Q&A Thread

[hairy otter]

4 slot the crosshatch problem.
took out the battery, wasn't leaking. Checked the traces near it anyways, all OK.
Checked the 74f138. The enable inputs get High Low High instead of Low Low High for single use as explained in the datasheet. Is this normal if the carts are not recognized?

Regards

[channelmaniac]

Check them while the board is booting to see if you have activity while it cycles through the slots.

However, you're better off checking for damaged traces/chips on the address lines and data lines on the cart slots.

[CraftyMech]

Back with another interesting board issue...

I recently picked up a Final Fight pcb with major graphics issues (the screen was split into fourths, and mixed up). I swapped out the A board for a known working A, and that problem cleared up.

However, now I have a character pattern across the screen that stays constant no matter what game screen I am on (including the test menu).

ROMs all check out OK on post. I've reseated the C board, checked the B/A board connectors, and reseated all the ROMs on the B board as well.

Could this be a problem with the C board smd chip?

I attached a screenshot of the display, covered in 'y' characters.

[CraftyMech]

Fixed!

Turns out I missed a ROM on the B board that had a bent under pin.

My guess is the last person who looked at the boardset tried reseating the B board ROMS to fix the graphical glitches caused by the bad A board. In the process they curled up a pin on the 05 ROM. I straightened out the pin, fired the game up and good as new! (when paired with a donor A board of course)

I found a nice tool for straightening pins, a pair of eye glasses pliers. The inside surface of the jaws are flat and smooth, so they work great for straightening out pins, or bring a whole row of pins flush.

[channelmaniac]

Nice fix!

I use a miniature pair of duck bill pliers. Look for a pair that have smooth jaws that meet perfectly. Grind the jaws down on the outside so that the tips will fit in between pins of the chip and you can use them to straighten individual pins from side to side and three pins at a time in a row.

Raymond

[ghandi]

Hey channelmaniac, I've got a neo geo 4 slot that isnt producing any sound. the caps look replaced but the HA130001 looks like it was poorly repaired. could you help me out with some repair work?

[channelmaniac]

Drop me an email at my user id at yahoo

[CraftyMech]

Hey Channelmanic, my adventures in board repair continue...

I just fixed the graphics on a Black Tiger board. The player/enemy sprites were all garbled, and I found that putting pressure on the big square graphics chip cleared up the problem. I knew from reading your repair logs that I should try reflowing the joints, but wasn't sure how to go about it. So what I did was to go pin to pin with my soldering iron and apply a little bit of fresh solder, holding the iron in place until the joint reflowed. Those pins are so close together that a number of times I bridged pins, which was kind of a pain to clear up even with flux and solder braid.

My question is if there is an easier way to reflow those pesky graphics chips that have so many pins? I thought about just laying down a solder coil on each side, and then cleaning up with my Hakko desoldering gun, but the angle is tough to get the nozzle in the right position. Curious what would be a good technique to leave the reflow work looking nice and clean, for resale purposes.

[channelmaniac]

Liquid rosin flux is your friend!

I use a 1/8" chisel tip, but a conical chisel works best as the dingle oval surface of the tip hold the ball of solder in the best position for the J legs on a PLCC chip. Lay down a thin layer of rosin flux on the pins where they meet the bard, put a small ball of solder on the tip of the iron, hold the board at an angle, and run the tip with the ball of solder down the legs where they meet the board.

The liquid rosin will help the solder flow and keep the solder from bridging. If you have bridges, simply use a bit more rosin flux and use a clean iron tip to draw the excess solder away.

[CraftyMech]

Thanks for the tips! I didn't know rosin flux was available in a liquid form, I have a tub of the sticky stuff.

The other problem that Black Tiger board had was no sound. I checked out the twin Yamaha sound chips, and matching op amps, and the outputs were dancing along as expected. Went I took the probe of my oscilliscope and ran it over the amp pins to see if I could get any static on the speaker, the sound suddenly came to life. Shorting pins 8 & 9 on the amp fully restored the sound. I'm not sure why that worked, the traces on both side looked good so that didn't seem to repair an existing connection that was broken. Board is working great now, so I'm not complaining

[channelmaniac]

Probably a bad cap that you zapped.

I use this: http://www.onlinecomponents.com/gc-electronics-104202.html?p=11386794

It's a good brand of flux.

[CraftyMech]

Probably a bad cap that you zapped.

Thanks for the link, I will give that stuff a try!

So I think I understand what you are saying about the cap. A bad cap was potentially killing the audio signal, but when I ran my probe over the amp pins I might have shorted it? I guess I should probably track that cap down and replace it, although the audio still sounds good as is.

[CraftyMech]

Just wanted to say thanks for writing up the CPS-1 audio repair guide that can be found in various places online. I picked up a SF II Championship Edition board set with no sound.

The YM3012 IC was not outputting audio, and the clock signal on pin 6 of the Z80 was ok. A couple of the Z80 data lines I checked though looked dead. Replaced the Z80 and fired the board back up, and sound was restored

I pulled the replacement chip from another A board with bad graphics, so it was not a Z80A like you have mentioned in your repair logs. I socketed the chip though, so at least if this one goes bad, it will be easy to replace!

[CraftyMech]

More adventures in repair land! I fixed a few boards since my last post, but now I've ran into a game that has me stumped.

Sky Shark from Taito. Powers up to black screen, nothing on the rgb or sync lines, and no audio. I tested the program roms (17 & 18) and they were good. With the board dead, I started at the CPU (68000). RESET & HALT are high, and there is a CLK signal that looks correct (8mhz). The address & data lines have activity, but the RW pin is stuck high (Read mode). A few of the address lines are low (A10-A12), while the rest have signals. A1-A13 are connected to a 8x8 64KB SRAM (HM6264LP), which is connected to EPROM 18 (program rom). I traced A1-A13 to the SRAM, and then to the corresponding pins on EPROM 18, and all the signals match (including the low value for A10-A12). I'm still suspicious of the low address lines, but also not sure if I'm barking up a tree without any cats so to speak.

Also interesting I found all three interrupt lines on the 68000 were stuck high.

Thinking the CPU might be dead, I pulled and socketed a new replacement with the exact same results.

So now I'm a bit stumped, any help or advice appreciated!

[channelmaniac]

What does the reset line do on power up?

[CraftyMech]

Good question, I had not thought to check!

I hooked up a logic analyzer and watched the reset line and a few of the address lines during boot. The reset line goes low for about 60ms, and that is followed by the address lines springing to life shortly after. 60ms sounds about right for a 68k, based on your "Reset lines & CPUs" post that mentioned a minimum period of 50ms.

I was hoping not to find that reset pulse there, that would have been a good lead.

[channelmaniac]

Check Address Decoding to see if the control lines for the program ROMs and Work RAM are working. If they aren't, you know where to check. If they are, replace the Work RAM.

[hairy otter]

A MC68000 CPU problem?
I've got a Demons World board which is dead. Black screen and no audio (appart from a click when turned on)
Took a look at the reset en hold lines of the CPU, they are OK. Clock is present, but there is no activity on the addres lines, does this mean the CPU is dead?

[channelmaniac]

Hard to say. What is the halt line doing and Bus Error?

You really need a storage oscilloscope or logic analyzer to tell for sure. Could be a bad CPU or could be that the CPU is trying to start but address decoding isn't working right for it to read the ROMs. The bus could be coming up and stopping faster than your logic probe will show.

[hairy otter]

I took some time to look into it further.
My cheap  storage oscilloscope tels me the reset and halt line is kept low for 280 mS. During this time address lines are high. When reset and halt go high, the address lines get busy for about 35 mS and then go high.
You ask about the Bus Error? Is this the BERR* pin? This looks the same as the address lines. So i think the CPU is OK?
I took out the Rom's and read them out, my programmer doesn't get problems reading them, so i think they are OK. Where to look now?

[Artemio]

If no schematic is available, you'll have to trace each address line, and maybe data line. They are probably multiplexed to different TTLs, to select between several chips. And the data might go into a 74LS244 or 74LS245. You'll have to check those and figure out which one is not responding.

Before that, I'd check if the data lines do get busy as well, if they don't and the CPU is reading, then you know it is probably the data bus.

[channelmaniac]

Yup... Artemio has it.

If your data bus is working then check after any 74LS244 buffers on the address lines and any 74LS245 latches on the data lines. If those inputs and outputs look OK then check address decoding to see if the ROMs are being accessed properly.

[hairy otter]

I checked all pins and this is what i came up with:

VCC 4.72V

Address lines:
A1 ON RESET 4.18V ACTIVE STOP 0.48V
A2 ON RESET 4.18V ACTIVE STOP 4.06V
A3 ON RESET 4.18V ACTIVE STOP 4.06V
A4 ON RESET 1.33V ACTIVE STOP 4.06V
A5 ON RESET 1.33V ACTIVE STOP 4.06V
A6 ON RESET 1.33V ACTIVE STOP 3.89V
A7 ON RESET 1.33V ACTIVE STOP 3.89V
A8 ON RESET 1.33V ACTIVE STOP 3.89V
A9 ON RESET 1.33V ACTIVE STOP 3.89V
A10 ON RESET 1.33V ACTIVE STOP 3.89V
A11 ON RESET 4.18V ACTIVE STOP 3.89V
A12 ON RESET 1.33V ACTIVE STOP 3.89V
A13 ON RESET 1.33V ACTIVE STOP 3.89V
A14 ON RESET 1.33V ACTIVE STOP 3.89V
A15 ON RESET 1.33V ACTIVE STOP 3.89V
A16 ON RESET 0V NOT ACTIVE STOP 3.89V
A17 ON RESET 0V NOT ACTIVE STOP 3.89V
A18 ON RESET 0V NOT ACTIVE STOP 3.89V
A19 ON RESET 0V NOT ACTIVE STOP 3.89V
A20 ON RESET 1.39V NOT ACTIVE STOP 3.89V
A21 ON RESET 1.39V NOT ACTIVE STOP 3.89V
A22 ON RESET 1.39V ACTIVE STOP 0V
A23 ON RESET 1.39V ACTIVE STOP 3.89V


Data lines:
D0 ON RESET 0V NOT ACTIVE STOP 0V
D1 ON RESET 1.14V ACTIVE STOP 0V
D2 ON RESET 1.32V ACTIVE STOP 2.63V
D3 ON RESET 1.17V ACTIVE STOP 0V
D4 ON RESET 1.17V NOT ACTIVE STOP 0V
D5 ON RESET 0V NOT ACTIVE STOP 0V
D6 ON RESET 0V ACTIVE STOP 0V
D7 ON RESET 4.62V ACTIVE STOP 0V
D8 ON RESET 4.28V ACTIVE STOP 4.02V
D9 ON RESET 4.56V ACTIVE STOP 4.34V
D10 ON RESET 4.56V ACTIVE STOP 4.34V
D11 ON RESET 4.56V ACTIVE STOP 1.71V
D12 ON RESET 4.56V ACTIVE STOP 0.81V
D13 ON RESET 4.56V ACTIVE STOP 3.72V
D14 ON RESET 4.56V ACTIVE STOP 2.06V
D15 ON RESET 4.56V ACTIVE STOP 1.78V


Other signals:
AS* ON RESET 1.52V ACTIVE STOP 0V
UDS* ON RESET 4.66V ACTIVE STOP 0V
LDS* ON RESET 4.62V ACTIVE STOP 0.46V
R/W* ON RESET 4.56V ACTIVE STOP 0.44V
DTACK* ON RESET 4.08V ACTIVE STOP 4.08V
BG* ON RESET 4.08V NOT ACTIVE STOP 4.08V
BGACK* ON RESET 4.08V NOT ACTIVE STOP 4.08V
BR* ON RESET 3.8V NOT ACTIVE STOP 3.8V
CLK 10MHz
HALT* ON RESET 0V 280mS STOP 4.66V
RESET* ON RESET 0V 280mS STOP 4.66V
VMA* ON RESET 0V NOT ACTIVE STOP 4.02V
E BLOCK SIGNAL 2.5MHz
VPA* ON RESET 4.2V NOT ACTIVE STOP 4.2V
BERR* ON RESET 4.7V NOT ACTIVE STOP 4.7V (NOT AS REPORTED BEFORE)
IPL2* ON RESET 3.4V ACTIVE STOP 3.4V
IPL1* ON RESET 4.7V NOT ACTIVE STOP 4.7V
IPL0* ON RESET 4.7V NOT ACTIVE STOP 4.7V
FC2 ON RESET 1.15V NOT ACTIVE STOP 3.84V
FC1 ON RESET 1.15V ACTIVE STOP 0V
FC0 ON RESET 1.15V ACTIVE STOP 4.2V

[hairy otter]

There was no activity on the D0 line and 0v so i checked this. Seems the pin gets grounded some where on the board. Its not easy to trace the line to other components as it dispears  beneath several IC's. So I cut the line near the CPU. (maybe I committed a sacrilege, but I'm only trying to save this board from the junk pile.) Now it's not grounded so the shortage isn't inside the CPU. If I measure the pin on start-up there is nothing happening on that pin it seems to be floating? This also happens with the D1 pin after "disconnecting", if I feed the line from D1 into the D0 pin it looks happy and gets some activity
Now I think I need to find the shortage somewhere on the board, also for the D5 pin. Or does the lack of activity on A16-21tell us there is allot more going on?

[channelmaniac]

You need to trace out where the connection goes and find the bad component. The CPU will never boot with the data bus being in a bad state.