Jump to content

SNESNESCUBE64

Member
 View Profile  See their activity

  • Posts

    615

  • Joined

  • Last visited

  • Feedback

    100%

 Content Type 

Blog Entries posted by SNESNESCUBE64

  1. SNESNESCUBE64
    On my bench (technically the trashcan next to my Helifire table) was Nintendo Space Firebird. This one was specifically the Gremlin license but they are the same game for the most part. The problem with this one is that some sprites are garbled or missing.



    This ended up being one of the quickest repairs I have done. It probably took longer to hook up the board. The problem chip ended up being the second chip I did a spot check on: the 74ls00 at 3D. Probing the chip, I noticed that it had good inputs.

    However, it had floating outputs:

    Stacking a new chip on it fixed my graphics. That was ultimately the problematic chip. Pulling it and testing it on my chipmaster shower it to be bad.


    Fully replacing the chip solved my graphics issue.



  2. SNESNESCUBE64

    Wii U kiosk Monitor - Audio Issues

    By SNESNESCUBE64,


    Kiosk Repair
    This is something a bit different that I did a couple months ago. I'll start off by saying the Wii U Kiosk owner should consider doing this fix to their monitor. It is cheap and easy to do if you have any experience with a soldering iron. The monitor is fairly easy to pull out and dissassemble.

    As an overview, there are three boards, a logic board, a power supply, and a backlight board. I theorized that it was just cheap capacitors as I still heard a buzz coming out of the speakers. Because there was a buzz, I figured that the audio amp was fine.

    I started out by making a cap kit. Most of the caps were cheap, I ended up replacing all of them except for one of them on the logic board, of which I mistaken it for another value. Most of the caps were mediocre in quality, brands like Chong. 

    The replacement went really quickly. Ultimately it ended up bringing the audio back. It turns out the sound quality isn't actually all too good. I might look into replacement audio amps and whatnot to try to reduce the noise. That said, I believe the issue was on the logic board, but it could have been elsewhere since I just shotgunned all the parts. With that in mind, all the power supply and backlight board are actually off-the-shelf PCBs that you can buy on ebay. The only thing that is proprietary is the logic board, which was made by Lynx. 
    Attached is the Excel file I used, it is not complete because of the previously mentioned capacitor being the wrong value, I apologize for that. I would take this list as-is, as mine may be different than yours ultimately.
     
    Wii U Kiosk monitor.xlsx
  3. SNESNESCUBE64
    The purpose of this is to put logs that I don't think need a proper writeup. It should be updated periodically.
    Date: November 26th, 2021
    Game: Williams Earthshaker
    Symptoms: Left flippers not working, fuse F6 popping after holding down left flipper button
    Solution: End of stroke switch for upper flipper was not aligned and disabling one of the windings on the flipper coil. Adjusting it fixed the issue.
    Date: June 12th, 2023
    Game: Demolition Man
    Symptoms: garbage on lower half of the display regardless of what DMD unit I use.
    Solution: Bad 6264 (don't remember the designator) on the display driver board.
    Date: July 8th, 2023
    Game: Gottlieb Buck Rogers Pinball
    Symptoms: no displays despite repro boardset being in the unit
    Troubleshooting steps: Noticed HV light for the displays was out. Checking the fuse, I could see that it was out. Replaced fuse, still popped once. Checked harness connections and one of the pins was pushed in. Replaced pushed pin and fuse didn't pop.
    Solution: replaced display fuse due to bad harness connections and replaced probable cause for popped fuse.
    Date: September 9th, 2023
    Game: Bally Strikes & Spares
    Symptoms: Sticking Flippers
    Solution: Used oil in the coils, that is a no-no. Cleaned up oil and replaced sleeves.
    Date: September 23th, 2023
    Game: Williams Terminator 2: Judgement Day
    Symptoms: Cannon not homing, missing some General Illumination
    Solution: Cannon not homing issue resolved by re-aligned the switch. General illumination, like with many games of this time, had burnt connectors that needed replacement.
    Date: November 18th, 2023
    Game: Rocky and Bullwinkle
    Symptoms: Right Flipper not working
    Solution: Bushing that held the piston to the rest of the flipper assembly no longer held the link. So the piston wouldn't actually flip the flipper
    Date: December 2nd, 2023
    Game: Williams Gorgar
    Symptoms: Both flippers weak
    Solution: End of stroke switches were not very well adjusted. Adjusting them properly increased the flipper's power to acceptable levels.
    Date: December 9th, 2023
    Game: Williams Black Out
    Symptoms: Player 1 score display flickering
    Solution: Display tube was gassing. Replacing the tube solved the issue.
    Date: December 9th, 2023
    Game: Williams Twilight Zone
    Symptoms: Balls sometimes getting lost in the gumball machine area
    Solution: This was a goofy one. It ended up being a combination of slightly dirty optos and the fact that the far left trough switch had it's actuator replaced with a home made one. The problem was is that it would sometimes hold up the left trough switch so that it wouldn't register. I believe that's where the ball becoming lost came into play but I am not sure. It worked after all of that.
    Date: December 10th, 2023
    Game: Bally Bobby Orr pinball
    Symptoms: Something was just off about the lights. They weren't making sense
    Solution: One of the light driver connectors was offset by one pin, meaning that the wrong things were lighting up.
    Date: December 10th, 2023
    Game: Bally Popeye Saves the Earth
    Symptoms: Bottom right flipper not working
    Solution: Flipper fuse blown for bottom right flipper on driver board. Cause of fault was because the end of stroke switches for the flippers were trashed.
    Date: December 17th, 2023
    Game: Williams Flash
    Symptoms: Game not starting
    Solution: This was a really weird one. It was struggling to recognize switches with column one. Every time you would hit any switch in that column, it would trigger every switch in the row. So the game wouldn't start unless you held down the start button as a result. The solution ended up being to replace the 13N10L at Q81 (Rottendog board) with an IRL540 as the 13N10L was no longer manufactured and the IRL540 is what I had on hand.
  4. SNESNESCUBE64
    This entry is meant for simple fixes that aren't worth going in-depth about the repair. It will be updated periodically to add new entries. I only have one to start but should have more later.
    Date: June 12th, 2023
    Game: VS Super Mario Bros
    Symptoms: Low scratchy sound on sub side
    Solution: replaced bad LM3900 at 3N
    Date: September 4th, 2023
    Game: VS Super Mario Bros
    Symptoms: monitor (20ez) experiencing weird issues, small screen and low/unadjustable B+.
    Solution: replaced bad flyback
    Date: September 23rd, 2023
    Game: NFL Blitz 2000
    Symptoms: Random game resets
    Solution: Reseated video card, adjusted voltage to 5V
    Date: December 2nd, 2023
    Game: Ms. Pacman
    Symptoms: Random game resets when the game warms up.
    Solution: repaired damage game board card edge. HASL tarnishing made a poor connection that resulted in it getting burnt
    Date: December 10th, 2023
    Game: Crusin USA
    Symptoms: Game not holding calibration
    Solution: Replaced battery.
    Date: December 16th, 2023
    Game: Ms. Pacman
    Symptoms: No boot, loud humming noise
    Solution: Bad game board fuse block/blown out incorrect fuses.
    Date: December 16th, 2023
    Game: Punch-out
    Symptoms: Game not freezing on the title screen
    Solution: Reseated program ROMs on the game board.
  5. SNESNESCUBE64
    Seems like I am doing a lot of display repairs as of late. This time it was for Pinbot by Williams. In this instance, it was missing the middle character for player 2 and player 4. This one is fairly easy to handle looking at the schematics. The other issue is that the credit digits are missing segments 'B' and 'D'.

    The first thing to tackle is the missing segments. each display unit uses it's own driver IC for turning on the digits. So it would be just finding out what is in common. Looking at the schematics, I can see that the middle digit is STB13 and STB13'. 

     
    With this identified, I can see that they each converge to a single NOR gate at U4. Odds are U4 is my problem with that. They used a CD4001. I've had the worst luck with late 70s and early 80s 4000 series CMOS ICs, so I didn't doubt that it was bad.

     
    After replacing the IC, I found that the two characters were back. So onto the next issue, the missing segments on the credit section. As previously mentioned, I was specifically missing segments 'B' and 'D'. The first thing to check in this circuit are the resistors that go in-between the driver and the display. In this case, they are 10k 1/2W resistors R21 and R23.

     
    Testing them in circuit, I found that they were open, meaning that they would not pass any current through. After replacing those two resistors, I found that all displays were working as they should.

  6. SNESNESCUBE64

    Tektronix 2232 Oscilloscope - CRT won't focus

    By SNESNESCUBE64,


    Equipment Repair
    I have been using this scope for quite some time now and it has served me well. It is my go to for quick troubleshooting as it is quick and easy to setup compared to the scope functionality of my logic analyzer's scope. So you can imagine that I was a bit bummed out when I started having issues. Fortunately this one was a simple repair.
    The CRT would not focus all the way. Basically, I would have trace and text if I turned it on, but it was blurry. Typically, it would be simple in the sense that you only have to adjust focus on with the adjustment knob. However, even turned all the way it wasn't quite working the way I needed it to.

     
    The first thing I did was look at the schematics for the CRT section. Fortunately the service manuals for a lot of these old Tektronix scopes are available online for free. So finding the area of interest was super straight forward. The way it works is that the supply for focus is about 2000V, the focus adjustment basically just divides that voltage down just a little bit to tune the CRT. From there it seems that the rest is biased using several 510k resistors in series. 

     
    With the knowledge of how focus voltage is attained, I tested all the CRT voltages. Everything seemed to be there, including the 2000V supply. Next thing I checked was that row of resistors. To my surprise, R888 (A1R888) was testing as open. That is an obvious problem so I went ahead and disassembled the rest of the scope and replaced the resistor. Unfortunately I didn't have 1/2W 510k resistors on hand, so I used a 1/2W 470k resistor since it would still have been in the 5% tolerance range anyway. That fixed it, I now can focus to a nice sharp image.


     
    Crazy what one little resistor can do.
  7. SNESNESCUBE64

    Andromeda SS - Can't Count Correctly

    By SNESNESCUBE64,


    Arcade Repair
    This was a really goofy one that a friend sent me for repair. It's an odd space shooter-type game made by Irem called Andromeda SS. It uses M11 hardware with several mods for sound. Fortunately for me though, the CPU board didn't have many mods.

    The main issue is that it wasn't counting correctly. In Andromeda SS it has a bonus counter on the bottom that counts down. Basically, it starts at 99, but once it got to 95 it would count weirdly. It would count kinda like this: 95->91->90->81->80->71...

    The big thing that I was asked to do was hook my Fluke 9010 to it just to check. I originally wasn't going to be looking into this much due to lack of available time. But I told him if the fluke pointed out something obvious I would try some things.

    Basically, first thing I did was do RAM short tests. Typically if RAM has something wrong it will fail during the short test. I tested both the scratch and video RAM as that is all the CPU can both read and write to. From I read the ROMs just to get the checksums, the goal was to be able to see if it matched what was in MAME. However, before I did the comparisons, I did RAM long tests, which is a more thorough RAM test that the fluke can do. To my suprise the video RAM was would fail about 20 seconds or so into the test.

    So if the video RAM was flakey, it means that things like counts could get messed up. Looking at it, there are four sets of RAM chips.
    Scratch RAM - six 2111 Video RAM - two 2114 Color RAM - eight 2102 Background/stars - three 2102 In my experience, 2114 SRAM ICs can be incredibly problematic. When they start to go they can flat out fail or be inconsistent. So I just went ahead and socketed/replaced them.

    After that, I fired the fluke back up and performed the same RAM long test, this time it succeeded. From there I ran the game with the fluke and saw that it was counting just fine. Glad that worked because I wouldn't have had much of an idea on where to go from here.

  8. SNESNESCUBE64

    Williams Stargate - Not booting

    By SNESNESCUBE64,


    Arcade Repair
    Today I had to work on a Stargate by Williams. The complaint was that it was stuck in the rug pattern. Taking a look at the back, I saw that all voltage LEDs were lit on the power supply. I also noticed that I was getting a RAM error code: 1-3-1.

    1-3-1 is a RAM error code that is supposed to indicated that the first RAM in bank 3 has failed. It is also the first RAM that get's checked, awfully suspicious. First thing that I checked was the voltages, all were within 5% so I figured that it was good. Next step I did was just swap the RAM chip. All the RAM chips on the main board are socketed and I had one on hand. So it was an easy check. Still not working.
    The next thing that I checked was the 74153 MUXs, they are responsible for addressing of the RAMs and after doing some reading they are common failures. Checking all the outputs, everything looked correct except for the output of pin 7 on 3E. 


    On the oscilloscope, it was looking like either just noise or that it wasn't going low enough to register as a logic low. Regardless of that, it looked faulty. Pulling it out of circuit and testing it, it seems to have failed. 

    Now that it has been replaced, it seems like the game is working just fine. 

  9. SNESNESCUBE64
    This is probably one of the more common issues that I've seen on these old Ballys. The way these work is each display is it's own module capible of displaying a score or whatnot. Sometimes what can happen is that segments can get stuck or not light up at all. This one is from a Kiss pinball.

    Essentially, the way that these modules are driven can be split into two parts: digit driver and segment driver. If a certian digit is not lighting up or is stuck (many times 8 because of all the data from the other segment) then you probably have a bad digit drive transistor.
    In this instance it is a stuck segment. We can see that the bottom segment, or segment d according to the schematics. Looking further at the schematics we can see that segment d is driven by Q16.

     
    Replacing Q16 ended up correcting the issue. There other failure points on the module, such as cold solder joints on the connector and a bad BCD chip. Reflowing the solder joints on these modules should be a given. It is very important to do such otherwise you will get inconsistencies ranging from improper display data and flickering digits all the way to nothing at all.
    When the BCD chip fails, I've seen it to where it will display either all 8's or nothing at all. But individual segments can get stuck as well potentially.
  10. SNESNESCUBE64

    Nintendo Helifire - Game Not Booting

    By SNESNESCUBE64,


    Arcade Repair
    I recently picked up a Nintendo Helifire cocktail table and been having a blast bringing it up. There was one main problem though: the original CPU board wasn't working.

     
    With this though, I was able to cheat because I knew it was a problem on the CPU board. Reason being that I had an extra, but I wanted this one working. The first thing I always do when working on a game that isn't booting is check the ROM data. ROM corruption is a very common issue with arcade games due to their use of EPROMs. Unfortunately, these use 2708 EPROMs. Problem is that none of my programmers can really read these due to them being goofball tri-voltage chips. My solution was just to replace all of them with chips that I can program. What I did was make an adapter board that combines a pair of eproms into one. This reduces the chip count and I was able to program my free play mod onto it as a result. It also assisted me in this repair as it garunteed that I had good ROMs.

    Testing with the new ROMs, I still was getting no boot. One thing I like to do with games that utilize 4116 tri-voltage DRAMs is to socket them. They are very high failure compared to other RAMs, so having them socketed on my games is nice. One thing I did find is that several of these RAMs were faulty. I decided to replace the entire bank of 8 to ensure the timings were all the same. Trying out the game, it still wasn't booting!
    One thing I like doing is checking all the signals coming in and out of the CPU, in this case an 8080. Everything looked fairly normal on the 8080, so I moved to the address buffers since it's an easy thing to check. On Space Fever/Helifire boards they are buffered using 74LS08 AND gate chips. One input is always tied high for each gate, so they just act as a passthrough.

    To my suprise, A8 was stuck low despite having a pulsing input.
    Input:

    Output:

    This is a very obvious problem, either the AND gate there was faulty or something in circuit was pulling A8 low. I went ahead and simply replaced the 74LS08. This chip was what was holding the board back. With a new one it now boots as it should.

  11. SNESNESCUBE64
    This was probably one of the nastier 20EZs I've worked on. I don't know what this thing's past life was but it was pretty rusty and coated in dirt.



    Regardless of how dirty it was, it will clean up fine. The problem is that the picture was screwed up. It was dim in and had partial vertical collapse. You can even see some waving.

    One thing that is super common on 20EZs is that the electrolytic capacitors go bad. This is especially common with the higher voltage (160V) capacitors. Typically the burst on the bottom like these. When they go bad they can cause brightness and deflection issues in my experience. Just replacing the capacitors will fix a lot of these issues.


    After a good wash and a recap, the monitor was back to working like it uses to back in the day. Lots of times with these it just needs new capacitors.

  12. SNESNESCUBE64
    Game: Williams Corvette Pinball
    Symptoms: DMD display characters was a little blotchy
    Troubleshooting Steps: 
    Looking at the display, I could see that it didn't look like, it almost looks like the characters were blotchy and distorted. You could see dots around the characters, it was super visible in person as well.

    The first thing that I did was checked was the voltages that the DMD was getting. One thing that I noticed was that the -125V line was rather out of spec, I was measuring -140V, where as other DMDs at the shop were measuring closer to -116V to -120V This difference can cause some biasing issues, which could cause display issues. With that in mind, I had to troubleshoot how the -125V rail was being generated.

    The -125V rail is ultimately generated by Q6, a MJE15031. Checking everything else in the circuit, everything seemed like it was working just fine. I did notice that the voltage Q6 was generating was pretty close to BR2, which implies that maybe it was shorted. So I went ahead and replaced it. After doing that, it resolved the issue with the blotched characters went away.

     
    Solution: Bad Q6 (MJE15031) on the display driver board
  13. SNESNESCUBE64

    Vision Pro MTG-2901 - DOA/clicking

    By SNESNESCUBE64,


    Arcade Monitor Repair
    Monitor: Vision Pro MTG-2901
    Symptoms: Completely dead, clicking when powered on
    Troubleshooting steps:
    Like with any monitor, the first thing I do is make a capacitor kit and replace all the capacitors on the chassis. Arcade monitors work long hours and have exceeded their life in many cases. This is especially important with these Happ and Vision Pro monitors, of which have terrible capacitors. Replacing the capacitors in this instance had no real effect on the issue.
    Upon inspection, I noticed that R812 got hot, although it tested good. It is involved in driving the switch mode power supply (SMPS). Looking at the circuit, it is involved with Q801, a 2SC2073 used in driving the SMPS. It looked like that got hot as well.


    It's safe to say, if this isn't working, the SMPS won't be able to operate correctly. Replacing it ended up solving the issue, I also replaced the resistor because I didn't want to return this with a burnt resistor.

    Solution: Replaced Q801 (2SC2073) and R812 (68ohm 1/2W)
  14. SNESNESCUBE64
    Symptom: Lower left drop targets not resetting when all three are knocked down.
    Troubleshooting Steps:
    The first thing that I do whenever I deal with a non-functional coil is I check the coil itself. On a pinball machine, it is super easy to do this typically. The first thing you would do is remove the lock bar and playfield glass. Next, be sure to remove the balls themselves. This is an important step, as when you are lifting up the playfield, they can free themselves from the trough, potentially causing damage to the plastics or if you are REALLY unlucky, the back glass. After the balls are removed, lift up the playfield and secure it on the kickstand (or whatever means that particular unit has for doing such. For Black Knight, it has a kickstand so that makes it easy.
    Upon visual inspection of the coil, I could see that it is super burnt, meaning that it was constantly engaged for way too long. You should be able to move the piston that solenoid pushes with your hand under normal circumstances and it should move smoothly. This one however was seized, indicating that the coil or sleeve needed to be replaced. This case was a bit extreme, so the coil is just going to be replaced. This happens for a few reasons: return side of the coil is shorted, the switch telling the mainboard to engage it was stuck, or something in the drive circuit was damaged. The drive transistor is typically the first place I check if there are no visible shorts, as those fail pretty often as they are hard working.

    Looking at the manual, it was really easy to identify which part of the circuit to look at. You can typically find pinball manuals online on websites like ipdb.org. Each manual is different, but manuals by Williams typically were comprehensive. Table 2 and 4 told me the solenoid number (02), what coil to replace it with, and which drive transistors were involved.

     
    With the transistor in question identified, I could easily take a look at the board. The only problem, is that Williams did NOT use silkscreen on the gameboards during this time, so you have to refer to the component layout diagram in the manual. You can also double check the schematics to make sure there wasn't a mistake, as well see the rest of the circuit to inspect.

     
    So now looking at the board, I could see little balls of solder on the cooling mount tab and discoloration, a strong indicator that it got pretty hot. At this point, I was fairly confident that this was the issue. The 2N4401 that drives the drive transistor could also have failed, but that happens a lot less than the drive transistor itself as it is working with a lot less. The 7408 could also fail, but many times you would see multiple coils having issues. So I decided to just replace the drive transistor at Q17 with a TIP102.

    After replacing the drive transistor, the coil performed as it should. In order to avoid excessive coil damage like this, make sure that your solenoid fuse is the correct value (the manual has that information and there is typically a fuse chart in the back box) and if you see or hear a solenoid fire and get stuck, it is a good idea to turn off the game and investigate. Coils are still made but are more expensive than a simple switch adjustment, just replacing the drive transistor or coil sleeve, and fuses are cheap so use the right value.
    Solution: Replace bad solenoid and Q17 (TIP102)
  15. SNESNESCUBE64

    B&K 467 CRT Tester/Restorer

    By SNESNESCUBE64,


    Equipment Repair
    This is a bit different than what I normally do, in fact this is the first peice of equipment that uses a vacuum tube in it that I have acually worked on. So for those unfamiliar with the B&K 467, it is a peice lf test equipment meant for analyzing the health of CRT-based displays and if necessary, attempt to "rejuvenate" the tube in order to squeeze a bit more life out of it. It is particularly useful in the arcade world as many of those cabinets still have a well-used CRT in them. It is calible of testing most CRTs as it utilizes a modular approach by using interchangeable sockets.

    When I got it, the tester was in a very well-used state, with the case being held together by duck tape. The first thing I did was add hinges and latches to make the case open and close as it should. What I used was just generic latches and hinges meant for small kitchen cabinets. It turned out very well in this case. I think it looks very nice for this application.

    The next step was to replace the line cord, the original was not in very good shape. It was starting to fray on both the line and neutral wires, because it risked a short, I replaced it before I even powered it up. While I had it dissassembled, I also relfowed the solder joints that were on the board, particularly with the connectors.
    After it was patched up, I was able to finally test it. It powered up fine, and even let me set the heater and G1 voltages. The problems arose when I tried to set the G2 cutoff voltages for each gun. To describe the problem, basically if you turned one knob, all the guns would adjust, which is incorrect as they should adjust individually. After the annoying disassembly, I realized that overall it was simple, but a mess. Thank goodness for the good dissassmbly instructions, which guided me through the process.

    So after reading the service manual and looking at the schematics, I was able to narrow down the points of interest. Since it was a more universal issue, it mostly had to do with the circuit containing the vacuum tube and the multiplexing circuit. 

    The service manual suggests checking a few places in this area, the 12BH7A vacuum tube, transistor Q1, diodes, D1 through D4, capacitors C11 and C12, resistors R41 and R43, and the two integrated circuits involved in the waveform generation. Unfortunately, none of these components tested bad, everything in theory was working. I even brought the tube to a local electronics shop that still had the oldschool tube tester. Afterword, I started just checking the components that were not specifically listed in the service manual as points of interest. To my surprise, the culprit ended up being a few bad resistors that showed no signs of damage. The major one was R70, which was connected to the tube itself was open. Two other resistors, R44 and R37 were also out of spec. After replacing these, I put the thing back together (frustratingly you have to reassemble the whole thing in order to properly test it.

    After everything was said and done, I fired it back up and was finally able to set the cutoff properly and move forward with the other functions. Everything seems to be working great now, I was even able to use it to clean and balance one of my Donkey Kong monitors, tested good afterwords as you can see in the image below. 

    All that is left now for me to do with this tester is to build more adapters, which I will do as I aquire more sockets. I also need to just build one of those universal harnesses in case I can't get the socket.
  16. SNESNESCUBE64

    Laser War Pinball - Random Resetting

    By SNESNESCUBE64,


    Pinball Repair
    Symptoms
    Random reset, sometimes resetting rapidly

    Solution Steps
    Looking at the schematics, reset is controlled using a Power On Reset (POR) IC. Basically how it works is that it sends a reset pulse to the CPU whenever a threshold voltage is reached, somewhere in the 4V range. With that knowledge, I had to monitor the reset line.

    In this image, the reset line is yellow and the +5V rail is in blue. What I wanted to do was see if there was maybe a short or if there was excessive noise going on that went below the threshold. To my surprise, it periodically dipped a lot lower than expected, causing the reset. We definitely have a power supply issue at this point.

    Looking at the power supply circuit for +5V, there isn't a whole lot going on here. How it works, is that an AC line from set of taps directly from the transformer gets rectified and filtered. This makes a unregulated +/- 12V DC. From there, the +12V DC circuit is regulated to 5V using a 2N6057 and MC1723 circuit. 

    Looking at it, my initial guess was the rectifier as these fails often. So I test it in circuit, and suprisingly one of the diodes in it was shorted, specifically on the ones creating the +12VDC line. So if it was failing, it could explain the dip in 5V. What was super interesting was that it didn't pop the fuse. Normally when one of these fails, they take out the fuse. Replacing the rectifier fixed the issue and it hasn't improperly reset since I replaced it with a new one.

    Solution
    Bad power supply bridge rectifier at DB1.
  17. SNESNESCUBE64

    Head On N - Blocky Graphics

    By SNESNESCUBE64,


    Arcade Repair
    I got a pretty interesting board on the bench this time: Head On N, released by Nintendo, Licensed (Or stolen) from Sega. This is an interesting boardset, as this is for the most part a clone of Sega Vic Dual hardware, except they spread it across two boards and included the sound board on the CPU board. I had already done a deal of work deconverting the board, unfortunately this thing was missing a lot of parts and was modded to play an unknown game (the roms weren't included).

    On to the issue. The problem at hand is that there is no detail, just blocks moving around on the screen.

    First thing I did was hook it up to my Fluke 9010A to do RAM and ROM checks. The RAM had been previously desoldered by someone else, so I socketed it (4027 DRAM fail pretty often) and fixed some damage traces done by the previous owner. Everything came back as good, so I figured that the issue was after the RAM, since graphics and system RAM are shared. There are no schematics, but since this is so similar to Sega Head On hardware, I just used schematics for that (I actually used the schematics for Tranquilizer Gun, since Head On used a custom IC for video stuff). Looking at it, the first thing that happens is the graphics data gets serialized. From there, it get's processed with sync and color gets overlayed. Also note that on Head On N, the 74ls166 was sometimes substituted with a 74ls198 (as other Nintendo games like Space Fever used that IC at the time).

    Looking at the 74ls166 (found at 3C) with my oscilloscope. Looking at all the signals, everything except for the output looked correct. Output was stuck high. I decided to hook it up on my logic analyzer just to make sure that it wasn't some weird timing. However, it just seems that the output was always high despite not supposed to be. Notice that it gets loaded, but the output doesn't change despite being enabled.

     
    Pulling the 74ls166 out of circuit to test in my ABI chipmaster, the chip tested bad!

     
    After replacing the 74LS166, the graphics were back to normal. Next step is going to be testing sound!

  18. SNESNESCUBE64
    This was a neat one because I have not encountered a soundboard issue like this one before. In my short experience with Williams sound boards for games like stargate and joust, they have been an "all or nothing" kind of deal. This one however was missing sounds.

    To troubleshoot this one, I did a few things. When I booted the game up I pressed the test button on the soundboard, it should play game sounds in general, but this just played tones. What this told me was that the CPU, DAC (probably) and amplifier area was working on this board. My guess immediately turned to the EPROM or 6810 Bipolar RAM. I started with the EPROM, what I did was that was dump the ROM and validate it with ROMIdent. The ROM checked out just fine. From there, I decided to test and socket the 6810 since I have had the worst luck with RAM ICs this year and I was replacing the sockets for the ROM and CPU anyway. Using my ABI Chipmaster, I verified that the RAM was indeed failing. After putting a new one in and testing the sound board again, it was working as it should.
    I did a little video demonstrating what I was hearing since sound is not the easiest thing to describe:
     
  19. SNESNESCUBE64

    Donkey Kong - No Boot

    By SNESNESCUBE64,


    Arcade Repair
    So I had gotten some semi-repaired boardsets off KLOV last week. I've been wanting to get another Donkey Kong boardset for a while since I plan on getting a cabinet soon. Both boardsets did not work, both having different errors. When hooked up, it booted to nothing. First thing I noticed was that a bunch of chips had been socketed. None of the sockets were in good shape (one was an old nasty machine pin socket, another was a damaged dual wipe, and there were two terrible single wipe sockets), so they needed to be replaced. After doing such, I finally got a picture on screen.

    Afterwords, I saw that it wasn't booting, so I was able to make use of my recently fixed Fluke 9010a. I used it to run RAM and ROM tests, all the CPU ROMs and CPU RAM tested good. So that meant that something controlling the video board was goofy. I tried writting to the start of the VRAM (Address 7400) and found I was unable to write the data and read it back.

    So I set the fluke to constantly write FF to address 7400 in hopes of trying to figure out where the disconnect was using my logic probe. I found that for some reason, all of the data lines were just floating, occasionally having some activity. This means that the buffer at 6A was not passing the data by. It was getting data in, but it wasn't outputing, implying that it wasn't being enabled, meaning it was stuck in triastate which makes sense for it not to be passing data. After looking at the enable line, it looked as if there was some activity. But after looking at it on my oscilloscope, the line wasn't being pulled low enough in order to enable the buffer. The chip that controls this is a 3 to 8 line decoder (74ls139) at 2A. After replacing it and running the ram check on the VRAM (7400 to 77FF), it passed. From there I ran the game and it booted up.


     
    The game worked just fine. Only problem was the colors of the characters were goofy. They would basically change color half way through being drawn. So this had to be a color selection error since it was being drawn properly otherwise. Basically I landed on 5M, a 2 to 1 multiplexer (74ls257). By this point I had just been measuring the inputs with my oscilloscope just to see if they were there but too low to be recognized, and while checking pin 2 on 5M, it had fixed itself. I had to think about it for a minute as to why this worked. I was looking at it and the ground clip had come off, after the clip came off the signal started acting goofy again. At that point it had all made sense. Inside my scope probe, there is a 22pF capacitor. This acts as a filter to get a better reading on the oscilloscope. By measuring it, I was actually slightly filtering the signal. 
    This can be seen below, I had added a 68pF capacitor as that is what I had on hand. Notice in the first image, there is a lot of wobble. If the wobble gets too high, it is basically tristate, which means mux may not interpret the signal properly.
    Ultimately a small 68pF capacitor fixed the issue.

    No capacitor:

    With 68pF capacitor

     
    After this was done, I now have yet another working Donkey Kong boardset. This one was a bit more challenging than the other nintendo boards that I've worked on. Especially with that wonky signal, I found that on accident ultimately. I'm just glad in the end it worked out.

    If you have any questions, feel free to ask in the comments, or if you are uncomfortable asking it publicly, shoot me a message and I will be more than happy to attempt to answer it.
    Update: chips outputing a noisy signal like that of which I fixed with a small capacitor are an indicator of a failing chip. After taking a look at the schematics again, I found that the noisy signal came from the 74LS245 buffer at 1S, after replacing that, the video board worked perfectly without the capacitor.
  20. SNESNESCUBE64
    This was a fun one. I have a whole box of these boardsets and I figured out which ones had working CPU boards, so I was able to immediately narrow it down to the video board. The initial problem with this board was that the board was doing nothing, just displaying a blue screen.

    Just randomly probing around randomly at the buffers going between the video board and the ribbon cables going to the CPU board. The overall goal was just to make sure that the two boards were talking to each other. I stumbled across H1, a 74ls245. The outputs looked really odd so I decided to just replace it. After replacing it, I was able to see the game! The only problem here was that it was missing ALL moving objects (Notice the guy doing the acrobatics is missing).

    The way hypersports works, is that there are two main sections in which graphics are generated (ignoring the parts that generate clocks and whatnot): Background graphics and moving objects. Down the line they are mixed together to create one full image. With this one, since it was just moving objects, I figured that it was either the data getting lost on the way to the mixing, the ROM/RAM is not getting activated, or there was a selection problem when mixing the graphics. 

    The first thing I did was double check the ROMs were being activated and sending data out. With that verified, I double checked that the Konami 083 custom was outputting data to the color PROM. From there it was checking to see if the color PROM was talking, it was so data was at least making it out. This meant that there should at least have been SOMETHING on screen, so the problem had to be with the mixing together.
    Probing around with the enable and select lines of all the multiplexors and the buffers I noticed that the signals 512 and 5̅1̅2̅ were always stuck low and high respectively. That is super odd as this is clock signal used for selecting these. Tracing it back I ended up at B4, which was a 74ls74, a dual flip flop. Looking at the signals, it was getting acting as it was supposed to, but the problem was that it wasn't getting a valid data input. Moving to the next part of that flip flop, I noticed that it wasn't getting a clock, which is not correct here. I traced it back to B3, which takes the 1H and 4H signals and makes a clock out of them. It is always activated so it should be doing something as long as the inputs are valid. Both 1H and 4H were toggling appropriately (they really should have been if I was getting anything on the screen at all as these are timing for video), but there was no output. This was our culprit. After replacing it, we got our objects back!

    After closely looking at the graphics, this board was all set.

  21. SNESNESCUBE64

    Gyruss - No Moving Sprites

    By SNESNESCUBE64,


    Arcade Repair
    First time that I've ever worked on a Gyruss. I have to say, the fact that it had any custom ICs at all kind of intimidated me a little bit. I am not the best when it comes to video issues, but is something I am working on getting better. But anyway, this was a pretty straightforward repair. So the problem at hand was that there were no moving sprites at all. Text would appear, but nothing that moved.

    This is great because it really helps narrow down the circuit. After doing a bit of research online, I found out that there was actually a test rom, which ended up making the repair even easier. What was nice about this, is that it was only adding one additional ROM in an unpopulated socket at J14. After burning the ROM and populating it, it showed that RAM 17C, all the between 11A and 14A, and the ROM at 19E was bad.

    This is a bit suspicious, because that implies that nothing is working at all with the sprite engine, you would think something would be working. Looking at the schematic, all the data gets sent from the main Z80 CPU to a Konami 501 custom IC, then through a buffer to the 17C RAM and beyond. Here's the problem, I don't have an easy way to verify the 501 is working other than checking it's signals. Everything looked fine as far as I can tell with the outputs of that, so I shifted focus to buffer at 13E (a 74ls245). Looking at the signals on my oscilloscope, I could see that it was getting all the signals but a few of the outputs were stuck low. That was awfully suspicious as it really should be toggling.

    After socketing and replacing 13E, the moving sprites were all back!  Great to know that it was the issue tying everything up. I've never played Gyruss before, it was really fun and I wouldn't mind owning one of these cabinets in the future.


  22. SNESNESCUBE64
    This was a quick fix as I've seen this issue before. So I was working on a Data East Laser war pinball machine, and some of the flashers were stuck on. This pinball machine just had all of its bulbs replaced with LEDs, so realistically there shouldn't have been a problem. There were also several stuck on so it was more than just a bad socket or a wiring issue. I identified the bulbs as the following:

     
    Flashers are driven by the solenoid circuits, so it seemed odd that the flashers were for solenoid drivers 6, 7, and 8. Looking at the schematic, I could see that the ones in question all had a warming resistor. What a warming resistor does is that it provides a current to a bulb that is less than what is necessary for it to glow. The idea is that it makes it a little bit quicker to warm up and flash than it would be from a cold start. This can be represented by the highlighted resistors below:

    LED bulbs do not need these warming resistors as they require not only a lower amount of current to flash but because they are LED they turn on much faster. Not even all the flashers have these warming resistors as I pointed out. 
    The solution here was literally to just remove the 660 ohm resistors from circuit. After removing them, they turned off as they should and even flashed as they should.
  23. SNESNESCUBE64

    Donkey Kong - No Boot/Resetting

    By SNESNESCUBE64,


    Arcade Repair
    Yet another Donkey Kong boardset that boots to garbage. This one was a bit more interesting as there were two main problems with this board. Whenever the game does not boot, I always check the ROMs and make sure that the game can read the ROMs if I can check that. The ROMs all checked out good using my programmer, so now it was time to check to see if the game board could read them. The way I check is by using a Fluke 9010a with a Z80 pod. What this does is it emulates the Z80 CPU and allows me to run troubleshooting, such as ROM and RAM checks. To check the ROMs on Donkey Kong, you are going to need to run checksums on the address range of 0000 to 3FFF. Attached are what the checksums should commonly return, a different rom set will return different values, but these are the most common for this type of board.

    After verifying all of the ROMs and that they could be read with no issue. The next check was to check the CPU RAM. There are six 2114 SRAMs (1kx4), resulting in three pairs of two. The way they work on Donkey Kong is that they are set up in pairs to form a full 8 bit bus since the 2114 only stores words that are 4 bits wide. They are separated into the upper and lower bits, the lower being D0 through D3 and the upper being D4 through D7. Often times when these fail, the data will be messed up for only one. 

    While running through my RAM checks, I found that there were two RAM chips that had actually failed: 3C and 4B. After replacing these, it now passed all of the RAM tests, but it still wasn't booting. It looked like it was actually resetting while trying to display part of the title screen.

    What was weird about it was this is that nothing looked out of place with any of the control signals, so I ended up doing some research and found that the TKG4-14 has a watchdog circuit (I believe that's what it is). One way to disable this for testing is to cut the jumper CR3. To my surprise after cutting the jumper, the game booted up! After probing around, I found that pin 5 was actually floating. It ended up being a problem with C164, that cap had tested open, the watchdog was always activating. After replacing it, the CPU board worked exactly as it should.

     
    One new capacitor and two new RAMs and this game is now back in working order, on to the next DK!

     
      
  24. SNESNESCUBE64

    Space Invaders - No boot

    By SNESNESCUBE64,


    Arcade Repair
    Yet another space invaders boardset on the bench, this time it isn't a deluxe boardset. After hooking it up it would boot to only showing one invader, not making it any further than that.

    First thing I do in a case like this is double check the ROM strapping so I can hook up the test ROM. This one was configured for 9316B mask ROMs, so I was able to just drop my test ROM in with no issue. From there, the test ROM told me the the exact issue, it was RAM C.


    Before I flat out replaced it, I swapped C and D to see if the problem moved to D, which it did. This indicated that the 2107 RAM that was now located at D was the issue. Replacing the RAM fixed the issue and the game now boots up with no issues.

  25. SNESNESCUBE64
    On the bench today was a Space Invaders Deluxe that I had grabbed off the pile. This one had unfortunately been a parts unit, as someone had taken several parts off the board, probably to fix another Space Invaders board. It was also pretty dirty, so I had to give it a bath.

    After washing the board, my first thing was to attend to the missing parts. It was missing a 9316 (Really a 74161 counter), a 74ls04, and an Intel 3245. The intel 3245 has become a bit more difficult to get a hold of as it was discontinued quite some time ago. So this was actually an opportunity to try to use a Soviet equivalent, which could be obtained for much cheaper because there is more NOS of that left behind. For this I used the K170AP4. 

    Another part that I noticed was missing was the 180pF tuning capacitor that was used to help the crystal oscillate. Unfortunately, I did not have that particular value available. I did however, have a bunch of 56pF capacitors. You can add capacitance by putting capacitors in parallel, so I solder three of them to leads and soldered them onto the board to get 168pF, which is probably close enough.

    After replacing all the missing parts, I found that the game had a dead short on the 12V rail, which is obviously no good. One way to kind of narrow down a short like this is to utilize your multimeter if it is sensitive enough. The closer you get to the short, the lower the resistance will be. First place I checked was the tantalum capacitors, as those have a tendency to short out. It ended up being one of the tantalums by the RAM (shown below). I removed it and replaced it with an electrolytic that I had since I don't stock tantalums.

    After the power situation was taken care of, I hooked it up to find that it booted to garbage. First thing I did was reseat the ROMs and the CPU as midway loved to use bad single wipe sockets.

    After reseating the chips, the game fired right up and worked as it was supposed to, another board back from the dead! It's always a shame to see people use these things for parts. Sometimes it's difficult, especially with custom chips and whatnot.

×
×
  • Create New...