When your games go on the fritz - Suteneko's repair & mod thread.

Taito F3 Converted Game Cart Conversion:

That is a mouthful but let me explain!

When I bought my Taito F3 ( in early 2019) it came bundled with a “Puzzle Bobble 4” cart.
The picture of the listing was not very good quality and at the time I believed it was a legitimate F3 Game.

However, when it arrived you could easily tell that the labels on it were fake:

I didn’t complain or care too much because the price I paid was exceptionally low and everything worked.

Looking underneath the F3 game we can see that the Taito sticker seal has been peeled off from over the screw in the center of the cart and it is marked as: M20J0116A which is the product ID for Puzzle Bobble 2:

Upon opening the F3 cartridge up we can see that where there should be Mask Rom’s on the top and bottom rows for Graphic and sound Data that we have EPROM’s with uncovered windows:

Upon turning over the PCB you can see all the flux residue that was not cleaned up and the poor soldering by whoever converted this F3 game originally:

Has I was not happy owning a bootleg/conversion, I eventually bought the original Puzzle Bobble 2-4:

So now I had a conversion I no longer required, I decided I would do something special with it and convert it into a game I could not realistically afford and really wanted to play on original hardware and do a good job of it.

For Taito F3 conversions you have to pay very special heed to what PAL you have at position IC21:

This will either be a D77-14 or a D77-15.

D77-14 is used for Program EPROM’s that are 27c2001 (0x40000).
D77-15 is used for Program EPROM’s that are 27c4001 (0x80000).

I have a D77-15 so I need to find a Taito F3 game that uses 27c4001 EPROM’s and of the games I could convert to I decided on Elevator Action Returns.

Now I needed to de-solder all the EPROM’s for graphic and sound data:

Unfortunately at this time I didn’t yet have a de-soldering gun and was resorting to using a manual pump and a cheap hot air station to help pry out the EPROM’s and of all the 336 via’s I accidentally damaged one:

Since I was going to put sockets in place I decided I would do the trace repair on the top of the circuit board with thin Kyna wire so it could easily be concealed under a socket:

Which was threaded through the via, to be soldered on the reverse of the PCB along with the Sockets legs:

With this problem resolved I then cleared the solder mask from IC4 where I needed to place a socket for a position which was not already populated and proceeded to solder in all the required sockets:

I then used my EPROM Eraser to erase all the old EPROM’s:

Using the information from the Mame Taito_F3.cpp I had figured out that to do this conversion I would need to programme the EPROM’s with the data from the contents of the MAME elvactr.zip as follows:


IC04 - 27C160  (100ns) Mame ROM : e02-01.4
IC08 - 27C160  (100ns) Mame ROM : e02-02.8
IC12 - 27C160  (100ns) Mame ROM : e02-03.12
IC17 - 27C4001 (100ns) Mame ROM : e02-16.17
IC18 - 27C4001 (100ns) Mame ROM : e02-10.18
IC19 - 27C4001 (100ns) Mame ROM : e02-11.19
IC20 - 27C4001 (100ns) Mame ROM : e02-12.20
IC32 - 27C2001 (120ns) Mame ROM : e02-13.32
IC33 - 27C2001 (120ns) Mame ROM : e02-14.33
IC38 - 27C160  (100ns) Mame ROM : e02-04.38
IC39 - 27C160  (100ns) Mame ROM : e02-05.39
IC43 - 27C160  (100ns) Mame ROM : e02-06.43
IC45 - 27C160  (100ns) Mame ROM : e02-07.45
IC47 - 27C160  (100ns) Mame ROM : e02-08.47

All the now programmed EPROM’s were then inserted in their appropriate sockets and EPROM windows covered up with electrical tape:

I then tested the cart on my Taito F3 and it booted fine, however I was getting some very strange random flickering single pixels at various certain places in the game as circled in the image below:

This problem is often caused in conversions by EPROM’s whose speed is too slow. Taito F3 hardware requires that everything apart from the two AUDIO PRG EPROM’s at positions IC32 & IC33 need to be running at 100ns or faster.

Since the game was running fine and the audio had no issues I only needed to look at the six EPROM’s that held graphic data (IC4, IC8, IC12, IC43, IC45 & IC47).

I dug out a spare known good 100ns 27C160 EPROM which I used to programme, replace and test each of those IC positions until I found the culprit bad EPROM which was then thrown away and I now had a perfect working version of Elevator Action Returns!

I then removed the old labels from the cart:

And as best as I could reproduced new labels for it making sure to add a note that it is a conversion:

I went to a local print shop and got a few printed out and then added them to my F3 cart:

And now to enjoy my new game:


Holy shit dude. That’s amazing work and exhibits a very high level of dedication and knowledge right there.

Very impressive. Also, that’s definitely the game to do it for too.


What an awesome post! I love seeing how you logically approach each step. Where’d you learn all this cryptic info?

1 Like

For a while I did also consider Darius Gaiden which was also possible but I’ve never really been a fan of that series and thought I’d get way more enjoyment out of Elevator Action Returns.

It has all been from learning as I go, taking everything a step at a time.

Mame’s github repository did the majority of the hard lifting for this project and the studying of various Taito F3 game PCB images that I found via google images so I could see what was different between games.

I wish I knew and understood a lot more.

Had a bit of a hit in my confidence recently after I’ve not been able to revive a Konami Asterix PCB I bought after many many hours of work trying to figure out its problems and making almost zero progress even after fixing a fair few minor things that I found to be bad on it. :sob:

1 Like

I know all about getting your confidence kicked around. I had the same thing happen on a Raiden Fighters Jet PCB I got on the cheap. I’ve also had a model 1 Sega CD I’ve been struggling with for months.

It’s funny after all kinds of successful work one or two rough projects can throw you a bit.

1 Like

Brilliant! Great work.

1 Like

Personally, I think I would’ve made the same choice. Elevator Action Returns is just so unique and interesting, not to take anything away from Darius Gaiden.

1 Like

“Crush the old order!”

Amazingly well presented repair.

Thanks for the great content.


SEGA Mega Drive PAL VA4 Refurbishment & Modding:

The Very Amateur Past:

One of the very first mods I tried to do when I was starting to get back into retro gaming in 2015 was putting a region switch on my Mega Drive, which is a rather simple mod and because of this a good mod for beginners to attempt.

I had no soldering experience and I was using cheap and nasty tools, but it worked and gave me enough confidence to later the same month remove the RF modulator to allow me to add a rear stereo audio jack to my Mega Drive so I would no longer have to have a cable reaching round the front of my Mega Drive to obtain stereo audio output.

The results obviously weren’t the greatest…

Poor solder joints, bad wire choice and barely soldered in frayed wires that could potentially short out:

I’m not even sure now how I even managed to remove the RF Modulator without damaging the via’s as I didn’t even know what de-soldering braid or pumps were then:

I ran speaker wire directly from the front stereo jack connections with no care about interference to the rear of my Mega Drive and didn’t bother to clean up any flux:

I then also did a bad job of connecting ground to the stereo jack via a capacitor leg:

Perhaps due to my mechanical engineering background I am fortunate that I choose good placements for my region switch and rear stereo jack and did very clean installs of them:

Starting Fresh:

In 2019 I decided I would properly refurbish my Mega Drive.

All the previous mods were removed, flux residue was cleaned up and via’s properly de-soldered.

PAL VA4 Mega Drive Recap:

After testing each individual capacitor as I was removing/replacing them I could see that every 100µF & 220µF value capacitor bar one at position C79 was capacitance testing outside of specification tolerances and all the 47µF capacitors were all just on the borderline of the 20% tolerance.


:x:C01: 220µF 16v 8mm
:x:C04: 100µF 10v 5mm
:white_check_mark:C05: 10µF 16v 5mm
:white_check_mark:C13: 10µF 16v 5mm
:x:C23: 100µF 10v 5mm
:x:C24: 100µF 16v 5mm
:white_check_mark:C30: 10µF 16v 5mm
:white_check_mark:C31: 10µF 16v 5mm
:x:C32: 220µF 16v 8mm
:white_check_mark:C38: 10µF 16v 5mm
:white_check_mark:C40: 1µF 50v 5mm
:white_check_mark:C41: 10µF 16v 5mm
:white_check_mark:C42: 1µF 50v 5mm
:white_check_mark:C43: 10µF 25v 5mm
:white_check_mark:C44: 1µF 50v 5mm
:x:C49: 100µF 16v 5mm
:x:C50: 220µF 10v 6.5mm
:x:C51: 220µF 10v 6.5mm
:x:C52: 100µF 16v 5mm
:x:C53: 100µF 16v 5mm
:white_check_mark:C55: 10µF 16v 5mm
:white_check_mark:C57: 10µF 16v 5mm
:white_check_mark:C58: 1µF 50v 5mm
:question:C59: 47µF 16v 5mm
:question:C60: 47µF 16v 5mm
:white_check_mark:C61: 1µF 50v 5mm
:white_check_mark:C62: 1µF 50v 5mm
:question:C63: 47µF 16v 5mm
:white_check_mark:C64: 10µF 16v 5mm
:question:C65: 47µF 16v 5mm
:question:C66: 47µF 16v 5mm
:question:C67: 47µF 16v 5mm
:question:C68: 47µF 16v 5mm
:white_check_mark:C71: 10µF 16v 5mm
:white_check_mark:C72: 10µF 16v 5mm
:x:C74: 100µF 16v 5mm
:white_check_mark:C79: 100µF 16v 5mm
:x:C86: 100µF 16v 8mm

Replacing some Additional “Optional” Components:

I then also replaced the two original 7805 voltage regulators with more efficient 78S05 voltage regulators and because this Mega Drive is used essentially 99.5% of the time in NTSC mode I removed the original PAL 53.203MHz clock crystal (OSCI):

and replaced it with an NTSC 53.693175MHz clock crystal so I could get exact timings for 60hz:

Re-Tension Cartridge Connector Pins:

By using a small flat headed screw drivers I very gently applied a little bit of pressure to each of the cartridge pins to increase the tension as they were a little loose from years of abuse:

Region Modding the VA4 Mega Drive:

On the Mega Drive you will find 4 sets of jumpers labeled JP1-4.

Simply JP1 & JP2 control language, JP3 & JP4 control video output and the regions are set as follows:

REGION Language Video
USA +5V +5V

To get a blank state on a PAL Mega Drive you need to cut the traces between JP3 & GND and JP2 & +5V:


For reference:


For a cleaner install and better routing I decided to run my wires underneath the Mega Drive and thread the wire through the via’s and soldered them on the topside of the PCB:

Using a DPDT ON OFF ON switch we want to connect it up as follows:

  • Left: JP4 or JP3
  • Middle: Ground
  • Right: JP1 or JP2

We then want to connect both the left and right terminals with 1.2K ohm resistors that are in turn tied together at the other end and connected to 5v to create a pull-up resistor so that when the language or video jumpers are not connected to ground through the switch they are “pulled” up to 5 volts.

NOTE: Instead of using a switch you could also try a switch-less mod

Adding a Halt Switch:

If you connect pin 17 of the 68000 CPU to ground the CPU will halt and stop running. This is great to be able to freeze the Mega Drive where you would usually be unable to pause and to be able to get a clean screen in those games that obscure the screen when paused.

So by connecting a SPDT ON/ON switch as follows we can have a halt switch:

  • Left: Floating
  • Middle: Ground
  • Right: CPU Pin 17

NOTE:This switch will not pause SEGA Master System games as they use the Z80 CPU.

Since I had previously installed a region switch which I had positioned to the left on the lower case house, I just needed to drill and file a second hole for this new second switch:

Again I routed the required wire from the 68000 CPU pin 17 underneath the PCB and piggy backed the ground from the region select switch.

Both Switches installed:

Due to the positioning and using slide switches, both switches are well hidden but easily accessible:

Installing a M1 Mini Mega:

To improve my Mega Drive’s audio output I decided I would install a M1 Mini Mega PCB:

Since I already had a stereo audio jack installed I hooked its audio output up to that.

NOTE: The caveat of using a M1 Mini Mega is that the volume slider is now sadly non-functional

Widening the Cartridge Slot:

So I could insert Japanese cartridges without the need of a pass through adapter I decided I would widen my cartridge slot by using a fine file and linishing paper to remove the bare minimum required:

With a Japanese cartridge inserted:

Additional Information:

VDP Pin 50:

Pin 50 of the SEGA 315-5313A is the video sub-carrier (marked in pink) and due to it running very close to the RGB line can cause video interference especially along the Blue line.

VDP pin 50 and RGB traces PCB top side:

VDP pin 50 and RGB traces PCB bottom side:

You can optionally lift leg 50 of the 315-5313A to remove this interference, however you will then be unable to use composite video for the dithering.


If you wish to fully convert a PAL Mega Drive into a NTSC machine you will also need to lift Pin 7 of the CXA1145P and connect it to +5v. This will then correct the composite colour burst frequency to NTSC.

However, if you are using RGB then there is no need to do this step at all.


This is a great write-up. It makes me want to revisit my shitty Genesis mod I did around the same time. I’m pretty terrible though so I’d probably do a worse job now though.

I should at least figure out a way to test caps and replace everything before it’s too late.


Awesome post! Super cool to see your original mod work and your update! Always fun to mark your progress. I need to do a region mod to mine as well.

1 Like

SEGA Mega CD Model 1 Refurbishment:

I recently purchased this PAL Mega CD and like with the Game Gear these are notorious for having bad capacitors that should be replaced, so a refurbishment was in order:

SEGA Mega CD 1690-18 Recap:

After opening up the Mega CD and gently removing both the ribbon cables I was able to remove the main motherboard PCB of the Mega CD:

Since the ML2016 T25 Rechargeable Battery (25 mAh) was dead and because I wanted to socket the BIOS chip I removed both of them along with all the capacitors in one go:

It was hard to notice until the capacitors were removed but the majority of the SMD capacitors had been leaking and causing some corrosion to the contact pads and especially bad are that the C41 and C47 had caused some corrosion to occur on some of the BIOS via’s:

SEGA Mega CD Motherboard PCB Capacitors

:x:C41 10µF 16v SMD LEAKING!
:x:C42 10µF 16v SMD
:x:C43 10µF 16v SMD LEAKING!
:x:C44 10µF 16v SMD
:x:C45 10µF 16v SMD LEAKING!
:x:C46 10µF16v SMD LEAKING!
:x:C47 10µF 16v SMD LEAKING!
:x:C48 10µF 16v SMD
:x:C49 10µF 16v SMD LEAKING!
:x:C50 10µF 16v SMD
:white_check_mark:C51 100µF 6.3v
:white_check_mark:C52 100µF 6.3v

After measuring the original SMD capacitors I was able to find some Nichicon UWX1C100MCL2GB SMD capacitors that are suitable drop in replacements. Sadly these parts are now EOL so stock is limited.

Since these SMD capacitors are tiny (only 3mm Diameter x 5mm Height) they are difficult to work with, so unless you are a bit OCD like me it is probably best just to go with mini electrolytic replacements.

NOTE: Care needs to be taken when removing and replacing the capacitor at position C51 because a resistor is soldered to it underneath the PCB.

I also removed all the capacitors from the Power PCB:

SEGA Mega CD Power PCB Capacitors

:white_check_mark:C1 10µF 16v
:white_check_mark:C2 10µF 16v
:white_check_mark:C3 10µF 16v
:white_check_mark:C4 10µF 16v
:white_check_mark:C5 10µF 16v
:white_check_mark:C6 10µF 16v
:white_check_mark:C7 10µF 16v
:white_check_mark:C8 10µF 16v
:white_check_mark:C9 10µF 16v
:white_check_mark:C10 10µF 16v
:white_check_mark:C11 10µF 16v
:white_check_mark:C12 10µF 16v
:white_check_mark:C13 10µF 16v
:white_check_mark:C14 10µF 16v
:white_check_mark:C40 100µF 6.3v 105c
:white_check_mark:C41 100µF 6.3v
:white_check_mark:C42 100µF 6.3v 105c
:white_check_mark:C43 100µF 16v

All the capacitors on the power board tested good with only a couple around the voltage regulator heat sink having any slight variance but well within acceptable he 20% tolerance.

There are also a few capacitors on the optical drive PCB but at this moment in time I am not going to touch them while the optical drive is functioning.

My take away from this recap is that you can probably ignore recapping the power board but you should 100% remove and replace all the original 10µF 16v SMD capacitors as soon as possible!

I now soldered in a socket for the BIOS chip and re-assembled the Mega CD for a quick test:

and everything was working as expected!

Modding the Mega CD for Multi Region BIOS loading:

With the original BIOS removed I now needed to dump its data contents:

To do this I inserted the BIOS Mask ROM into my TOP3000 universal programmer and set it to 27C1024:


As with SEGA Mega Drive Mask ROM’s the dumped data needs to be byte-swapped to be readable:


After byte-swapping my BIOS dump I got a CRC of: 529ac15a which is identical to other PAL Mega CD 1 BIOS files you can find preserved online.

At this point I could have just burnt a modified universal SEGA Mega CD BIOS. However, there are apparently some minor and very trivial issues with doing that and since my Mega Drive is region modded already I decided I would create a multi-BIOS that would change the Mega CD region based on what region my Mega Drive is set to.

For this I needed to use a 27c4096 EPROM which could hold all three region BIOS ROM’s and use its bank selecting legs 38 and 39 to toggle between them.

This would be achieved by connecting leg 38 to JP3 or JP4 (video) on the Mega Drive and connecting leg 38 to JP2 or JP1 (language) to leg 39 as follows:

BANK 01 Ground Ground USA
BANK 02 +5 Volt Ground JAPAN
BANK 03 Ground +5 Volt PAL
BANK 04 +5 Volt +5 Volt USA

With this figured out I obtained the original USA SEGA CD BIOS and a Japanese Mega CD BIOS from online and byte-swapped them ready to be used.

Then in BINman I loaded them into each file slot in the correct order along with my original PAL BIOS dump:


and then combined them altogether:


This gave me a 4Mbit file with the CRC of ffa52927 that was ready to be programmed to the 27C4096. Leg 38 and 39 then needed to be bent upwards before being inserted into the BIOS socket:

We now need to make a connection from the Mega CD to the Mega Drive utilizing some unused pins on the expansion connectors.

  • Expansion Connector PIN 4 needs to be connected to LEG 39 of the 27c4096
  • Expansion Connector PIN 6 needs to be connected to LEG 38 of the 27c4096

NOTE: After doing this make sure to bend legs 38 & 39 down the outside of the socket otherwise they could potentially short on the metal shielding when everything is put back together.

We now need to look at the interconnecting piece of the Mega CD:

This needs opening up:

The Ferrite bead around the PCB needs to also be removed which just clips together:

On the reverse of the PCB carefully lift up the ceramic capacitor that is in the way and solder wires connecting expansion ports pin 4 and 8 between both connectors in parallel:

With the wires in place you will not be able to re-assemble this interconnect and we need to modify its housing slightly by filing a 10mm wide space with a 2mm depth at around 12mm from the left hand side:

The Mega CD now can be re-assembled.

Finally on the Mega Drive side we need to solder two wires between the expansion connector over to the JP jumpers:

  • Expansion Connector PIN 4 needs to be connected to either JP1 or JP2
  • Expansion Connector PIN 6 needs to be connected to either JP3 or JP4

With everything put back together and hooked up:

Mega Drive region switch set to PAL:

Mega Drive region switch set to USA:

Mega Drive region switch set to JAPAN:

NOTE: Switching Mega Drive regions while the Mega CD is powered up is not advised and can lead to the Mega CD crashing. So Power off the Mega Drive before changing regions.


Replacing the Mega CD Internal Battery:

In the previous post a few of you might have wondered why I didn’t have any mention of replacing the soldered in ML2016 T25 rechargeable battery (25 mAh).

There is a good reason for that and I wanted to create a separate post dedicated to the Mega CD internal Memory and rechargeable battery circuit.

Unfortunately ML (Lithium Manganese Dioxide) type rechargeable batteries are no longer produced and even worse the ML battery charging circuit is not compatible with the available alternatives (LIR and VL).

This leaves us with EOL stock that is running out and has a shelf life of about 10 to 20 years.

ML2016 T25 “tabbed” batteries are already very difficult to find in the wild with non tabbed ML2016 batteries being a little easier to find but getting expensive and require a 16mm coin battery holder that are not themselves that common either.

ML2032 non tabbed batteries on the other hand (which are used in Dreamcast consoles) while also EOL still have a fair amount of stock available in the wild and is the same size as the common CR2032 so a coin battery holder is easily obtain cheaply.

ML2032 batteries also have a higher 65 mAH capacity so they will last longer between charges.

So I bought some ML2032 batteries and soldered in a CR2032 coin battery holder:

NOTE: It doesn’t matter if you can’t solder the positive terminal of the battery holder into both the old T25 positive via’s on the PCB. You only really need one to be connected.

Due to the lack of space around the position of the battery on the PCB it is likely you will need to cut away a little bit of the metal shielding to get it to sit flush again:

In the future when the stock of ML batteries is exhausted we will need to use a diode to disable the charging circuit and then use CR2032 batteries which don’t look like they will be disappearing any time soon and having a compatible battery holder in place already is nice.

SEGA Mega CD Internal Memory Repair:

Initially my Mega CD was saving to internal memory just fine but a few days of use later my saves were gone and the ML2032 battery I had inserted was exhausted.

After placing in a new battery the internal memory was incorrectly reporting it had 21885 blocks of free memory:

But it was refusing to format or save:

When I opened up the Mega CD and was able to test the battery with my multi meter while it was turned on I found that the battery was not charging at all and was draining at a rate of 0.01v per second.


The battery charging circuit was obviously not working.

The Fujitsu MB3790 at position IC6 is responsible for providing power to the SRAM and for charging the ML battery and upon closer inspection of this IC we can see the top of the IC is burnt which is a common sign that the IC has gone bad and is over heating:

However, even without a battery, while the system is powered on we should still be able to temporarily save to the internal memory so it is likely that when the charging circuit went bad it also damaged the SRAM which is a Fujitsu MB8464A-10L.

In the arcade domain Fujitsu IC have a bad reputation of going bad after decades of use and unfortunately for us SEGA decided to use a lot of Fujitsu chips within the Mega CD.

The SRAM used by the Mega CD for internal memory is located at position IC16 and is a 8192x8bit (64Kb) CMOS SRAM in SOP28 form factor with the following pin out:


Since I wanted to find a more reliable non Fujitsu replacement for the SRAM, using the pin out above I was able find a suitable compatible replacement Hyundai HY6264ALJ-10 SRAM to purchase.

Using my hot air station I removed both IC6 and IC16:

The removed faulty IC:

Using de-solder braid I then cleaned up the left over solder from the pads:

I then carefully placed the replacement SRAM IC in position with tweasers:

Using a T18-C2 soldering tip I tagged down the top right and bottom left corners of the IC to hold it in place:

This was repeated for the IC6 replacement and then using drag soldering both IC’s were soldered down:

Complete view of the PCB with the new IC in place:

After re-assembly the ML battery now charges correctly and the internal memory now fully functions properly:


SEGA Game Gear (837-9130) 1 ASIC Refurbishment:

I was recently able to pick up another Game Gear in a bundle from a dumpster diver:

As expected the Game Gear has issues, but at least it powers up and has sound (albeit very low volume):

At the bottom of the original Game Gear screen you will find three IC that are embedded within the plastic housing. These IC are LCD driver chips and if their internal connections start to fail you will get a dark 1/3 portion of the screen per failed IC.

Pressing down hard on the center IC removes the dark section of the screen:

However, as soon as you remove pressure the middle section of the screen goes blank again. You can try hot air to try to re-flow the internal connections but unfortunately this was not working for me and coloured vertical lines on the screen usually mean a new screen will be required anyway.

Checking inside we can see some previous liquid ingress damage, fortunately without any corrosion:

So this goes onto my workbench for a recap, screen replacement and general clean up:

Motherboard Capacitor Summary:

:x:C1  -  33µf (6.3v) ~ TESTED: 490nf and LEAKING
:white_check_mark:C4  -  10µf (6.3v) ~ TESTED: 9.01µf
:x:C11 - 10µf (6.3v) ~ TESTED: 6.5µf
:white_check_mark:C14 - 10µf (6.3v) ~ TESTED: 10µf
:white_check_mark:C42 - 10µf (6.3v) ~ TESTED: 10µf
:x:C43 - 22µf (6.3v) ~ TESTED: 7.5µf
:x:C45 - 4.7µf (35v) ~ TESTED: 4.2µf but LEAKING
:x:C48 - 68µf (6.3v) ~ TESTED: 0.88µf and LEAKING
:x:C49 - 100µf (4v) ~ TESTED: 126µf and LEAKING
:x:C54 - 0.47µf (50v) ~ TESTED: 2.48µf and LEAKING
:x:C55 - 0.47µf (50v) ~ TESTED: 3.10µf and LEAKING
:x:C68 - 100µf (6.3v) ~ TESTED: 122µf

Power Board Capacitor Summary:

:white_check_mark:C5  -  22µf  (35v)   ~ TESTED: 23µf but LEAKING
:white_check_mark:C11 - 100µf (25v)  ~ TESTED: 117µf
:white_check_mark:C13 - 820µf (6.3v) ~ TESTED: 858µf

Audio Board Capacitor Summary:

I was unable to test any of the 5 capacitors on this PCB because they were all BADLY LEAKING electrolytic fluid and I didn’t wish to risk damaging the solder pads so I used flush cutters to cut the capacitors clean off and only then removed the remains of the smd capacitor legs.

I then had to resort to using a Fibreglass pen to clean up corrosion:

After everything was re-capped I tested the Game Gear again, but it was not powering on…

I found the issue was that the 1.28v supply wire was sometimes shorting to Ground inside the connector:

Because the short was inside the connector header I had to cut it off and re-solder the wires to the motherboard manually, which fixed the problem and the Game Gear now powered up again.

I then removed all the no longer required components for a new screen installation and replaced several SMD components using my hot air station as the new screen requires some different resistor values:

Replacing the Screen:

I really wasn’t willing to pay the price of another genuine McWill Game Gear Screen and I knew Chinese clones of it had appeared and were available to purchase for up to half the price.

While looking on AliExpress I came across a Funny Playing Version 2.0B Clone screen that advertised among other things:

  • Power saving design with DC-DC power supply chip.
  • Adjustable Brightness Control

Which got me intrigued as the genuine McWill does not feature any of these and this appeared to be a dubiously “improved” version and not a mere clone so I bought one:

This arrived from China to the UK in only 4 days time and was very well packaged.

Unlike the genuine McWill the screen does not come pre-attached to the screen logic PCB.

Straight away you can notice that this PCB is much thinner, has smaller (and not gold flashed) solder pads and uses smaller SMD (but better rated) voltage regulators.

That said the clone is much better labeled and uses faster rated ram:

Some wire is provided which appears to be solid core 32 or 34 AWG wire. While I did use this as I could thread it easily behind the screen, I think it would be better to use 28 AWG wire instead which I did for the Clock, Sync, Ground and voltage wires as they are the most important.

I almost forgot to connect a wire from T10 to R23!
This small connection is only required for Sega Master System games to display correctly:

Since the screen does not come pre-attached you need to buy some double sided sticky pads with which to affix the screen after everything is installed:

This does make aligning the screen a bit more difficult, but as long as you don’t press the screen down you will be able to gently pry it free to adjust its position.

I then finally attached the two wires required for brightness control to function:

Everything was then re-assembled.

It is rather difficult to show the comparison of brightness settings in photographs but I tried anyway.
Please note that in person the difference is much more pronounced.

Lowest Setting:

Highest Setting:

Personally I can not tell any difference between this clone screen and a genuine McWill.
Hopefully McWill updates his design to include brightness control in the future.


Awesome post as usual! Somehow I had missed your previous post about the Mega CD. Really great stuff there too. I love the how clearly you format your posts. Very inspiring. Keep up the good work and keep em coming!

1 Like

Neo Geo MVS 1 Slot (MV1A) Refurbishment:

This is an old repair/refurbishment log from a few years back that I revisited today to sort a couple of things out, which at the time I just didn’t have the correct tools or experience for.

##Please excuse that this log will be presented somewhat out of chronological order to maintain flow!##

Back in early 2017 I purchased a Neo Geo MVS 1 Slot from China for around £30 including shipping.

The listings picture looked like it was in pristine condition and was a 1 slot with a BIOS Socket, however what I ended up receiving was a huge disappointment.

The BIOS was not socketed so a different model was sent, but even worse…
The plastic housing had “CHINA ONLY SALE” gouged deeply into it:

Upon testing it I found that sound was very scratchy (non functional) and games were not able to save any data to the internal memory, but games booted fine.

After taking off the plastic housing to inspect the MVS closer I found out that the capacitors for audio were in very bad shape, leaking electrolytic fluid and heavily corroded:

The Lithium battery was also speckled with rust and had also started to leak:

I also later noticed that the hard DIP switches were not functioning correctly and even had a broken off switch lever on DIP8 which controls CPU Halt:

I contacted the seller who apologized profusely and said if I sent it back she would get a replacement sent out to me (she was just a third party reseller). However, I really didn’t want to deal with sending it back and waiting again for another one.

So I suggested that I would be happy to repair it myself instead if she would give me a small rebate to allow me to purchase the required parts to do so and unexpectedly she agreed as long as I would provide her a receipt for the components purchased:

The first thing I did was recap the MVS.

Neo Geo MV1A Capacitor List:

  • 470µF 16v x4
  • 220µF 16v x1
  •   47µF 16v x1
  •  4.7µF 25v x5

However, at this point I only had a manual de-soldering pump, was really in-experienced at de-soldering and struggled removing them especially the ones that were attached to large ground planes.

What was worse is that the capacitor leg at position AC6 was badly corroded and broke off the original capacitor that was installed while trying to remove it. Leading to part of the capacitor leg getting stuck inside the via.

Luckily this was a ground so I was able to scrape a bit of the top of the ground plane on the top of the PCB away and surface mounted the capacitor to the top for its ground leg:

Briefly fast forwarding to today, I now own a de-soldering gun and thought I would revisit this and attempt to get that stuck in the via corroded leg out:

Success! And now that capacitor is correctly installed:

Replacing these capacitors fixed the audio issues.

And now returning to 2017:

I then removed the original VG2430 3v rechargeable battery and replaced it with a suitable VL2330:

Now the MVS could save slot data again and keep time.

Next was removing the DIP Switch which I really struggled to remove but after a lot of time and effort managed to remove it cleanly:

I was unable to find a matching blue DIP switch replacement, but I did find a suitable red one to solder in its place:

I then test the replacement Hard DIPs in the service menu and they all function correctly:

Jumping forward a little bit again…

About 8 months later, I wanted to finally put a Universe BIOS into this MVS but the BIOS as previously mentioned was not socketed:

I had a bit more experience now along with a Hakko soldering station and a cheap Chinese hot air-station, however the removal didn’t go perfect and one trace was damaged on the removal of the BIOS:

Luckily only the top side of the via was damaged so after soldering in a socket i was able to connect Kyna wire to the correct socket pin on the underside of the PCB:

and then thread it through a via to the top side of the PCB where I was able to reconnect the damaged trace directly to the Motorola 68000 CPU:

At the time I did not yet have a Universal Programmer and bought a Universe BIOS v3.3 from Razoola which I then used in this MVS.

Later I contacted him to email me a binary file of the newly released v4.0 BIOS which I burnt and instead placed into my then newly acquired 2 slot MVS and left this running on v3.3.

Jumping forward in time briefly again:

So while I have this MVS out today I decided I would update the 27C1024 EPROM.

First I read the data on it and saved a backup of my original v3.3 Universe BIOS:

Then I erased the data, wrote and verified the new v4.0 Universe BIOS:

Back to 2017:

Lastly I wanted to deal with the housing and for this I filed out the gouged letters and using black epoxy putty I filled them all in. I then used fine grade linishing paper to sand everything down level:

Then I spray painted the result in matt black and left it to dry:

The final results:

Hope this was an interesting read!
If people are interested I’ll also post up my MVS 2 Slot refurbishment eventually.


Awesome post! Be sure to show us the 2-slot refurb!

1 Like

Thank you for the encouragement to post logs up @Danexmurder

I have to admit your own recent thread gave me a bit of a kick up the backside to write up a few more and to get on with some outstanding projects.

Spent roughly two and a half hours writing up the MVS 1slot refurbishment post.

I waste way too long sorting out images, taking photographs for things that I didn’t already or to replace bad quality ones and then to sort out the jumbled mess of information into some sort of coherent and logical time line that can be followed.

Quite surprised in myself that I’ve somehow made it to 18 written up logs already and I could easily get the number up to 30+ with existing and past projects.

I think I also have a few that are hidden away in the Arcade Collectors Thread and perhaps I should move some of them over to here so they are easier to locate?


Really glad to know there’s more where that came from! My mod posts do take quite a bit of time too. Lately I’ve been trying to limit my picture taking to 3 at each step so there’s less to wade through.

Posts like this are a huge service to the community. Not only are they interesting content to any retro gamer but they’re potential help to other people looking for a guide on the subject at some point down the road.

Really glad that my own posts could be a small catalyst. I think we’re in a positive feedback loop! Keep em coming!


Since the dark cold days of autumn are swiftly approaching, and a second covid lockdown seems right around the corner, it seemed about time to start a desperate hopeless gamegear repair challenge.
I recently obtained it from a junk lot from Japan, and its clear it had been stored for years with batteries inside, upside down. The battery contact themselves are gone. The powerboard, although heavily damaged might be salvageable, although on a first inspection after a round of cleaning reveals a number of soldered pins floating above the pcb.

The main pcb is going to be a real challenge. First round of cleaning with vinegar and alcohol, reveals no obvious major damage expect the pads for on of the buttons having been completely eaten away. A quick first test after cleaning reveals the unit powers on, the backlight circuit is dead, and the lcd shows a flew flickering signs of life. A lot of solder joints look really dull and need to be redone, probably 50% of the board. My only concern is some of the really small through vias that might have been damaged by the corrosion. Next step is removing all caps, known bad components, and slowly start reworking the board.