Got this RTYPE IREM M72 board for repair from my friend Andrea, one of my favourite “Shoot’Em Up” ever!

Irem R-Type! WooooooooooooW !!

It’s a side scrolling shoot-em-up arcade game produced by Irem in 1987. The player controls a space fighter named the R-9 to defend humanity against a mysterious powerful alien life-form known as the “Bydo”.

(taken from Wikipedia)

The board was in very good condition and very clean. As usually, the first visual inspection didn’t reveal anything strange. As the owner said the game suffers o some graphics issues. Now it’s time to power up the board and see what’s going on …

STAGE 1 – Foregroud Graphics Issues

The PCB reveals its weakness …

This is bad … very bad. By the way the game is almost running, but all graphics planes and colours were currupted. I unstacked the CPU board from the VIDEO boards, and as it’s not my first Irem M72 repair so i already know were i should look for.

I also included the ROMs because they’re part of the same graphic sub-system, but the chance they are the cause of this kind of issue is almost near to 0 (zero) 🙂

The “Usual Suspects” point my finger on this guy …

I looked at the schematics, the chip marked KNA6034201 (in the red square) is probably an array of a pair of 74LS166 (in the blue square) with reversed inputs. So the fact it should be a 3x dual 8-bit Shift Registers it’s my best guess.

Google as always is my friend so while i’m searching on the internet i landed at this repair log on Jammarcade.net which confirm my thoughts …

R-Type repair log and Irem ‘KNA6034201’ reproduction

I probed inputs and outputs with the scope that confirmed it was bad, two outputs were stuck. So it should be replaced …

I installed an ic socket and replaced it with a good one from my RType PCB and this lead me to some improvements. Now the foreground/sprite layer seems to be fixed, but …

STAGE 2  – Background Graphics Issues

… the starfield in the first level background was corrupted suffering also with jailbars and some kind of priority issue, or like this one in the second level:

That graphics layer is managed same way as the foreground one by another KNA6034201 placed in the bottom half of the VIDEO board. A deeper looking at the circuitiry and ICs involved reveals a very little hope that the custom chip wasn’t damaged this time …

Probing IC5 with the scope confirm the rust made it’s job and damaged the 74LS166, so it should be replaced. It’s not the only IC damaged this way also IC6 and IC14 were rusty and needs to be replaced.

This lead me to a little further improvement. At the same time probing the outputs of the other KNA6034201 the chances it was good was dramatically dropping down. Two outputs were stucked so as last resort i desoldered also IC15 were the output of the custom chip lead to hoping they’re stucked due to a short on the input side of the 74LS157. But out of circuit it was tested good so no more chance and no more hope, the KNA6034201 should be replaced too.

Testing out of circuit also IC6 and IC14 confirmed IC14 as damaged but IC6 as good, I installed the socket for the custom too. So now i should transplant the other KNA6034201 from my Rtype board too.

STAGE 3 – Mr.Ozzi words of wisdom

Now i don’t remember exacly for how long i’ve been cried at that moment. By the way, i always remember the words the wise Mr.Ozzi said talking about friendship :

“Friends … the more they’re close to you the more they will hurt you!”

… so my karma impose me proceeding with the second transplant. 🙂

Now with two working KNA6034201 the board was playing quite perfectly, but still some graphics issues in the background were present … damn!

OMG! I forgot to replace the rusty IC5, once replaced the game seemed to be fixed …

… but i was wrong! Looking at the title screen sometimes the tiles had missed pixel data, still bad news!

STAGE 4  – Bad Video RAM

The issue wasn’t always present so it was quite annoying to spot and fix. There are four video ram chips on the CPU board, so which one will be the guilty one?

After a lot of power cycles the IC39 video ram confirmed weak outputs on data bus. I removed it with confidence and after i replaced it with a good one, the issue is still there. So i was wrong …

According to the schematics, if the RAM was good, the guilty IC could be the 74ALS374 latches at IC28, removed and replaced:

Now the board was playing fine but my RType PCB wants its KNA6034201 custom chips back! So what can i do now? I need a donor board, but i have no one in my stash. Buying an M72 even damaged is still expensive and the custom chips may be damaged too. As an another option I can buy some reproductions, i know, I found them on ebay but they’re expensive like hell … so only one option was left …. DO IT YOURSELF !!

BONUS STAGE – KNA6034201 Reproduction

After some “rat-nesting” and signals path routing to keep the footprint as small as the Dip40 package is … I see the light at the end of the tunnel! 😉

Now i must send the gerbers to put it in production somewhere, and wait … so STAY TUNED !

After all these efforts, before put the board on the stash waiting for the home made KNA6034201 delivered, it’s time to enjoy this amazing shooter and deep test the board … PLAYER ONE READY?

Yes I’m ready for the second bonus stage … holy sh*t!

BONUS STAGE – The board didn’t boot anymore

The board didn’t boot anymore, so what can be happened? It’s unbelievable. I started from the beginning analysing control signals on main CPU. I found that the RESET signal isn’t working as it should … here is a snippet from the schematics:

The RESET signal was always LOW preventing the Z80 CPU starting executing the code … but what can be wrong in the circuit?

Handling the board, i should bent the capacitor pins this way. I reseated the cap and removed the short. And when i powered up the board it finally played flawlessly!

RTYPE IREM M72 and KNA6034201 … CASE COLSED! … or not? 🙂

Another WonderBoy bootleg board was laying around on my stash, so it’s time to pick it up and check out what’s wrong with it …

The usual visual inspection didn’t reveal anything strange, but once powered up the board seemed to suffer several graphics issues and also the audio part isn’t working as it should.
The sprites were totally missing, and there are some jailbars over some graphics tiles:

The sprite generation circuit on SEGA system 2 is almost handled by a couple of custom ICs and on the bootlegs board they were replaced by the big doughter board in the middle of the PCB

Probing with the scope its connections with the mainboard i found at least one floating input pin and tracing the signal backwards in its path i was lead to a Fujitsu 74LS157 with floating outputs. It’s a common issue for these TTL IC chips, some of them are prone to fail over the ages, they probably loose it’s internal connections between the pins and the die.

Piggybacking it with a good working chip restore the sprite graphics. The piggybacking technique it’s always a good thing if the damaged chip has lost its internal connections, while is totally useless in case of internally shorted pins.

We will take care later about this chip because as you can see, jailbars over the graphic tiles are still there …

The circuit which handles the tiles is close to the EPROMs which stores them, you can find it in this section of the PCB:

Working on this layout of PCBs stack can be quite annoying because you cannot check and diagnose easily the ICs which are laying underneath the doughterboard, but i own another Wonderboy bootleg board, so i tried to swap the small doughterboard hoping it could help me to diagnose the culprit faster … This daughterboard in the bootleg boards replaces three custom Sega ICs in the original boards.

Swapping it resolved the jailbars issue, so i can easily check the TTLs involved, because they should be located in the daughterboard. It ends up with 2 TTL replaced : a 74ls157 with floating outputs and a damaged 74LS299 with internally shorted pins:

I replaced all the three TTL ICs damaged, two in the tiles handling doughterboard and one in the sprite generator circuit

Now the board is playing graphically fine:

The only issue left is in the audio section: sometimes it was totally silent and sometimes it played a some kind of motorboating ramp sound … the culprit can be easily spotted, it’s due to a leaked bad capacitor near the audio power amp:

Once replaced everything was OK in this board! Case Closed 🙂

Got this board from a friend of mine for repair, the game is a side scrolling beat’em up released by Konami in 1993.

The board played fine except for a strange graphic issue involving backgrounds and sprites. Sometimes wrong background graphics were displayed:

… sometimes the sprites had bad data and colours:

… sometimes the sprites were missed at all:

Touching and bending the PCB, partially recover or almost changed the graphics, so the issue should be related to bad traces or lifted pins on one or more IC chip in QFP package a very common issue found on Konami boards.

After a quick and dirty investigation i spotted some lifted and bridged pins on the Konami 055555 TQFP IC. The Asic seems to be a

“5-bit-per-pixel priority encoder with 6 inputs. It can do variable alpha blending and probably a few other special graphics tricks that are currently undocumented” (taken from system16.com)

Probably this bad rework is due to a previous repair attempt, cleaning and reflowing the pins resolves all the graphics issues!

Case closed! 🙂