Tech: Space Invaders

[John Bennett]

Aye up you lot,

Seen as people are talking again thought I would join in the fun

Fingers on your buzzers now.

Who can name the chip that causes the following error :

The screen is rolled around about 50% vertically. i.e. the bases are
in the middle of the screen, and when you fire a bullet it goes off
the top of the screen and comes back on again at the bottom where the
invaders are.

I've had this one before, but didn't have the sense to doccument how I
fixed it last time, doh!

Oh forgot to mention Taito 3 layer pcb, fault is on the bottom (rom)
board.

first person to name the chip wins a bag of used capacitors

Thanks

Luke

__._,_.___
Messages in this topic (1) Reply (via web post) | Start a new topic

DATA Imported from archives: originally posted by Luke Wells (arcadeuk@hotmail.com)

 

[andrewhannay]

It sounds like the MSB of the address line going to the RAM is some how
inverted either during writing to or reading from the RAM. One of the
74ls157s I think.
Andrew

>

 

[Zektor]

It's one of the 7474 or 7404 I forget, in the clock circuit down at one end near the edge connector

 

[andrewhannay]

Possible that piggy backing wouldn't fix the problem.
Did you look at the counter circuit after the clock?
I can't quite remember the circuit operation now, but the main clock feeds a
load of frequency dividers and it's one of the last gates that has got the
problem.
I had a very similar problem on a midway board only the screen was shifted
up (or was it down) 8 or 16 pixels - that turned out to be an standard
inverter IC (7404) that I had replaced with a 74LS04 (or was it a 74HS04)
and was switching too quick. I put a capacitor across the output and it
fixed the problem. Now, I believe that the Tatio boards use the higher speed
74LS stuff anyway but I seem to remember there being little capacitors on
the outputs of the gates. Maybe it's a capacitor that has fallen off or
broken. I love the old 70's stuff.
Andrew

>

 

[John Bennett]

Hello thanks for the sugestions guys,

Anoyingly I have not managed to track this one down yet.

I piggy packed most of the chips on the rom board, but didn't get a
change that I could see relating to the screen position, also there
are no obviously dead/stuck or floating lines on anything on the rom
board.

The fault definatly lies on the rom board!

Whats the best next step? Replace all the 157's one at a time?

I did manage to fix another fault which occured while looking at this.
I started getting intermitant ram 2 errors. Scoping about I found that
the 74174 (IC 44 i think it was in the top left of the CPU board) was
dropping its signal on the D1 line every now and then, replacing this
fixed that fault.

Thanks

Luke

> It sounds like the MSB of the address line going to the RAM is some how
> inverted either during writing to or reading from the RAM. One of the
> 74ls157s I think.
> Andrew
>
>
> >

DATA Imported from archives: originally posted by Luke Wells (arcadeuk@hotmail.com)

 

[andrewhannay]

In addition to what I've just said.
You say it's on the ROM board other things could be:-
There is a load of address multiplexing going on on the ROM board could be a
problem with that.
Is there a fault on the ROM board shorting a line on the other board which
is causing the problem?
Andrew

>

 

[John Bennett]

Oh forgot to say the solution was IC28 74LS83, some of the outputs
were static

Luke
>
> In addition to what I've just said.
> You say it's on the ROM board other things could be:-
> There is a load of address multiplexing going on on the ROM board
could be a
> problem with that.
> Is there a fault on the ROM board shorting a line on the other board
which
> is causing the problem?
> Andrew
>
>
>
> >

DATA Imported from archives: originally posted by Luke Wells (arcadeuk@hotmail.com)

 

[John Bennett]

After fixing the screen wrap fault and playing a few games, the board
has developed the "White lines through the invaders and you can't
shoot them" fault.

On a midway set its easy, short the outputs of each of the bank of
74151's till the lines go black, but where would the fault lie on a
taito 3 layer set?

Don't think i will ever finish fixing this damn set

Luke
> In addition to what I've just said.
> You say it's on the ROM board other things could be:-
> There is a load of address multiplexing going on on the ROM board
could be a
> problem with that.
> Is there a fault on the ROM board shorting a line on the other board
which
> is causing the problem?
> Andrew
>
>
>
> >

DATA Imported from archives: originally posted by Luke Wells (arcadeuk@hotmail.com)

 

[Zektor]

I can't stand Taito SI 3 layer boards, I have about 15 sets with one fault or another....just when you think they are working they go wrong again! the L shaped ones seeem to be better for some reason. Good luck!

 

[andrewhannay]

Still on the 74151s.
On my board set, all my 74151s were socketed - Which was nice.
Andrew

>

 

[andrewhannay]

Ooooooo, I'll know where to come when I fancy another board then.
Andrew

>

 

[Dave2084]

At 09:22 am 10 10 2006, you wrote:
>Still on the 74151s.
>On my board set, all my 74151s were socketed - Which was nice.

If the board has an MB14241 (custom chip) then it may be that that
has gone bad.

If the custom is bad I'll let you (and the world) into a *secret*!

1. Pull the duff custom from the socket and leave it empty.
2. Burn a set of Taito CV invaders ROMs and fit them. The CV set
DOES NOT NEED the custom chip! )

If you need part 2 invaders though you'll need a new custom!

Dave Langley
www.robotron-2084.co.uk

 

[andrewhannay]

Yes,
I must say I have noticed that. I had the Space Invaders 8 Multi game fitted
on my Taito boardset.
During self test at startup, the Multigame complains of a bit error but
Space Invaders plays fine, it's only when I select any of the other games I
get the line across the invaders.
Some clever software bod must have worked out how to do the scrolling
without using hardware.
Andrew

>

 

[John Bennett]

Nice tip Dave!

> At 09:22 am 10 10 2006, you wrote:
> >Still on the 74151s.
> >On my board set, all my 74151s were socketed - Which was nice.
>
> If the board has an MB14241 (custom chip) then it may be that that
> has gone bad.
>
> If the custom is bad I'll let you (and the world) into a *secret*!
>
> 1. Pull the duff custom from the socket and leave it empty.
> 2. Burn a set of Taito CV invaders ROMs and fit them. The CV set
> DOES NOT NEED the custom chip! )
>
> If you need part 2 invaders though you'll need a new custom!
>
>
>
> Dave Langley
> www.robotron-2084.co.uk
>
>

DATA Imported from archives: originally posted by Luke Wells (arcadeuk@hotmail.com)

 

[John Bennett]

I hate them as well, L-shaped are so much easier to work on.

1) It takes 20 minutes with a toothbrush and a bottle of isoprpyl to
get the damn black paint off the top of the chips so you can see what
you are working with

2) The intercconnects are just another failing point that the L's
don't have

3) The schematics are impossible to read. I have to use the L-shape
schematics and then try and work out which chip relates to which on
the 3-layer.

Does anyone have a good copy of the 3 layer schematics? The scans /
pdfs on the net are too poor to work from.

Would be interested in buying if the price is reasonable.

Thanks

Luke

> I can't stand Taito SI 3 layer boards, I have about 15 sets with one
>fault or another....just when you think they are working they go
wrong >again! the L shaped ones seeem to be better for some reason.
Good luck!

__._,_.___
Messages in this topic (15) Reply (via web post) | Start a new topic

DATA Imported from archives: originally posted by Luke Wells (arcadeuk@hotmail.com)

 

[Zektor]

Dave Langley has some good schematics on his website, I think the 3 layer schems are the same as the SI bootleg ones he lists

 

[John Bennett]

You are indeed right they look simular (not sure how much?) and the
quality is 100 times better then the real taito one.

Thanks

Luke
>
> Dave Langley has some good schematics on his website, I think the 3
layer schems are the same as the SI bootleg ones he lists
>

DATA Imported from archives: originally posted by Luke Wells (arcadeuk@hotmail.com)