Williams Joust Playing Blind

So taking the advice by Gamefixer and Mecha, I've adjusted the L351. I thought it might be useful to describe the response.

The image you see in the pictures above is exactly what I see. Because the photo represents a still image, I can imagine someone would think they are seeing the same artifact being imaged over and over across the screen (like when horizontal or vertical hold is way out of fact). But its really not acting that way, meaning the image is not rolling in any way and is stable. As I adjust the L351 the image THEN starts moving and rolling and becomes unstable.

Does that change the analysis at all? For sure, L351 isn't fixing the image.

Might rule out a monitor problem but it means you really have a board problem. My guess is some where in the SYNC chain/timing on the MPU board. Since thats not just one IC its going to take some poling around with a scope or logic probe to figure it out.
 
Okay

My thought
Send the boards it to Eldorado games.
Have them fix and ship them back

Well worth the time and money
 
So after letting this sit for a week or so, I had the chance to remove the boards and started inspecting. I view some heat damage located at H1, H3, and H4 as displayed in the pic below. It looks to my untrained eye that one of the socket legs at H3 would not be making good contact at the board. Remember I was getting a "0" displayed after self check indicating everything was okay at self test.

Anyone have any thoughts why: (1) this wouldn't cause an error on the segmented display? and (2) whether any or all the chips located at the H1, 3 and 4 locations would cause any problems in displaying game images as mentioned initially in the thread?

I continue to believe that this presents as a monitor problem -- but as noted above, the game itself was acting wonky in booting up but would coin up and play blind upon reset, suggesting a board issue and, contrary to monitor issues, no adjustment on the neck seems to do anything except roll and shake the already garbled display)?

(Gamefixer, sent you a PM. Thanks for responding so quickly!)
 

Attachments

  • mpuback markup.jpg
    mpuback markup.jpg
    157.1 KB · Views: 19
  • Joust H3 socket.jpg
    Joust H3 socket.jpg
    140.6 KB · Views: 22
Not exactly sure since I'm relying on memory however those look to be the CPU/RAM data buffers. There are no unused pins on those ICs as far as I can tell (20 pins, 2 pins power, 1 pin clock, 1 pin output enable, 8 pins in, 8 pins out). All three of those look to have been replaced, what revision is the CPU board? Does it have the speed wire factory mod? Does it have the three caps soldered onto the data buffers? Always a problem when trying to fix something that has had prior repair.
 
Heat damage is pretty common on these boards. A lot of the parts are NON LS parts (Low-power Schottky) so they operate a bit hotter.

As far as that open pin on H3... If that pin was truly disconnected I doubt the board would come up at all let alone boot to a "0"
 
From looking at the pic that pin is either pin 8 or pin 13. Both are connected to the data bus according to the schematics.
 
Game fixer: I'm with you. It's making contact or it's not but I don't see how that part of the board would cause the jacked up display issues. When you mention the ram being NON LS, that gets modified with the Williams ram conversion I pmd you about wouldn't it?

Crazy-k: As to version identification, the board reads "5770-09656-00 rev b" below that it reads "4-82", I assume that's a production date. I don't know what the mod would look like that your referencing but FWIW there is a very high gauge wire that weaves around the board starting near one of the ribbon cables. ( I'll post pics to show what I'm referring to and see if it clarifies what you're referring to or if you or someone can identify it. That might be helpful for someone searching later and coming across this post). If that's not factory installed the modification was made unnecessarily difficult in how it was routed because there was definitely a faster way to get from a to b.

Thanks for the info so far.
 
Game fixer: I'm with you. It's making contact or it's not but I don't see how that part of the board would cause the jacked up display issues. When you mention the ram being NON LS, that gets modified with the Williams ram conversion I pmd you about wouldn't it?

Not the RAM being non LS (although its not) the other parts on the board are not LS parts. Like the 3 you mentioned showing heat stress on the back of the board.
 
That empty hole is on the data buffer for bank 2 read access. The write is unaffected. 1-2-1 is RAM error bank 2. I would expect the game to be able to boot and run in that situation. At one time, I disassembled the Stargate program and found that a lot of the scratch memory is in bank 3. Either way, verifying that that pin is connected would be my first priority. The B revision is, I think, the original Robotron which has two speed wire mods for the BS*BA inversion needed by the Special Chips.

This board also has three capacitors soldered to the data buffers from the enable to ground. Get a Robotron drawing set and verify the mods have been made. Dave Langley has a good site which shows all of the Williams hardware.

Since the board shows signs of prior, possibly bad, repair, I would first look very closely at the board for other signs of such repairs.
 
So here's a wrap-up. Hopefully Matt/Gamefixer is lurking around here and can clean up my own comments with accurate and coherent helpful information for masses who might come across this thread later.

As noted in the beginning of this thread, when powered on at self test there were no errors showing on the segmented LED data display. The game had sound and appeared to play blind. However, there was no display appearing on the monitor at all except when a hard reset was done at the upper left corner of the RAM board. At that point I could force it to display an image like the first pic below. I couldn't tell if it was it a board problem, the dreaded data cable, the monitor, or all the above.

Gamefixer found the first issue. (I'd be curious if anyone had a good explanation). The ROM board in the cabinet was loaded with the twelve roms specifically marked "Joust" from the factory. Big problem though. The ROM board literally had no place for the two special Williams chips. I'm not saying they were missing -- I mean there was literally no spot on the board for it. See, Second pic. To my untrained eye it appears to me that that ROM board is completely original with nothing evidencing from the solder on the back that anything was added or modified.

Upon receiving my boards, Gamefixer sent me a pic comparing a known good Joust ROM board (center) with my specimen (left) and, for comparison, a Williams Defender ROM board (right). Clearly this looks more like the Defender Rom set -- but with all Joust ROM chips on it. (In case anyone has ever seen something like this before, there was an extra board provided in the purchase of the cab that looks to be a Williams board. Perhaps it did all the logic the special chips did?) In any event I fail to understand how the game could coin up blind and show no errors as it was when the two Special Chips were not there.

There did not appear to be any solution to that puzzle so Gamefixer gave me a very reasonable deal on a replacement ROM board he had laying around somewhere. For hardly more than the price of the parts, he did the lithium battery conversion and the conversion to the newer 4164 RAM to run on +5. With the ROM board swapped out and the mods done, everything was working on Gamefixer's bench better than new. Problem solved and now, for certain, no further display issue could be attributed to the boards.

Once, I got everything back I threw it back in the cabinet. I received the same garbled display. See, Pic 1. (Very much like the pics in #11 above. Now we definitely had a monitor problem. I'm no monitor expert, but the image you see in the pic is exactly what you'd see on the screen in person, meaning specifically there was no rolling or movement on the screen at all with still images. I futzed around with L351 and turned it more than two full turns and suddenly got a correct display and reasonable picture. (Gamefixer previously called that one too earlier in this thread, with some modest change of a half to full turn in either direction I assumed I was on the wrong path of the decision tree. Instead, it turn's out I just didn't go far enough to make the difference. With a reasonably good display and working boards, it's now worthy of dragging in the house. (Here in Ohio it's colder than a witch's something so it hasn't been thoroughly cleaned, etc. just yet until it gets inside). (I still have a couple of other small issues related to a elevated voltage thing, and some monitor display stuff I want more info on, but I'll see if it's in any other threads and will post separately with that question if necessary).

Lastly, a point of honorable mention. Gamefixer is a good dude and so far as I can tell is extremely knowledgable. As noted above, he tested everything and gave me a more than far price on the missing hardware I needed and couldn't just buy online. His work was fast, thorough and more than reasonably priced. I tried not to contact him much figuring I'd learn what I needed to know when I needed to know it -- but if I did send a PM or email he'd be back to me faster than my kids. Even with the time involved in sourcing parts, I still got my stuff back within a little over a week. I was impressed. All said and done, when I figure out this rep points stuff, he's deserving of something there too. (I never would have figured out that part of my ROM board was literally not there).

Hey Gamefixer, did I miss anything on the boards that should be further noted?
 

Attachments

  • Pic 3 (2).JPG
    Pic 3 (2).JPG
    78.4 KB · Views: 19
  • Boards to mail.jpg
    Boards to mail.jpg
    165.9 KB · Views: 25
  • Boards for compare.jpg
    Boards for compare.jpg
    155.1 KB · Views: 27
  • Mystery Board.jpg
    Mystery Board.jpg
    157.1 KB · Views: 28
Something that I've seen with K4600's when L351 is way out of whack is that you can hear the monitor giving off a high pitch whine. When you get L351 where it belongs that seems to go away. I've had a couple of these where the adjustment on L351 has been more than 4 or 5 full turns off before syncing properly. I imaging that was from constant over adjusting as caps begin to fail. Then when rebuilt it is way off.
 
Hey Gamefixer, did I miss anything on the boards that should be further noted?

Only thing of note was that I think your original ROM board was a Stargate ROM board not a Defender.

That and the BS.BA mod was present on your CPU board but undone. I think what you had there was a Prototype Joust and that extra board does in fact do the work of the blitters on the Joust and later ROM board.

I was going to email Eugene about it and ask but forgot too. If anyone would know what that board is it would be him.
 
I'm pretty sure that extra board is doing the work of the blitters.

I've some old photos of one somewhere on my PC from years ago, no idea where I got them.

Any chance you could post some high resolution photos of the board?
 
As requested Philmurr, two higher res photos are attached. It's definitely Williams hardware but I couldn't tell you what it goes to. Nothing is printed on the board itself except "front" and "rev A." FWIW the cabinet definitely wears red paint or primer underneath the brown. The serial number on this particular cab is 550851.

+ Rep Points added for Gamefixer for his hard work.
 

Attachments

  • mystery board1.jpg
    mystery board1.jpg
    512.1 KB · Views: 33
  • mystery board2.jpg
    mystery board2.jpg
    241.1 KB · Views: 24
As requested Philmurr, two higher res photos are attached. It's definitely Williams hardware but I couldn't tell you what it goes to. Nothing is printed on the board itself except "front" and "rev A." FWIW the cabinet definitely wears red paint or primer underneath the brown. The serial number on this particular cab is 550851.

+ Rep Points added for Gamefixer for his hard work.

Are you interested in selling that board? I would like to compare it with my FPGA code for the special chips to see how close my code is to the actual hardware. And see if it has the XORed bit 4 like SC1's
 
Toledo -- Philmurr and I had a brief similar discussion. Most of what you explained with your "XORed Bit 4" pitch -- which I assume is something other than a passing run (and, more importantly, is hopefully not hugs and kisses!);) is over my head. Literally, to me you're saying nonsense words, except the SC1 -- unfortunately that one I know too well. All that aside I can tell you have a plan of some sort you're thinking about.

Let me state it this way. For me, discovering this thing seems like a really cool novelty to have happened upon. However, honestly it probably deserves a better home. I collect some things, games among them, but not random silicon boards. Therefore, were I to keep it simply to hoard it, it would go in a box on a shelf until I die and I assume my kids would throw it out not knowing what it is -- hell, I don't REALLY know what it is. If I knew where it was going and was convinced that by reverse engineering the technology some caretaker would distribute the information in a way so as to keep more Williams games in play when the special chips go out, I think I could be persuaded. Besides I need as much good Karma as I can get!
 
Toledo -- Philmurr and I had a brief similar discussion. Most of what you explained with your "XORed Bit 4" pitch -- which I assume is something other than a passing run (and, more importantly, is hopefully not hugs and kisses!);) is over my head.

Sorry, I'm describing an error that was inherent in the SC1 blitters that required the game code to compensate for when the blitters were accessed. This error was corrected in the later SC2 parts.

If I knew where it was going and was convinced that by reverse engineering the technology some caretaker would distribute the information in a way so as to keep more Williams games in play when the special chips go out, I think I could be persuaded. Besides I need as much good Karma as I can get!

I have reverse engineered the SC1 and am using it in an FPGA based 1:3.4 scale Robotron game my sons and I built this last summer. (I'll post some picks tomorrow) But, being a perfectionist I'm always looking to improve or verify my work in whatever way I can. I would be willing to post all the information I extract including my current SC1 Verilog code (FPGA code) to allow others to use it to keep the games running as the blitters become harder and harder to come by. I can definitely create a replacement part based on this board (It would be a small PCB with a 40 pin header to plug in to the board and a small CPLD/FPGA to hold the code)
 
With the holidays took me a little longer to post than I planned. It's in the general section under Mini Robotron. Still interested in reverse engineering this board to verify my work as well as document it for the community. Please PM me if you would be willing to allow me to buy OR borrow this board for awhile.
 
Back
Top Bottom