(Topic ID: 122295)

Gott Sys80 MPU Problem

By minnesota13

9 years ago


Topic Heartbeat

Topic Stats

You

Linked Games

Topic Gallery

View topic image gallery

image.jpg

#1 9 years ago

I've got a system 80 version 1 converted to version 2 with the GPE U2/U3 module loaded with a combined file of U2 and U3 starting at address 0x2000 on a 27c512 eprom. The game eprom at Prom 1 is a 2716 loaded with Mars God of War.

The board passes the Quickscan80.

When the board is placed in a MGOW it boots OK without the driver board attached. With the driver board attached, the board will not boot. Without the driver board attached but all other connectors attached, the MPU boots and the game will coin-up and will start a game showing ball one in the credit display.

Attach the driver board and the MPU will not boot.

This game is fully working with another MPU so the machine and driver board can be ruled out as possible sources of trouble.

The voltage is 5.04 VDC at the MPU capacitor by the power connector with the driver board attached.

All ground mods are in place. New cap in the bottom.

What do you think is broken?

#2 9 years ago

No sense, BUT...

I converted a "lot" of SYS80 MPU boards (in early 80' ) and never had a problem like yours
(frankly, I didn't use 27C512.... )

On driver board there is no address or data lines, so... first I'll try pull ALL dip SW's - OFF !!!

Second...do you have some "extra wires" on connector between MPU and driver board ?
(can not remember is there any diference on A1 J4 )

If "first&second" is useless, I'll try repining connector between MPU & driver board to "isolate"
problem ) (removing connections - one by one , with labeling of course

#3 9 years ago

I agree with CP above and will add;

U2/3 are rules for the game, obviously, there's a conflict when the circuit lines are completed.

Assuming the MGoW is a functioning game, go over your work on the U2/3 conversion, look for a solder bridge where one shouldn't be. Also, if this is the first gen board, there should be 4 jumpers and a few traces cut on it to convert it to a second gen. If the conversion was done by you, you'll have to look it over too for an error.

Steve

#4 9 years ago

This has been a problem board. I have more that double checked the U2 and U3 sockets. All is good - now shorts and all pins go to where they are supposed to go.

I tried setting all the switches to off and experienced no change.

The conversion from version 1 to version 2 does have the white wire adds and the trace cuts
outline to make the conversion. I have checked this a number of times.

I have a new MPU to Driverboard cable assembly.

I changed the RIOTs no change.

It may be how I have the loaded the U2 U3 combo.

I combined using:

copy /b U2_80.BIN + U3_80.BIN U2U3.BIN

I burned the combined U2U3.BIN into the 27C512 starting address 02000. The chip verifies with a second combine I made. I am using a Pocket Programmer - an old one that uses the parallel port. I haven't had any problems in the past burning eproms.

I don't know about my source files as I haven't used these in the past for U2 and U3.

I wonder if someone has a check sum of the combined files that I can compare with my result.

#5 9 years ago

My guess is that either the starting address isn't right or, if you are using Ed's adapter, the jumpers aren't right. Just a guess.
--
Chris Hibler - CARGPB #31
http://www.Team-EM.com
http://webpages.charter.net/chibler/Pinball/index.htm
http://www.PinWiki.com - The new place for pinball repair info

#6 9 years ago
Quoted from minnesota13:

When the board is placed in a MGOW it boots OK without the driver board attached.

Boot, or not to boot..that's THE question !! : )

If MPU boots up without driver bord, it's "answer"...

Have no (new) ideas, "The answer" is probably in a little thing "between" two boards

#7 9 years ago

To your information....the combined U2U3 image I use (I merged them with a HEX editor) has checksum $7e18. I always program it in a 2764, a 27C512 wouldn't be my choice.

To make things a bit more clear.....you modified the board to use a single 2716 gamerom in the PROM1 socket (gamerules) in stead of 2 512 byte game PROM's?!

And you use a single Eprom in a GPE piggyback in stead of the original U2U3 PROM's (no modifications needed for that).

If you want, you can have my system80 test Eprom. It also tests the PROM1 socket (with a dedicated 2716 testrom) so you can check if the conversion succeeded. Maybe that gives some new ideas.

Weird problem and like cro_pinman (good technician!) i don't have any new ideas....

Good luck!

Marco

#8 9 years ago

I used the 27C512 as it is required for use with the GPE board for System 80 applications. My check sum matches yours 7E18 for the combined U2 and U3.

The only modifications made to the board are for conversion from version 1 to version 2. I have not made any modification specific to using a 2716 in PROM 1 as I thought the conversion wiring was all that was needed. Are additional changes needed?

I would try the procedure suggested by cro_pinman but I do not have the bifurigated pins for the connector. They are obsolete and prices have increased. I found a supplier at .25 USD each.

What I am thinking is to cut the wiring between the two connectors and splice them back one at a time,
starting with power and ground. There is extra length wiring in a spare connector assembly I have.

I leave tomorrow for the Texas Pinfest so I will not be able to report back until next week.

Thank you all for your suggestions. Your help really appreciated.

#9 9 years ago

Hi Barry,

First, a 27C512 is not required. I typically use a 2764 on Ed's board.
Second, STOP...don't cut any wires! 8-)
I've had to do exactly what you are talking about to figure out which MPU connection was causing a problem. I cut short pieces of business card (or equivalent), and wedged them between the connector pins and the card edge. Thin tape would work too.

--
Chris Hibler - CARGPB #31
http://www.Team-EM.com
http://webpages.charter.net/chibler/Pinball/index.htm
http://www.PinWiki.com - The Place to go for Pinball Repair Info

#10 9 years ago

Hi Chris, great idea. I will try a 2764 on Ed's board. I will have to look in my eprom collection to see if I have a 2764.

Indeed, that's a good plan using pin blocking rather than wire cutting and splicing!

Barry

#11 9 years ago
Quoted from minnesota13:

The voltage is 5.04 VDC at the MPU capacitor

Did you measure +5V voltage on "27C512" board ?

1 week later
#12 8 years ago

There is 5V on the daughter board. I replaced the daughter board as well. I changed the U2+U3 from a 27C512 to a 2764. No change in the problem.

The board continues to pass with the quickscan tester. But after futzing with the board it no longer will boot in the machine but does OK with the quickscan and test chips on the bench.

I tried Marco's test chips. First without the socket test then with the test chip installed in PROM1.

I used a 2764 for the test program installed on the daughter board. It appears to pass all tests and then does a steady blink for testing the RIOTs. It tested the RIOTs and found all OK. I tested the option switches and all was OK.

Sometimes when powered on I do get an error but the blinks are very fast -- it indicates a RIOT failure either the first or second test. But a restart and there is no problem.

Does this flash/blink sequence look OK for Marco's test chip?

I am testing this in a working machine.

#13 8 years ago

Hi Barry,

Does the board boot OK with the lower/center MPU connector removed?
What about the lower/left connector removed?
These are the cabinet and PF switches, respectively.
--
Chris Hibler - CARGPB #31
http://www.Team-EM.com
http://webpages.charter.ne/chibler/Pinball/index.htm
http://www.PinWiki.com - The Place to go for Pinball Repair Info

#14 8 years ago

Hi Chris,

It used to boot with all connectors attached with the exception of the driver board. It would even coin-up.

It has steadily deteriorated to now it will not boot in the game. It came up with all displays zeros - too quick for a boot and only with the small display connector not connected. Seems to indicate a prom problem.

I have moved and replaced the RIOT chips with no change in the problem.

Marco's test didn't see a problem with the PROM socket or the RIOT outputs.

I have the GPE board optioned for use of a 2764 address starting 0x0000 as in an 80B. I have a game 2716 in PROM1. I haven't done any board optioning for PROM1 using a 2716. Since Marco's test program ran in a 2764 and test a PROM1 socket containing a 2716 I am surmising all is well with the current set-up.

I am thinking of sending this bad actor off to you to see if you can determine what is wrong.

#15 8 years ago
Quoted from minnesota13:

Sometimes when powered on I do get an error but the blinks are very fast

With "morning coffie"..new "ideas"

I forgot a lot of "exotic" failures like this, but this could be a problem with establishing + 5V.
( hope that I used the right phrase : )

Try start sys with the RESET on LOW - (semi) FIXED !!
This means that AFTER power_ON you can remove reset from GND,and see what happens

#16 8 years ago

Another shot in the dark, have you tried adding a pull up resistor to the CPU board? Below is a quote from Ed at GPE that he posted in another thread a while ago. It's possible that you may need this resistor installed.

Steve

"Small resistor on the plug into TC1 -- pullup resistor of 3.3K to the Read/Write signal. Some MPU boards run fine without it, some will kind-of work without it, and some refuse to work at all without it.
So... if you leave the connector on the TC1 40-pin DIP and unplug the 7-pin connector from the reset board - does they work then?
If it does work, you need to add a 3.3K resistor to bottom of TC1 between pins 7 and 11 and then leave the plug off of TC1. GPE"

#17 8 years ago

Try what Steve mentioned first.
I have seen boards that work fine with the test fixture attached and no longer work when it is removed. The test fixture has a 3.3K pullup resistor added to pullup the Read/Write line, removing the test fixture removes the pullup resistor. I have seen boards come back to life simply by putting the 40-pin TC1 reset connector back into place (without the plug attached to the reset board at the other end). If this works, you can either leave the TC1 plug in place or install a 3.3K resistor between pins 7 and 11 on the back of the board.

The strange thing is - in first post, you said the board works fine if the driver board is disconnected. To me, this indicates that the CPU itself is running and the problem isn't in the CPU/memory/EPROM section.

Ed

#18 8 years ago

Thank you Steve and Ed. I added a pull-up resistor to the wiring side of the board when I did the modification to bring the board up to rev 2.

What I don't like is the inconsistent way the board seems to work and not work. I have pushed on the chips, daughter board and bent the board during the Quickscan80 operation and the tests continue to pass.

#19 8 years ago

I also tried holding the reset low then letting it go. Unfortunately this didn't make a difference.

#20 8 years ago

Send it up here Barry...I need something outside the "run of the mill" board repair that I can bite my teeth into...

--
Chris Hibler - CARGPB #31
http://www.Team-EM.com
http://webpages.charter.net/chibler/Pinball/index.htm
http://www.PinWiki.com - The Place to go for Pinball Repir Info

#21 8 years ago

My test ROM is running ok...you see an initial flash as soon as the board is powered up. Then:
(111): a long flash inidicating the test is running
(1): a short flash inidating the 5101 tested ok
(12): a burst of 2 flashes indicating the RAM in RIOT U4 tested ok
(123): a burst of 3 flashes indicating the RAM in RIOT U5 tested ok
(1234): a burst of 4 flashes inidicating the RAM in RIOT U6 tested ok
(12345): a burst of 5 flashes indicating socket PROM1 tested ok

Memory tests are preceded by a faint flash caused by accessing the memory chip.

You tested your board already with Ed's quickscan tester so I didn't expect to see RIOT/RAM problems but I was curious about the last test but you performed the modification okay.

Marco

#22 8 years ago

Thank you Marco. I'm sending the board up to Chris to apply another set of eyes and hands to see what is wrong with this board.

Thank you for your test chip information and the confirmation that the rev1 to rev2 conversion appears to be good.

Barry

#23 8 years ago

I have sent this board off to Chris in St.Louis. He should have it in hands in a day or two. It's too bad we live a few hours apart, as it is always helpful to have two working together on a stubborn problem.

Hopefully........Chris will not have report that I brought on the problem myself with poor workmanship or some other "operator error." At this point it will be helpful to know what the root cause of the problem with this board.

Thank you all for your help!

Barry

#24 8 years ago
Quoted from minnesota13:

Hopefully........Chris will not have report that I brought on the problem myself with poor workmanship or some other "operator error." At this point it will be helpful to know what the root cause of the problem with this board.

I'd keep that under wraps anyway Barry...it's not nice to kiss and tell...
I'll get a report out on this ASAP.

Regards all...
--
Chris Hibler - CARGPB #31
http://www.Team-EM.com
http://webpages.charter.net/chibler/Pinball/index.htm
http://www.PinWiki.com - The Place to go for Pinball Repair Info

#25 8 years ago

I just learned from Chris that the 6532s were or appeared to be counterfeits. Once replaced with known working 6532 chips the board worked mostly fine.

He found that the board clock would sometimes run 2X speed and sometimes cause the board not to boot.

He also found a bad 7404 in the switch matrix circuit.

I received the 6532 counterfeit chips as a gift from someone - I believe they found them in their game after purchase.

#26 8 years ago
Quoted from minnesota13:

I just learned from Chris that the 6532s were or appeared to be counterfeits. Once replaced with known working 6532 chips the board worked mostly fine.
He found that the board clock would sometimes run 2X speed and sometimes cause the board not to boot.
He also found a bad 7404 in the switch matrix circuit.
I received the 6532 counterfeit chips as a gift from someone - I believe they found them in their game after purchase.

I knew one person that bought a bunch of Chinese 6532's that didn't work. He took them to work (ST Microelectronics, I think) and had them X-ray'd. They were -empty-. Nothing inside the IC but the fame. Actually had that picture as my computer's background screen at work for awhile but have since lost the pix.

Ed

#27 8 years ago

image.jpgimage.jpg

Date code of 32nd week of 2006.
Was Rockwell still making these in 06?
Per some old posts, I don't think so.

Anyway, despite what I'd told Barry earlier, whatever these are, they appear to work. Once I got the board working, I put them back in and it continued to work.

The problems that the board had were...

* Failed switch matrix 7404, which wasn't detectable until the board was running.
* Suspect 4069. It had rust and corrosion. It checked good with my IC tester, but I didn't like the looks of it.
* Failed 7474 in the clock circuit. This was the biggest problem. Sometimes the board would fail to boot. Sometimes it would boot normally. And sometimes it wold boot and run double speed for a while until it crashed. Very odd. The 7474 tested good in my Chinese IC tester, but failed under load. I replaced the chip and all was well. Put the suspect 7474 back in, just to verify, and the game would again run double speed. Definitely a bad 7474.

It's back in service now, ready to go back to Barry.
--
Chris Hibler - CARGPB #31
http://www.Team-EM.com
http://webpages.charter.net/chibler/Pinball/index.htm
http://www.PinWiki.com - The new place for pinball repair info

#28 8 years ago

No, Rockwell wasn't making these in 2006... those are fakes.

Rockwell stopped making semiconductors under the "Rockwell" name in 1998. They spun off the semiconductor business in 1998 and went under the name of Conexant Systems. Conexant sold the 6500 series parts for only a couple years before dropping them but I'm not sure when. However, anything from year 2000 on had the Conexant name and logo on it - no more Rockwell logo's.

Since then, Conexant spun off most of their semiconductor divisions.
Western Design Center has the rights to the old Rockwell 65xx series components and has been around since early 2000's.
Western Design Center also has the rights to 6532's but won't act on them. I asked if they would start making them - they wanted $50K up front as setup charges plus a huge min order after that.

#29 8 years ago

Thanks Ed. I was hoping you'd drop in.
--
Chris Hibler - CARGPB #31
http://www.Team-EM.com
http://webpages.charter.net/chibler/Pinball/index.htm
http://www.PinWiki.com - The new place for pinball repair info

Promoted items from the Pinside Marketplace
$ 83.00
Electronics
PinballReplacementParts
Electronics
From: $ 1.25
Playfield - Other
Rocket City Pinball
Other
From: $ 2.99
$ 34.99
Cleaner
Pinball Pro
Cleaner
€ 175.00
Cabinet - Toppers
Kami's Pinball Parts
Toppers

Reply

Wanna join the discussion? Please sign in to reply to this topic.

Hey there! Welcome to Pinside!

Donate to Pinside

Great to see you're enjoying Pinside! Did you know Pinside is able to run without any 3rd-party banners or ads, thanks to the support from our visitors? Please consider a donation to Pinside and get anext to your username to show for it! Or better yet, subscribe to Pinside+!


This page was printed from https://pinside.com/pinball/forum/topic/gott-sys80-mpu-problem and we tried optimising it for printing. Some page elements may have been deliberately hidden.

Scan the QR code on the left to jump to the URL this document was printed from.