so reliably failing on the same games? Shouldn't the games that have problems continue the issue beyond the character select screen?
 
After shelving this for several months, I have decided to address it again.

I've found this : http://wiki.neogeodev.org/index.php?title=CRE401. It defines what the CRE401 is and whats inside. The values I measured on mine are consistent with what is inside. Just to be sure everything was checked out, I removed the two that were connected to the player 1 directions and buttons and replaced them with 8 1/4 watt 330 ohm resistors.

Afterwards there was absolultely no change to which games work and don't work. I have done the battery mod on this board as well. I figured I would leave the battery out for 24 hours to ensure the backup RAM was cleared. It did get cleared out, but again, no change.

I have a theory as to what is going on, though I don't know why or what is causing it : The games don't really work without issue beyond the character select screen, it just isn't as noticeable. The inputs seem to pulse very rapidly, for example, on Fatal Fury - Real Bout the selector will not move off of Terry Bogard at character select. However, durring gameplay movement has no noticeable issues but if you hold any attack button down it acts as if you have some type of rapid fire on that button.

My guess is that all eight of those switches (4 directions and 4 buttons) are pulsing very rapidly and some games don't mind a pulsing signal while others don't sense the switch closed long enough to register that it was even pressed. I don't know if what I just said makes enough sense, but rereading it I can't think of a better way to put it.

Player 2 controls still work fine in all games, and a logic probe on the player 1 buttons doesn't show a pulsing signal when tested at any point all the way up to the NEO-1C chip. It is just a steady ground signal like it is supposed to be.

Still looking for any ideas.

Brian.
 
Look for a gouged trace or bad plated thru hole around that C1 chip. Something is keeping it from chatting with the rest of the system properly or the chip is bad. In all my Neo Geo repairs I've never had to replace one of those.
 
I'll take another look.

From the chip pinout here : http://wiki.neogeodev.org/index.php?title=NEO-C1, it appears that all the P1 and P2 input data processing happens in the chip itself. I cannot see anything that indicates the control data leaving chip. I am inclined to agree with channelmanic that NEO-C1 isn't the problem, I just can't figure out where the input data goes next.

Brian.
 
In the pic it looks lie the legs are corroded, are they?

I've dealt with a lot of non-functioning controllers on Neo Geos myself and it is rarely the C1 itself like CM said. It's usually a gouged trace, corroded trace, bad resistor pack or an accompanying chip. There's a slight possibility that it's the F0.
 
In the pic it looks lie the legs are corroded, are they?

Are you talking about the NEO-C1 or the CRE401? I don't remeber if that October picture is from before or after I reflowed all the pins on the C1 (it looks like before to me), but the C1 is completely corrosion free. The CRE401 is corrosion free as well, but I have replaced it will a set of four 330 ohm 1/4 watt resistors.

I've dealt with a lot of non-functioning controllers on Neo Geos myself and it is rarely the C1 itself like CM said. It's usually a gouged trace, corroded trace, bad resistor pack or an accompanying chip. There's a slight possibility that it's the F0.

I have tested continuity from the JAMMA connector the the the C1 input pins IN00-IN07 (42-48,59) and other than the resistance from the pull-ups it is good. There is no difference from the values when testing P1 vs P2 on the solder side of the board either... well, now there is a little difference since I've replaced the CRE401's that control the four directions and four buttons for P1 with the 330 ohm resistors. The problem was there before I did that, though, and hasn't changed afterwards.

The most interesting thing is that the board's test mode shows all P1 and P2 inputs functioning correctly.... heck, half my games work without any noticable issues. It's just on the 26 or so games that can't seem to tolerate the quick on-off signal that seems to be reaching the game code.

I'll see if I can do a better job of explaining what is happening (as far as I can tell). Using Real Bout - Fatal Fury as an example again, when the game is started the character select for player one starts at Terry Bogard. Player 1 is unable to move the slector to any other character. The buttons will allow you to slect Terry, and then the games starts. While playing the game, the move controls work normally, although if you press an attack button it repeats the attack while you are holding the button down. For example if you press light puch and hold it down, you will see Terry do : punch, punch, punch, punch, punch.... etc. in rapid succession. The same is true for all the P1 buttons. Player 2 can insert a coin, join the game, select any character he wants at the selection screen, and then durring the fight if an attack button is pressed and held the character will only perform one attack. If the P2 attack button is released and then pressed and held again, just one attack. This is the normal behavior.

The directional controls for Player one have the same type of quick succession repeating function as the buttons. If I put Kabuki Clash in, you can move the P1 character selection, but it just zooms from character to character so fast that it is in effect a random selection. You have to just tap the joystick and hope it stops on the correct character that you want, if not, repeat.

Finally, on King of Fighers '94, Player 1 and Player 2 work correctly at all times, character select, gameplay, attacks don't rapid fire..... The same is true for all the Metal Slug games, and the other 20 some odd games that have no issue.

When I put the logic probe on the JAMMA pins, they just short to ground when the button is pressed. Logic probe on the pins for the NEO-C1 are the same, short to ground when the button is held down, floating when released. There is no indication that the button signals are going on-off from the JAMMA connector to the C1.

Does anyone know how the input signal travels after it leaves C1? Does it leave C1, or is it entirely processed there?

Brian.
 
I've replaced the CRE401's that control the four directions and four buttons for P1 with the 330 ohm resistors. The problem was there before I did that, though, and hasn't changed afterwards.

Hey Brian,

Are you saying you removed the 401 and replaced everything with just 330ohm resistors or just the outputs?
 
Are you saying you removed the 401 and replaced everything with just 330ohm resistors or just the outputs?

I removed the 401 and then placed a 330 ohm resistor between pins 2-3, 4-5, 6-7, and 8-9. Pin 1 and 10, being ground and 5v were left open. That is what I took the statement "Drop-in replacements seem hard to find but using simple 1/4W 330 ohms resistors seem to be a good quick fix." on the Neo-Geo Development Wiki page about the CRE401 to mean.

Brian.
 
Back
Top Bottom