Sidam explorer pcb repair

[Lurch666]

OK another day of ups and downs.

The second explorer board has turned up today. It's a sidam 10800 and it has the two sound roms but unfortunately is also has two AY-3-8912 audio chips as opposed to the one on the 10900 board so getting scramble sound working on the 10900 board looks to be impossible.

But Macro has kindly agreed to part with one of his sidam 10800 boards so I will check and if necessary fix that one so forceful can have two working explorer boards and a bonus 'the end' boardset as well.

Macro can't send the board till December 2nd so will have to wait a while but it will give me a chance to fix the 10800 that arrived today.

At first I was getting no output from the board apart from some beeps from the speaker.

Found some of the timing signals were missing and after a bit of a chase finding a missing signal I finally tracked it to one of the four 161s that handle vertical and horizontal signals. Replaced the faulty 161 and got a screen of garbage.

Still had my test rom but trying that made no difference.

Found the reset on the main CPU was stuck low and this seems to come from an LS02 that has one input held high.

This comes from another 161 near the reset circuit but the 161 seems fine so I'm going to have to figure out how the reset actually works to get this one going.

I can get activity if I ground the pin on the LS02 and the test rom starts up but it's showing some sort of error that I can't read because the graphics are also incorrect so I'm also assuming some of the 2114 rams are faulty.

Onto the 10900 with bad sound I think I might have found what was causing the graphics to glitch out but because it was so intermittent I can't be sure till it's been tested more.

If it is fixed then I'll post the fix and that board will be fully repaired but unable to run explorer because of the sound issue.

 

[Lurch666]

Reset has been fixed.

It turns out the LS161 was actually faulty-I missed it because I hadn't correctly traced the connection between the 161 and the 02.

Pin 11 of the 161 is connected to pin 7 of the 02 and pin 11 of the 161 was stuck high.

By comparing the two boards I have it was obvious where the fault lay.

Replacing the LS161 now has the board resetting properly.

The two chips involved on the reset are shown here in the red boxes for reference if anyone else has this problem.

The board is now running the test rom but failing on the first test so will be swapping out the 2114s that provide the CPU memory next.

Lurch6662020-11-25 12:39:50

 

[Lurch666]

While attempting to find the faulty work ram using my slice last night I noticed that sometimes the self test was passing.

Since it turned out to be too tricky to try and find faulty ram when it was sometimes passing I decided to fix the graphics issue so replaced the two rams in column J and it looks like these are the Vram chips because now the graphics appear correctly but with incorrect colours.

Testing the colour prom shows it to have one bad output so replacing that should fix the colour problem but for now it's getting through the ram tests (sometimes) but failing the sprite test.

Sprites appear to be correct but they are stuck on the bottom corner so will fix those next.

 

[aeroflott]

This is like a Netflix series. Enjoying the new episodes and updates!

Great work here Lurch - can't wait to see this one working again.

 

[Lurch666]

Ended up replacing all the work ram since some was intermittent and now the game is running.

Sound is just random beeps and sprites still not working.

This sprite problem is proving difficult as the sprites are being generated properly but are stuck at the bottom of the screen and I can't yet figure out which part of the circuit is responsible for the position of the sprites.

The sprite ram is good and there is an LS245 feeding into the data lines of the sprite ram so maybe that could be something to do with it.

Will have a search around the net to see of I can find anything but unfortunately searching for 'scramble' brings up lots of stuff that has nothing to do with the game.

Anyone else any idea what part of the scramble circuit handles sprite position since knowing that would help with the sidam?

 

[Lurch666]

So while using the sidam 10900 board that's running 'take off' to check what the readings should be for a board with working sprites it decided to crap out on me again.

The screen started wavering then disappeared completely like there was no signal going to the monitor.

Checking the R,G,B and sync there were signals but for some reason the picture was not showing on my monitor.

I tried starting a game and it was running but the music seemed to be going at double speed.

I compared the sync to the other sidam board and noticed the sync was pulsing twice as fast as well.

This led to the timing circuit and the LS161 chips that divide the main system clock up for the vertical and horizontal timing signals.

On one pair of 161s the signal was 6mhz like it should be but the other pair was getting 30mhz instead of 15mhz-Well there's your problem.

Tracked the issue to the 74LS74 next to the crystal.

Piggybacked a good chip and everything is back again.

The chip must have just died on me suddenly as I wasn't testing anywhere near to it.

So at least I got the working board back to working and now it's back to figuring out sprites.

Found this thread today and it's proving useful:

http://www.ukvac.com/forum/galaxian-video-circuit-description_topic349186.html

Philmurr helping me again.

Wouldn't be able to do as much fixing if it wasn't for people like him.

 

[Lurch666]

After all this time tinkering with these boards I have just
noticed that there are numbers to go with the letters at the top to
identify the chips.

I looked for them at first but they aren't
on the ends they are in the middle so I missed them hence the reason I
haven't been giving the chip locations.

Anyway I finally figured out the sprite issue on the explorer.

It turned out to be the 74LS174 at E2.

Couldn't figure out which part handles sprite position so ended up 'shot-gunning' the board.

Normally
this entails replacing any chip to see if it fixes the board but with
the slice I'm able to test every chip and piggyback any that look
suspect.

The LS174 was showing lots or errors on slice and
piggybacking caused shimmer where the sprites should actually be and
replacing it has fixed all the sprites.

Game appears to be now playing properly but there is no sound so onto that next.

The other board is currently evading any attempt to fix.

I
replaced the LS74 that I thought was the fault but it worked for .002
milliseconds before the screen went wonky and disappeared again.

So
somehow the clocks being generated are double speed. I've got the 6khz
(not mhz like I said in my previous comment) going into the 1st pair of
LS161s (6M and 6L on scramble) but it's output is double of what it
should be.

I've replaced the first 161 (6M on scramble) but it's making no difference and I can't figure out why.

The
slice is saying it has errors on all the outputs and it's even getting
errors showing on pin 10 and that's an input but the chip is known good
so something else is happening here.

The Board sort of follows the scramble schematic as far as the timing generation goes But I still can't figure it out.

More digging required but at least the explorer is looking better.

 

[forcefulfil]

Great work glen.
Great read too. Are you sure you want to continue?

Pcb looking Great too btw.

 

[Lurch666]

Finally nailed the problem with the 'take off' board not working.

I mentioned in the last post about the slice giving errors for the input of the 161 at pin 10 well I now know it indicates an input that doesn't follow normal digital parameters.

Once I checked it with my scope I saw it was being pulled high so it never actually reached 0V-it was going between +1V and +5V.

Now since I had replaced the two chips at 6M and 7M on the scramble schematics (which are 10F and 12F on the sidam) I knew something that was attached to the 1H signal that comes from there must be holding it high.

This turned out to be another LS74 at 12D (4C on the scramble board) and replacing that has got the board working again but now it's resetting after a minute or two.

I'm not bothering about it resetting since I can now use the board to help diagnose the non working sound on the explorer.

Once I get the explorer working 100% I'll see if I can find out what's causing the reset.

 

[Lurch666]

Macros board turned up today-Thanks again dude.

Plugged it in and it fired right up.

Sound not working again but this time it's down to the amp as my audio probe was able to detect the sound.

It was nearly correct but with some issues. Turned out to be a rom problem as once I had replaced the sound roms with the ones from forcefulfils board the sound was perfect.

Haven't been getting very far with the sound fix on because the 10900 board has a different sound layout due to the take off boardset being simpler.

Turns but the sound ram on this new board is in a socket so was able to check the spare ram I had (CM8104) and it's bad so I've ordered some more from Europe.

I'm thinking it's a sound ram issue since the new board with the bad ram in it shows the same symptoms and the explorer with no sound.

Will be a while before the new ram turns up but hopefully this will fix the sound.

The other problem with this new board is it resets at random like the take off does so I'll try and fix that while I wait for the ram.

 

[forcefulfil]

Patience of a saint glen.
Good news about macros pcbs. Like you said a good starter for checking.
Awesome reading as always

Thx
Philforcefulfil2020-12-03 19:14:11

 

[Lurch666]

Think I may have found the cause of the reset problem on the take off and second explorer board.

I found that if I got the voltage to an exact point (about 4.7V at the back of the board) the problem went away.

This is usually down to memory and I remembered when sorting out earlier problems I swapped the work ram with the graphics/sprite rams and this I think is when the problem started on the take off but because it was so intermittent I never realised what it was that I had changed to cause the problem to appear since I changed a lot of things.

Anyway swapping the four work ram chips for the four sprite/graphic rams seems to have solved the problem.

What I think happens is when ram gets old it seems to start failing to store numbers correctly but only on a rare occasion so on a test it seems to be working but when a board is running it's constantly changing the contents of the memory so eventually one byte is written wrong and if it's the work ram the game then crashes and the watchdog resets it.

Now the problem ram is either in the graphic/sprite memory small glitches don't crash the game.

It will have to stay this way since I'm out of 2114 ram but the game appears to be stable so it's good enough.

The explorer I got from Macro also has this problem but if I adjust the voltage to as low as possible for the board to run it's stable. I played a loop and a half on it last night and it's running fine.

Fixed the sound by replacing the 470uf cap that's on the sound output so this board is fixed except for needing a precise voltage.

Since the memory isn't socketed I'll just leave it as it is.

@Forcefulfil:

Does your explorer cab have a way to adjust the voltage so if you have a problem with this explorer you can set the voltage to something the PCB prefers?

 

[Lurch666]

Since Forcefulfil doesn't have a way of adjusting the voltage I have socketed all the work ram in the sidam 10800 I got from Macro so if he has problems he can use the work ram from the take off boardset that I know is 100% good.

Replaced the amp on the take off since the sound was a little flaky and got round to removing the sound ram on the explorer with no sound.

Since the sound ram in the explorer that does have sound is socketed I put it into the other PCB (albeit without a socket as it's a none standard size and I am waiting for the correct size sockets to arrive) and got all the sounds.

What this means is that once the replacement ram arrives (and assuming the ram is good this time) all I will have left is replaceing the colour prom and all the boards will be fixed.

 

[forcefulfil]

Absolutely brilliant Glenn!
I cant thankyou enough for all your help here.
And a great write up too.
Also like to thank everyone else who assisted.
Many thx
Phil

 

[Lurch666]

There might be a problem with the replacement memory chip I have ordered to sort out the sound.

Now when i searched for an equivalent since the only place I could find CM8104-P was the same place in Europe I bought the first one from that didn't work I found this datasheet:

http://www.datasheetarchive.com/pdf/download.php?id=58aed133f42dad924fb82e0289c9d921b0a1f3&type=P&term=CM8104

And on page 4 you can see the CM8104 being equivalent to the intel 2101.

This led me to look for these ram chips and found an ebay seller (also in Europe) who sold 2101A ram that's the same size so I figured it would be compatible.

Last night I was winding my way round the net and discovered that Galaxian uses 2101 ram so I thought what I could do is use a chip from my Galaxian bootleg in the sidam so it would be fixed and use the ram coming from ebay to replace my Galaxian ram.

But when I tried it it didn't work. Checking the pins I saw on galaxian GND is connected to pin 8 instead of pin 11 (It's a 22 pin chip) like I expected.

On the sidam board GND is indeed connected to pin 11.

I have been unable to find a pinout for the CM8104 but a pinout of 2101 ram shows GND is pin 8.

What this means is I have been misled into thinking these ram chips are compatible. There are a few listings on ebay that sell 2101 chips as equivalents of the CM8104 but this would appear to be untrue.

Anyway it might be that CM8104 ram is unobtanium.

I have one in the sidam 10900 and one in one of the two sidan 10800 boards but if (when it eventually arrives) this 2101 ram doesn't work I'm not going to be able to replace the faulty ram in the other board.

 

[Purity]

I believe you have the wrong 8104?

I think it is a 1024x8 N-MOS TTL In/Out Static Ram

Do a search for a GTE 8104-3CD

That's an 11 pin chip with GND on pin 11

Zoom in on the chip here:

http://www.wolfgangrobel.de/arcadetalk/img_pcbs/explorer.jpg

Datasheet attached

https://www.ukvac.com/forum/data/uploads/665/DSAP0048964.zip

Purity2020-12-16 16:25:03

 

[Lurch666]

That's probably why the one I got didn't work then

It's hard sometimes to know which part if a chip ID is the important part and I assumed 8104 would do but it looks like it needs to be 8104-3CD as Purity pointed out.

Unfortunately a look around hasn't shown anywhere I can get those either.

The search continues.

 

[Purity]

Unfortunately a look around hasn't shown anywhere I can get those either.

Yea don't look easy to find

 

[Macro]

You could make a little adapter PCB to put a 6116 in there. just need to do a little swop around from pins 17 onwards

I *think* you can connect /OE and /CS together and just wire the other pins to their namesakes - just put the extra address line low.

(as long as /WE goes low after /CS!)

obviously need to stick a socket in and jury rig one to test first!

 

[Lurch666]

So the 'take off' PCB is back to haunt me.
It's working without sound so I'm thinking I need to replace the ram chip but as stated before there nothing available for that type of ram.

So I'm looking to follow Macros advice and use a 6116 with an adapter but one thing that I don't know is what to do with the OE input since the origonal chip doesn't have one.
If I keep it high it won't be able to output and if I hold it low will it be able to write?