C̶r̶a̶z̶y̶ b̶a̶l̶l̶s̶ Rally Race (ttl only).

[Lurch666]

Even though I didn't want to buy any more boards until I fixed the ones I'm currently trying (failing) to fix I saw a TTL only board set on ebay and I've been interested in playing with one since it's a different challenge than a CPU based board.

It was advertised as 'Crazy Rally' and I couldn't find any info about that game.

When it arrived I did some more digging and it might actually be a 'Crazy Balls'.

The PCB has the letters EGS and this lead me to an Italian website that showed the game.

I also found a reference here that shows the same board set I have:

http://www.citylan.it/wiki/index.php/Crazy_Balls_(TTL)

It's possibly got the distinction of being the 1st game that has high score name entry beating asteroids since it appears to have a little keypad on the control panel and it possibly predates the release of asteroids.

Anyway after wiring it up for +5V since I couldn't see anything that needed -5V on this board but it appeared to be dead.

I tried to move the crystal to see if there was any markings apart from the 27.075 on the top but unfortunately it snapped off. I replaced it with a spare 24 Mhz crystal just to see if I would get any sign of life until I got a replacement 27Mhz crystal but the crystal wasn't outputting a signal.

Checking the LS04 it was connected to I found the chip was bad so I replaced it and am now getting activity.

Finding what looked like a sync signal on the edge connector I connected to that and the pin next to it that was showing possible video data but still got a blank screen and I assumed it was because the 24 Mhz crystal was producing the wrong sync timing.

Next I tried to hold the broken crystal in place so the wires made contact with the base and started to get a glimpse of out of sync graphics.

Anyway after testing I have found the board set is active even without a crystal present. I get intermittent graphics when the sync catches up on my test monitor and in my cab where I can adjust the V and H hold I can get a picture that shows me parts of the board are active. I can see a set of zeros and some squares.

What's puzzling me is that the board seems active even without the crystal and I would have thought that the crystal was the timing for all the electronics. There is a 556 next to the LS04 but that can't be providing the logic timing can it and if so why is the crystal needed?

Anyone with any experience of TTL only boards?

Lurch6662021-04-14 13:27:31

 

[Lurch666]

Fitted a new crystal but made no difference. Still puzzled why I had a clock signal without a crystal but ignoring that for now.

Still can't get the sync sorted. I tried tracing where it came from but it goes in multiple directions so I decided to try some other things.

There are 7 proms on this boardset that I'm assuming do the job of eproms as there are no eproms on this board.

I used the slice to check them and 6 of them seem to be coming back with bad output that don't match the files I found in mame. The 6 are the ones on the main board with the one on the daughterboard the only one that came up fine.

I know something connected to the prom output could be throwing out the signals but I decided to remove all 6 so I could check them in my reader and since proms are prone to deterioration having them in sockets will help futureproof the board.

I noticed that 5 of the proms had signs of being soldered with resin traces left where the proms are so I was worried they might have been replaced with different proms as the boardset was sold as 'crazy rally' but the only thing I could find were files for 'crazy balls'.

Anyway after removing the proms all but one checked out with the mame files with the one prom having two locations giving bad data.

Will burn a new prom to replace the bad one then see where that gets me.

 

[Lurch666]

Replaced the faulty prom but got no difference.

Looks like fixing a TTL only without schematics is quite the challenge since I can't start at the CPU and work from there like I normally do.

I've decided to replace the only two ram chips (2111) since I have no way of checking them.

After that I'm going to have to check random parts trying to find any faulty since I don't know where to start.

 

[RaveN]

You're brave, TTL only games are a nightmare to fix. In the past it has taken me ages to fix asteroids/breakout pcbs and that's with great schematics/explanations.

"Finding what looked like a sync signal on the edge connector I connected
to that and the pin next to it that was showing possible video data but
still got a blank screen"

I'm assuming as this is a TTL only game, then it's black and white?. If so, there won't be a separate sync line, it's just a basic single composite-type video output.

Trace this back. The usual way to do it was a point on the board where numerous video data lines met, combined with the sync signal and then this was fed through an ac-coupling capacitor to the edge. If you find this point, you can extract the sync and video separately to feed to your monitor.
However, this isn't always possible (as in the case of Breakout) and it's easier to test with a tv with composite input.

My guess is that tracing back the sync signal is probably the best way to determine the official crystal speed as well.

You should be able to see this on an oscilloscope as well. After the decoupling capacitor you'll see about 0-2v P-P, with horizontal retrace periods.

 

[Lurch666]

Doesn't asteroids have a CPU?

Anyway I did try to trace backwards from the sync/composite pins but they split off in so many directions through various chips I gave up.

Was hoping it would be like space invaders with a 166 or something like that but it's video generation is beyond my experience.

I'll just keep trying to find faulty parts as that's the way I fix most boards but here I'm in the dark as to where to look hence the random search until I understand things better.

I wanted a challenge and I got one.

 

[Lurch666]

So since I have no hope with my limited knowledge of figuring this out logically I've decided to cheat and slicegun it.

I'm using my slice to methodically check every chip I can on this PCB.

Anything that looks slightly out of whack I'm replacing. By doing this I should find every faulty chip (there's been a couple already) and get this working.

Not the most satisfying or efficient way of fixing a PCB but it's the only way I'm going to get this one working.

Looking forward to being able to put up the only youtube video of this board-set working (fame at last!!!!).

 

[Morty]

Would be great to see a youtube of your progress, it's always interesting to see what turns out to be the faults.

Also be good to see how much the slice helps

 

[Lurch666]

Thought of doing youtube videos of actual repairs but can never bring myself to do it so I'll just stick to using it to show faults.

I'm just not the sort of person to appear in a video (my wife has a channel called 'little urban homestead' that she regularly puts videos on but I refuse to appear in them) and my repairs are usually a comedy of errors until I get something right.

With this boardset it's just going to be me replacing chip after chip till the damn thing shows signs of improvement.

 

[jonhughes]

I once read a post on a repair of a similar TTL type board and the repairer mentioned that modern logic chips weren’t compatible because of timings. Thought it was something out of the ordinary and maybe hidden advice, maybe not.

 

[Lurch666]

I have been a little worried that most of the chips in the PCB are 74 and not 74LS but since all I have is the LS I'll keep using them.

If replacing all the bad chips doesn't make a difference then I guess I'll have to try and source some 74 chips.

 

[RaveN]

LS are roughly the same speed, but because of their inherent lower power they provide about 1/5th of the output current. So obviously problems arise with fan-out with other older TTLs. It all depends on where the outputs go.

I'd forget about S or ALS etc as they're too fast and will probably cause timing issues with their shorter propagation delays

If it's just 1 gate to another (or few) then I don't imagine it matters too much what you use.

 

[Lurch666]

I recently bought a small CRT B/W TV to make a mini 'the invaders' tabletop machine.
The PCB has a very strange control wiring setup that uses different common connection so a standards jamma adapter cannot be used:


So then I had the idea of trying this TV with the crazy balls PCB since with 'the invaders' PCB is very difficult to find a TV that can handle the output signal properly the fact that this TV could deal with it had me wondering if would be able to show the output of crazy balls better.

After digging the PCB set out I tried it and still got a rolling picture but it was a little better than before so decided to try and trace from the crystal to see If I could find out where the composite sync signal is generated.
I looked at the schematics of a few other TTL only games to get a general idea and was following traces and checking signals with my oscilloscope when suddenly the picture went steady.

I think I might be close to actually seeing what this game actually looks like:



(Ignore the bunny)

It's not as clear in this photo but there are two sets three zeros with a set of four zeros below that could be scores or balls remaining,a bat in the top left that has what might be a ghost image next to it and lots of white dots moving down the screen.

It might be that the dots are supposed to be static but I have no information so the movement could be part of the game.

The puzzle I have now is why do I only get a good signal when I'm probing a certain pin:


It's the output of a 7473 which is a flipflop. What property of my oscilloscope probe could be changing the signal that causes the signal to be cleaner and thus fixing the sync problem?
Is it pulling it down to GND or something else?

 

[Lurch666]

Been searching for info and found this:

https://www.tilt.it/deb/cda-en.html



So it looks like my previous research was wrong and it is a rally game.
Still got a lot of work to do but at least I now know what it's supposed to look like.

Got a new PSU to replace the one I blew up (see the 'be honest' thread for more details) so I can carry on working on this.

 

[Lurch666]

So first thing was to see if I could find any 74 chips to replace the 74LS chips I had used to replace the bad chips I initially found.
Found a few tucked away in my collection (pile) of chips and got a result changing a 74LS151 with a 74151:




Looks like you were right RaveN-I now have Trees. Now to continue looking to see if I can find any more 74 chips to replace the few 74LS left in this PCB.

 

[Lurch666]

So I figured out how to mess with the timing to get the sync to match the small CRT without having to attach the scope probe to the PCB.
There's a small capacitor next to the crystal (not soldered in at the moment so I can experiment) that was marked 42 and changing the value changed the sync.
After a bit of guessing and swapping out different values I got it to sync on the small CRT at any voltage-with the scope probe attached it would only sync at about 4.5V.


Since in the picture of the cab of rally race I posted earlier the game is in colour I next wanted to see if I could get it to appear on a colour CRT to see if it's a B/W or colour.
So after a bit more experimentation with the capacitor I managed to get a picture on my old commodore 1084 monitor but it's still B/W.

Also after messing with the voltage I found I could get parts of the cars and road to appear:

That looks to me like a multiplexer of some sort is failing so I need to track that down.

Now if you read my earlier comments in this thread I said I replaced one of the proms and I realized today that the prom image I got was from the mame crazy balls romset.
If I removed the prom the road and cars disappeared so I was worried I had lost the original prom which may have had different code in it for this game.
But like I said I have a pile of old chips and I found the original prom:


Not in the best of conditions but considering it's been kept in a plastic bag with a bunch of other chips for two years it's not so bad.
After sorting out the legs there were two missing but enough metal was sticking out of the case to fix them.
Put it in the PCB and got the same result as before-road and cars trying to appear.

So now I'm wondering if the mame romset for crazy balls is mislabeled because someone made the same mistake as me in trying to identify this boardset which is understandable since there's so little info on these..

Anyway I'll read this prom just to verify it has the same contents as the one from crazy balls and start looking for a bad multiplexer.

 

[Mc-Q]

i have converted several B/W games to RGB colour,
are schematics available for this board?

 

[Lurch666]

No. I'm just figuring out the connections by trial and error.
There appears to be a sync signal on one pin and I'm getting the composite output from the pin next to it.
It could be that it needs a separate PCB for colour or I'm using the wrong outputs on the edge connector.
Or it could just be a B/W version of the game.

 

[funk]

Awesome stuff this @Lurch666 envious of the talent and willpower!

 

[Mc-Q]

B/W usually is just a composite signal.
maybe the board has seperate colour and B/W pins

if you have them seperate it will save some time, you can trace the video back to the seperate circuits for each subject (score, player, background etc) and then tap them to individual colours.
long ago i converted a Pong board to RGB that way

 

[Lurch666]

Was wondering if there's any old ttl only games that use composite output that are in colour?
If one of those had a schematic it could give me clues to this baordset.

I've checked all the pins on the edge connector and there's only two showing any sort of signal.
I can kinda figure out which part of this baordset generates what by removing the proms since the proms appear to contain graphic information.