Sky soldiers graphic issue

    This site uses cookies. By continuing to browse this site, you are agreeing to our Cookie Policy.

    • Sky soldiers graphic issue

      hello guy, I recently acquired a sky soldiers pbc which was said to have graphic issues.
      the board looks mint physically but has some strange dots all over the picture.
      the board boots normally everything is fine but the graphic issue.
      the jamma edge had some pins work on but nothing to bad.

      I tested in a super neo 29 with big red jamma adapter.
      after looking online i found that most of the board had jumper wires that mine does have.
      also I read in the game manual that it need -5 volt (coin door) not sure
      also it need a negative sync

      I dont know how all that factor in the problem.

      so can you guys give me your opinion and and any leads where to start?
      Images
      • 72692479_374290630142359_107624415784599552_n.jpg

        244.46 kB, 960×720, viewed 106 times
      • 72760534_1366930086819025_6141791366285361152_n.jpg

        286.33 kB, 720×960, viewed 22 times
      • 73425110_542774239883416_3281045316937187328_n.jpg

        169.18 kB, 960×720, viewed 19 times
      • Inked73079877_756675658114633_4208504248507301888_n_LI.jpg

        326.14 kB, 697×929, viewed 35 times
      • yyyyyyyyyyyyyyyyyyyyyyyyyyy.jpg

        446.01 kB, 1,086×815, viewed 35 times
    • I'm pretty sure the issue is due a bad RAM (or more) in the row of eight devices located on the opposite side of edge connector, they should do the line buffer.Try to do piggybacking or replace them (from pictures I can see some have been already socketed)

      The post was edited 1 time, last by caius ().

    • kazanova225 wrote:

      piggybacking ? How do I do that?

      Buy an new, identical chip to the chip on the PCB that you're going to piggyback. Place the new chip on top of the PCB chip in such a way that all the legs line up identically. Make sure the legs on the new chip are bent in slightly so that they make tight contact with the legs on the PCB chip that it's sitting on top of.
    • Got it already ordered 10 chips. But it shipping from china. So November 16 is when I supposed to get them. Until them hope I will be resettled properly. I will keep you updated. In the mean time I have some questions:
      1- sky soldiers required a -5V, if it is not been provided to the board what will be the effect?

      2- sky soldiers manual says that the board takes negative sync, my monitor is sanwa. I have only one sync cable from the jamma, can that basly affect the game?
    • I am not familiar with Sky Soldiers in particular, but the foreground image data looks fine on top of the buggy backdrop. I don't have a schematic handy, but I would expect there is some degree of pixel data order selection that changes when flip is enabled. At the end of that chain it would not surprise me to see a 74LS157, swapping even and odd pixels (since two pixels are represented with one byte). It is possible some of the outputs on an IC like that could be problematic. Data errors every other pixel are sometimes a clue.

      It's possible that some RAMs are used as a line buffer for the backgrounds, though that's a little bit unusual. If addressing is broken there, then that could explain the scrambled appearance when it's not flipped.

      First figure out which graphics bus and corresponding ROMs are giving you trouble. You can rule out the ROM itself if the flip helps. Then, see what logic is downstream of that doing the buffering.

      The post was edited 1 time, last by Hatsune Mike ().

    • I had a similar issue on a P.O.W- PCB (so another SNK game on similar hardware).The pixels spread on screen belonged to sprites and not backgrounds as it may seem on first glance.The issue was due a bad RAM in the row of eight devices, they acts as sprites line buffer.

      The post was edited 1 time, last by caius ().

    • Hatsune Mike wrote:

      I am not familiar with Sky Soldiers in particular, but the foreground image data looks fine on top of the buggy backdrop. I don't have a schematic handy, but I would expect there is some degree of pixel data order selection that changes when flip is enabled. At the end of that chain it would not surprise me to see a 74LS157, swapping even and odd pixels (since two pixels are represented with one byte). It is possible some of the outputs on an IC like that could be problematic. Data errors every other pixel are sometimes a clue.

      It's possible that some RAMs are used as a line buffer for the backgrounds, though that's a little bit unusual. If addressing is broken there, then that could explain the scrambled appearance when it's not flipped.

      First figure out which graphics bus and corresponding ROMs are giving you trouble. You can rule out the ROM itself if the flip helps. Then, see what logic is downstream of that doing the buffering.
      I have been looking for the schematics, so far no luck, I also ordered the ram to test them, waiting to be delivered but it is going to take some time.
    • caius wrote:

      I had a similar issue on a P.O.W- PCB (so another SNK game on similar hardware).The pixels spread on screen belonged to sprites and not backgrounds as it may seem on first glance.The issue was due a bad RAM in the row of eight devices, they acts as sprites line buffer.
      I have been looking for the schematics, so far no luck, I also ordered the ram to test them, waiting to be delivered but it is going to take some time.
    • As @caius suggested, it was indeed one of the line buffer SRAMs that was bad (see attached). There's multiple layers of sprites; the only tilemap is the text layer. It was easy to identify which SRAM was bad on a 'scope because it was outputting random shite at times when there shouldn't be any visible sprites :)

      I couldn't find any schematics for Sky Soldiers but it does share some similarities with Prehistoric Isle, for which I did find schematics.
      Images
      • Inked73079877_756675658114633_4208504248507301888_n_LI.jpg

        68.07 kB, 325×375, viewed 11 times