GGEAD>
<TITLE="Sega/Gremlin Carnival PCB Repair logs">
<style>
p { color:black; }
table { color:black; width:100%;}
<!--body {background-image:url(pics/bg.png);} -->

</style>
</HEAD>

<div style="background-color:white;background-image:url(pics/marquee.jpg);background-repeat:no-repeat;background-position:center top;padding:5px;width:750;border: 0px solid black; margin-left:auto;margin-right:auto;">

<div id=monitor style="background-color:white;margin-left:auto;margin-right:auto;margin-top:190px;color:green;overflow: auto;">

Brian E. Brzezicki<BR>
Copyright 2014<BR>
Version 1.00 06/30/2014<BR>
<P>
<B>Disclaimer</B><BR>
Use this at your OWN risk, do NOT sue me if it breaks your board... if your doing this your board is probably already broken though ;)<P>

<TABLE BORDER=2>
<TR>
  <TH>Game Board #1 June 30 2014</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>All game boots to is this screen of yellow blocks. Owner said game was working, then one day he just got yellow blocks.

</LI>
<A HREF="pics/1.jpg" TARGET=_NEW WIDTH=2592 HEIGHT=1936><IMG SRC="pics/1.jpg"></A>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
  <LI>Did Fluke 9010a long ROM test of RAM range 0x8000 - 0x8FFF got BTS error 80.
  <LI>Swapped the RAM that corresponds to data bit 0x80. When swapped the suspectd bad RAM with a known good RAM the error moved too
  <LI>RAM was a MK4015/MK4027 a 4116 should have worked.. it did not. ordered NOS MK4027s.
  <LI>Replaced bad RAM with a MK4027
  <LI>Game works!
</LI>
</UL>
</UL>
</TD>
</TR>
</TABLE>

<TABLE BORDER=2>
<TR>
  <TH>Game Board #2 1/24/2015</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>All game boots to is a blue grid.
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
  <LI>Did Fluke 9010a long RAM test of RAM range 0x8000 - 0x8FFF got BTS error FF, this means ALL RAM are bad or soemthign else in the memory circuit is bad, which is more likely.
  <LI>Pulled one 4027 RAM and replaced it with a known good one, reran Fluke long RAM test, surprised to find that the Fluke responded with BTS error indicating that all RAMs were bad EXCEPT that one!
  <LI>Looked and RAMS and found they were the wrong type! (actually they should be acceptible replacments, but I learned before that those RAMs will not work on a carnival, for reasons that I'm not sure of.)
  <LI>Replaced with the proper RAMs from a secondary board as well as a few NOS RAMS.
  <LI>Reran RAM long test, OK!
  <LI>Game now has different pattern, but will not reboot (Note on carnival to get the CPU working on a JAMMA test bench you need to make a custom reset circuit (pin 26 on Z80 -> 10K Ohm resistor -> +5) then ground that pin once. That said the game would not boot
  <LI>Verified all ROMS worked OK
  <LI>noticed that CPU socket was lose, removed, found at least 2 pins seemed to fall off as desoldering them, not sure how the Fluke passed the test, but after replacing socket game booted fine.
  <LI>Game works!
</LI>
</UL>
</UL>
</TD>
</TR>
</TABLE>

<TABLE BORDER=2>
<TR>
  <TH>Game Board #3 2/9/2015</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>All game boots to is a blue line pattern
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
  <LI>Did Fluke 9010a long RAM test of RAM range 0x8000 - 0x8FFF got BTS erron bits 0. This means that writes to address 8000 and 8001 are colliding in some way. read from 8000 and found data bit D6 was stuck. 
  <LI>replaced the 4027 RAM corresponding to Data bit 6 (4027 @ U70)
  <LI>Fluke passes ROM test, but game will not boot, have colored pattern.
  <LI>Tested ROM range 0x0000 - 0x1FFF - OK
  <LI>Tested ROMs - 0x0000 - 0x3FFF Fluke Signature 271A - OK
  <LI>Game works!
</LI>
</UL>
</UL>
</TD>
</TR>
</TABLE>

<TABLE BORDER=2>
<TR>
  <TH>Game Board #4 2/9/2015</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>All game boots to is a blue line pattern
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
  <LI>Did Fluke 9010a long RAM test of RAM range 0x8000 - 0x8FFF got BTS erron bits 4. This should correspond to addresses 0x8000 and 0x8010.
  <LI>Read 0x8000 and x8010 both FF
  <LI>wrote 0x00 to 0x8000 noticed that 0x8010 also became 0x00
  <LI>Wrote 0xff to 0x8010 noticed that 0x8000 also became 0xff
  <LI>Looks like something is permanently stuck on address line 4 (which makes sense for a BTS error on address bit 4)
  <LI>did a loop read on address 0x8000 and 0xFFFF, used osciliscope to test output of 74LS08 @ U11 pin #8, looks good
  <LI>looked at next chip in circuit, custom chip @ U57 notied output wierd on A4, tried with another U57 from another board, still same problem. 
  <LI>Traced address line A4 to next chip, the program ROMS, followed continuity from ROM to rom, notice from from row to back row, lost continuity. 
  <LI>Found cut trace on Address line #4 from one row of ROMs to the next
  <LI>Repaired traced
  <LI>Tested ROMs - 0x0000 - 0x3FFF Fluke Signature 271A - OK
  <LI>Game works!
</LI>
</UL>
</UL>
</TD>
</TR>
</TABLE>

<TABLE BORDER=2>
<TR>
  <TH>Game Board #5 2/9/2015</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game is dead, no video at all
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
<LI>Did fluke long RAM test on 0x8000 - 0x8FFF - OK
<LI>Did fluke ROM - 0x0000 - 0x3FFF - FAILED
<LI>Tried upper half or ROMS 0x2000 - 0x3FFF - OK
<LI>Tested 2708 individually from 0x0000 - 0x1FFF, found that the ROM in range 0x1300 - 0x17FF read bad, switched to another socket/range and it read OK
<LI>Replaced the physical socket
<LI>Retested - all ROMs OK
<LI>Not getting any video still
<LI>Replaced 40 pin socket U 57
<LI>Found pin #15 (COMPSYNC) is dead, switched with known good one - OK
<LI>Game works!
</LI>
</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=6>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #6 2/10/2016</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game is dead,blue pattern on monitor
<LI>Game is missing some ROMs and RAMs
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Replaced missing RAMs
<LI>Ran Fluke long RAM test on 0x8000 - 0x8FFF - BTS 1 error
<LI>Replaced RAM associated with data bit 1 (4027 @ U76), when pulling noted that pin was broken off of the original
<LI>Failed ROM test, noticed on reading some addresses the data values were rapidly flipping on a loop
<LI>Removed all ROMS but u33 (address range 0x0000 - 0x03FF), but problem still occurred
<LI>Replaced socket for u33 as all sockets looked bad, transitioning values went away.
<LI>did Fluke ROM long test on range 0x0000 - 0x03FF - Failed
<LI>read some values from addresses 0x0000 - 0x0007 and compared them against the known values, bit 3 (4th) bit was always set low
<LI>using multimeter found continuity between ground and data output and bit 3 (4th bit)
<LI>removed 74LS125 @ u51 (and 74LS125 @ u52, by accident)  and replaced socket, no change
<LI>Pulled ALL 16 rusty ROM sockets from the ROM location. Problem went away
<LI>While replacing sockets, found some bad ROMS, replaced with ROM from another board.
<LI>Did a ROM test on all ROM locations, all passed except the ROM associated with 0x2400 - 0x27FF, Fluke was giving address error. 
<LI>replaced with another ROM, same problem.
<LI>verified physical ROM in another slot - OK
<LI>Replaced corroded CPU socket
<LI>Board works

</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=7>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #7 3/12/2016</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game is dead,blue pattern on monitor
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Plugged in Fluke 9010A with z80 pod, did bus test, all kinds of failures on databus, looks like high 4 bits are bad
<LI>removed one of the 8216s (1/2 of the data bus driver) and tried again, same problem
<LI>removed the current 8216 and placed back in the pulled 8216, no issues
<LI>did not have any new 8216s so pulled one from a old Galaxian board and replaced the one that seemed bad
<LI>Ran RAM test on 0x8000-0x8FFF - OK
<LI>Ran ROM test on ROM RANGE - OK
<LI>Booted game up
<LI>Board works

</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=8>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #8 3/14/2016</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game is dead,blue pattern on monitor
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Plugged in Fluke 9010A with z80 pod, did bus test - OK
<LI>Ran RAM long ROM test on 0x8000-0x8FFF -  got DCD error on Bit 3.
<LI>Noticed that all 16 ROMs sockets were replaced with machined sockets, the soldering on some looks questionable.
<LI>Found a bridged trace, verified bridge between 2 pins, address line a2 and another pin, found bridge and removed
<LI>Ran RAM long ROM test on 0x8000-0x8FFF -  more errors
<LI>Ran ROM signature check on ROM test on 0x0000-0x3FFF -  more errors
<LI>Ended up finding 3 more bridges underneath sockets, fixed those 3

<LI>Ran RAM test on RAM RANGE - OK
<LI>Ran ROM test on ROM RANGE - OK
<LI>Game will still not boot up
<LI>Replaced Z80 with new Z80
<LI>Board works!

</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=9>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #9 5/21/2016</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game works, but won't coin up
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Using Oscilloscope verified that pin #6 on U62 transitioned as coin switch was pressed.
<LI>measured pin #2 on U35 (74LS08) that triggers the CPU reset, noted that that pin 2 (coin up input) was always low (it should be high except when a coin is inserted)
<LI>measured transistor at Q4, found at base, collector and emitter voltage difference to ground was always 0.
<LI>Pulled transistor, found all pins shorted
<LI>Replaced with a new 4401 transistor
<LI>Game works!
</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=10>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #10 6/22/2016</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
  <UL>
    <LI>Game does not work, pure blue screen</LI>
  </UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
<LI>extended ROM video logic board
<LI>fluke z80 bus test OK
<LI>RAM long 0x8000 - 0x8FFF BTS 0xFF pulled a RAM and replaced no change
<LI>looping write test pin #3 it goes low as expected
<LI>replaced 8216s "@ u47 and u48 with known good ones, no change, put back originals
<LI>pulled 82S123 @ U14 (316-0206) and tested - OK
<LI>pulled 82S123 @ u49 (316 - 0633) - OK
<LI>looping read test on address 0x8000 to see if buffer read is happening used pin #4 on 74LS125 @ u54 - read signal OK
<LI>test 8216s data buffers to make sure they are getting the signal pin #15 on either, there is a signal, not sure if its OK
<LI>signal is controlled by 74LS04 @u25 - tested OK
<LI>74LS08 @ U35 looks good
<LI>74LS04 @ U68 looks good

Checking address buffers
<LI>74LS08 @ a10 - OK
<LI>74LS08 @ a11 - OK
<LI>74LS08 @ a12 - OK

<LI>started tested random chips since I had my logic comparator out, this was probably silly as it just distracted by from "good" diagnosis, still that board has so few chips.
<Li>74LS175 @ u19 - OK
<LI>74LS175 @ u20 - OK

<LI>74LS10 @ u77 - OK
<LI>74LS10 @ u39 - OK

<LI>output of RAMs is always floating
<LI>correct data signal seems to get through when writing to RAM
<LI>VRAS/CAS last chip 74LS04 @ U68 is OK

<LI>74LS163 @ u15 - OK
<LI>Tested pins 5 and 6 of u13 - OK
<LI>74LS74 @ u22 - OK
<LI>74LS74 @ u43 OK

<LI>OK need to get back on track, let's use a ROM read/checksum to validate address buffers, and data buffers
<LI>replaced U33 with known good ROM, read 0x0000 - OK
<LI>did fluke checksum on 0x0000 - 0x03FF (u33) got back correct signature, 
<LI>address lines at least out of the first address buffer are working at least A0 - A9
<LI>data bus is working

<LI>now we know the RAMs out is floating on read, why?
<LI>power to RAMs is good
<LI>RAS and CAS are always low on reads that's not correct
<LI>vcas and rras are coming from U19 and U20, those have been checked
<LI>signal coming into U20 is really dirty let's put in a replacement for the bipolar@u14
<LI>socket came off when removing the chip at u14
<LI>CPU socket also came off
<LI>will need to replace both of these

<LI>replaced U14 socket, I now have vRAS ans vCAS on the RAM but still rams not working on all bits @ 0x8000 :(
<LI>replaced 40 pin CPU socket no change
<LI>replaced a 4027 again, I get valid output on this new one, now that VRAS and VCAS is good, so ,all rams must be bad
<LI>swapped in new rams - getting good read/writes - doing long FLUKE ram 0x8000 - 0x8FFF
<LI>Put in single ROM kit - OK
<LI>Low ROMS test OK, top row not testing OK, sockets are horrible.

<LI>removed ROMs installed single ROM kit
<LI>Game works!
</OL>

</TD>
</TR>
</TABLE>

<A ID=11>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #11 2/25/2018</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game does not work, vertical blue lines every other line on screen
<LI><A HREF=https://youtu.be/CuaLrV0D7CE>Repair Video here</A>
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Checked voltage - OK
<LI>Checked CPU clock signal - OK
<LI>Checked /RESET - OK
<LI>Fluke Bus test - OK
<LI>Fluke RAM long - 0x8000 - 0x8FFF - OK
<LI>Did a fluke ROM test on each individual ROM - OK
<LI>Huh this should work
<LI>Replaced ROMs with my <A HREF=http://www.arcade-cabinets.com/board_hacks/VICDUAL-DB/>single ROM kit</A>
<LI>It worked!
<LI>Moved CPU then it started acting up again
<LI>jiggled CPU was able to get it to stop working and start working
<LI>Removed CPU socket, pins were falling out as I de-soldered them
<LI>Installed new CPU socket, game works great with my kit, no issues
<LI>Installed the original ROMS, game works great
<LI>Just for preventative measurement installed new sockets for all sockets EXCEPT the ROM sockets (since the owner is going to use my single ROM kit)
<LI>Game works
</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=12>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #12 3/12/2018</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game works but has yellow garbage on the screen in various places.
<BR><IMG SRC=pics/12.png>
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Fluke bus test - ok
<LI>Fluke RAM long 0x8000 - 0x8FFF - BTS error 0x80
<LI>pulled and replaced MK4027 RAM @ U73
</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=13>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #13 3/16/2018</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game does not work, "checkered screen"
<LI>Board has had alot of work done, all chips seem to be removed
and socketed
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Voltage - OK
<LI>u49 looked wierd, pushed on it and colors on screen changed, 
    pulled old socket and replaced
<LI>Fluke bus test, address bit 11 stuck high
<LI>It was not stuck high but did have a square wave pattern
<LI>Found continuity with chip at u48, a tiny bridge under the CPU socket
<LI>Cut the bridge and replaced the CPU socket
<LI>Fluke bus test - OK
<LI>Fluke RAM long 0x8000 - 0x8FFF BTS C0
<LI>Swapped 8216 @ u47 and u48, ran test again, this time BTS error 0C
<LI>Pulled bad 8216, replaced with another
<LI>RAM now ok
<LI>Game works!
</UL>
</UL>
</TD>
</TR>
</TABLE>

<A ID=14>
<TABLE BORDER=2>
<TR>
  <TH>Game Board #14 4/14/2028</TH>
</TR>
<TR>
<TD>
<P>
<B>Problem:</B>
<UL>
<LI>Game does not work
</LI>
</UL>
<P>
<B>Troubleshooting/diagnosis steps:</B>
<OL>
</LI>
<LI>Voltage - OK
<LI>Clock - OK
<LI>Fluke Bus test OK
<LI>Fluke RAM long 0x8000 - 0x8FFF - DCD bits 5,7,11
<LI>Replaced all RAMS with known good RAM, no change, put back originals
<LI>tested all chips in path from address bits 5,11 U56
<LI>ram sockets were horrible loooking, replaced
<LI>Fluke RAM long 0x8000 - 0x8FFF again - BTS error 7
<LI>Board was scratched up on top and bottom, so did a bunch of continuity checks that I normally would not have done.
<LI>Tested continuity from #37 (A7) on CPU to pin #5 of u11 - OK
<LI>tested continuity from #4 of u11 to #1 - OK
<LI>tesetd 74ls08 @ u11 - ok
<LI>tested continuity from #6 of u11 to pin #6 of U60 - OK
<LI>Tested continutity of #5 of U60 to pin #7 of 4027 RAMs - OK
<LI>Tested 74ls151 with logic comparitor, FAILED
<LI>Pulled 74LS151 and replaced
<LI>Game works!

</UL>
</UL>
</TD>
</TR>
</TABLE>

</BODY>
</div>
</div>