Centipede nothing but white dots...

thebman80

Well-known member
Joined
May 22, 2006
Messages
7,394
Reaction score
30
Location
Wallingford, Pennsylvania
I have a Centipede PCB that displays nothing but white dots plus two red dots in the center. I have tested all socketed custom chips, processor and ROM's in a separate PCB and they all test fine. I thought they where bad until I realized that my spare PCB has a 200 series ROM images and this PCB has a 300 series ROM images. I just took them all out at one time and they worked in my Spare PCB one at a time just made the screen yellow was really weird. So Has anyone ever come across this one in their repairs? I know there are a few more 74ls chips that looked a little corroded and I have started shotgunning them already. I was hoping someone may have run across this error before to cut down on the needless soldering.


On the bright side of things I did fix a totally dead Centipede board today was just a missing crystal and two bad character ROM sockets.:D

Sorry about the glowing ET fingers in the picture I tried my best to block the flash LOL.
DSCF3814.jpg
 
Last edited:
Yeah there in the center of the grid but there red, other than that just white dots.

I tried to put the thing in test mode on the test bench but nothing happens.
 
I have seen this before. IIRC the inputs (self test, coin, slam, etc) where all tied low due to a bad LS257. I just checked the schematics and it is at K9 on the pcb. Meter the test, slam, and coin c pins. I'll bet that they are all tied to ground. Could be a bad chip or damaged traces.

Brian
 
Well in the cabinet I don't have the test, slam, coin door hooked up because I don't have the molex connector on the coin door. Someone cut the thing off the coin door I have. But my spare PCB works just fine without that hooked up. On Saturday I just sorted through all my wire harnesses and scrap wire and came up with some molex connectors and will fix it with some nonoriginal stuff later.

My guess is that it is the LS257 because all of them on the PCB show signs of rust/corrosion on the legs of the chips. I already replaced a few with some dm74ls257BN's I had, it must have been a bad run of chips because most every other chip except those and some 7474's had the same issue. Looks like I was on the right track just have more to replace.
 
Ok replaced K9 no go as of yet... also replaced L9, M9, M8, P7, also a no go to solve the issue. So I still have D9, D/E10 left of the LS257's hopefully one of them will solve the issue since they also have Corrosion damage. Hell of a bad lot of chips I wounder if other boards with serial numbers close to mine with possible chips from the same batch suffer form this LS257 corrosion issue.

Just for future reference the serial number of my board is 19465 so close serial numbers might have a corrosive batch of LS257's on it.
 
Last edited:
Test to see if pins 3, 6, and 14 of K9 to ground to see if there are shorted. If they are not, the look at the DB4, DB5, and DB6 lines. If they are grounded, you have another issue like shorted caps. If you take a working Centipede pcb, ground the test, slam, and coin c pins and turn it on, you will get that same screen of dots and color squares. Your problem lies with them.

Brian
 
So to test if there shorted use the logic probe? what would it read on the probe if any of those pins are shorted? high/low ect.. And what exactly are the the DB4, DB5, and DB6 lines? I get some of the troubleshooting with boards still not anywhere near a pro yet. :)
 
So to test if there shorted use the logic probe? what would it read on the probe if any of those pins are shorted? high/low ect.. And what exactly are the the DB4, DB5, and DB6 lines? I get some of the troubleshooting with boards still not anywhere near a pro yet. :)

Yes, you can use a logic probe. If they are all low then your problem is before K9 due to the fact that those lines have pull-up resistors. They should all be high. If they are high, the your problem is after K9. Those DB lines are your data bus lines that go to the processor.

Brian
 
All reading high on those pins 3, 6 and 14.

Not sure if this helps but the 2 LS257 chips right under K9 read as follows on those pins: L9 pin 3 High, pin 6 High, pin 14 low: M9 Pin 3 Low, Pin 6 low, Pin 14 Low. Both where replace new already.
 
Last edited:
All reading high on those pins 3, 6 and 14.

Not sure if this helps but the 2 LS257 chips right under K9 read as follows on those pins: L9 pin 3 High, pin 6 High, pin 14 low: M9 Pin 3 Low, Pin 6 low, Pin 14 Low. Both where replace new already.

Well, that's good news. Those on K9 should be high. The bad news is we now know the problem is past K9. Could be a bad chip in the address decoding area, bad ram at F2, or a bad LS245 at E2. It could be also be shorted or damaged traces anywhere in between.

Brian
 
dots on screen (convergance test mode) problem solved

I had the same exact problem. I removed the 74LS257 at K9, no change. While checking the address decoder circuitry as suggested in a previous post I found pins 3 and 4 on the 74LS42 at H3 seemed to be stuck high. I then compared those signals to a working board and found that when I shorted those two pins together it created the exact same screen on my good board. I replaced H3 and it solved my problem.
 
I just wanted to bump this as that I finally got 2 hours to look at the board and indeed it was the same IC H3 74ls42N. Now the PCB is working again thanks for the tip. I used my logic prob on that IC and indeed way to many pins read High so desoldered it checked it in my chip tester and read bad.

Finally starting to get better at repairing these boards only took like a decade lol. Plus buying some decent troubleshooting equipment. I would highly recommend buying a a chip tested for 74ls chips to anyone wanting to do repair work.
 
Back
Top Bottom