Fluke 9010a and Omega Race PCB Question

bstnguy

Well-known member
Joined
Jan 30, 2008
Messages
1,812
Reaction score
36
Location
Boston, Massachusetts
Anyone have any experience with using a Fluke 9010a to fix an Omega Race PCB?

I have one with repaired minimal corrosion damage (two ICs, one socket, few diodes and resistors, ohmed out all traces) and it's giving me some funny readings on the Bus Test.

The manual isn't much help, so figured I'd toss this up to the group. Error messages I'm getting is:

Active Force @FFFFFFFF - Loop?
Active Force @FFFFFBFF - Loop?
Active Force @400000 - Loop?
Active Force @100000 - Loop?
Active Force @FFEFFFFF - Loop?

And so on until Bus Test Fail.

I'm leaning towards a thermal issue because these errors don't repeat. And if I turn the board off and redo the test, the errors again don't repeat.

Short RAM Test will pass some high and NOVRAM, but will then start failing on repeat.

I'm able to write to high memory, read high memory, then after a couple of minutes, I can read, fail write, re-read failed write which is correct and then fail read, and everything fails. Again, power down, power up does not repeat the failure.

Same on the NOVRAM.

Suggestions? Comments?

Thanks
 
When I first hook up the 9010, I disable the reset. I THEN TRY to Read ram at 4000 to 4bff. IF that fails, I will pull the 373 near the batter and TRY again.
The 373 isolates the main bus from the vector generator circuitry. I sometimes go as far as pulling the ROMS, and the 5101 to ensure nothing but the 4000 series RAM is on the buss. Once I get the 4000 to 4bff working (chip swaps, socket swaps, address decoder) I put the 373 back IN and work on 8000 to 8fff, the vector ram.
Then I will Check out the ROMS and finaly it should boot. Invariably, the vector circuitry is dorked, and I do not believe the 9010 helps muche there.
You cannot test the novram with the ram test, as it is only 4 bits. Without a program IN your 9010, just write and Read and verify.
One thing. Once you get the short ram test done, always do the long ram test. It uncovered a problem on one of my boards.

Here is a quote from a post I made a while back....
typos and all :)

Well, I decided to get back into the shop to work on some omega race boards I have laying around.

I took out the one with the least amount of base damage, and hooked it up. i am using an extender to bring power from my cabinet to my test bench, then using my scope ans an xy display. This works okay, but is not perfect, I'll tell you why later.

First kick at the can, showed lots of watchdoggin going on. I replaced the chip nearest the battery corner, a 373, and a socket for a ram chip near that. Watchdoggin ceased.

I hooked my 9010a into the cpu socket, and looked on the net for a memory map. I tried a ram test from 4000 to 4bff and it failed around 4400. This indicates a problem with the second back of ram chips. A pair of 2114's gives the 8 bits. Since the error was - write 00, read FF, I figured a chip select was buggered. Interestingly enough, the board, when put in the machine, will go through a self test, and tell you some things via either sounds or start light flashing. Looking at the flasshing patterns, I was getting 6, 9, and 14. I replace a ram chip at H1, and reseated both ROMS at E1 and F1. That fixed up the 9, and 14 from the self test but 6, the 4400 address was still buggered.
I hooked up my logic comparitor to PN2, a 74138, as it selects each bacnk of ram cips, and the camparitor indicated that pin 14 was not behaving. This could be the 138, or one of the ram chips dragging it down. Since i had moved the ram chips arond I pulled and replaced the 138.
I used my 9010a to verify I could read and write all ram chips 4000 to 4bff, and 8000 to 8fff. All is well.
However, hooking up my scope to the xy outputs, show a nasty screen. Looks like X is only positive, and Y was noisy. The nois apperaed to be eminating from the op amp tl082 at bc8, so I pulled and swapped it. Y now looks good.
Not really understanding much about these vector genrators, I decided to socke the rest of the op amps and swapped them from x to Y. The problem with X only being positive stayed with X. I then soceted and swapped the A to D converters and again the problem stayed with X. Looking at the vector List archives, I found a post from a few years ago that indicated the MSB going into the DAC could be at fault. Furthermore, I found a ling to a website "the secret life of a vector generator" which enlightened me on the operation of the circuit, and indeed, the MSB going into the DAC was likely to blame. I swapped the 74174 at D6 after my logic comparitor hinted that was a problem with pin 7.

And she roared to Life!
 
thanks for the reply!

I mistakenly thought the NOVRAM was at 4000, I'm able to read/write for a couple of minutes to that.

373 at the battery was replaced, so pulling that to isolate the Vector Ram and then pulling other RAMs to isolate.

I'll also see if I'm getting naything on the scope for the X-Y, the monitor is giving me the spot killer, and occassionaly the spot will drift if I turn the brightness up.

Oh, any Sigs for the RAMs by chance? I'd like to run them through the Fluke and see if I'm getting correct Sigs. The EPROMs verified correctly on an EPROM burner.

Thanks!
 
thanks for the reply!

I mistakenly thought the NOVRAM was at 4000, I'm able to read/write for a couple of minutes to that.

373 at the battery was replaced, so pulling that to isolate the Vector Ram and then pulling other RAMs to isolate.

I'll also see if I'm getting naything on the scope for the X-Y, the monitor is giving me the spot killer, and occassionaly the spot will drift if I turn the brightness up.

Oh, any Sigs for the RAMs by chance? I'd like to run them through the Fluke and see if I'm getting correct Sigs. The EPROMs verified correctly on an EPROM burner.

Thanks!

You can't do signature checks on the RAMs because their content is always changing. The content within ROMSs remains unchanged.
 
Fat fingers. SIGs of the ROMs. :)

Pulled the 245 at the battery, no change to the 4000 RAM. Will pull everything socketed and see if that does anything.
 
UPDATE

Came across a video of a Fluke in use (Adam: Tron ROM Error), and disabled my Control Line Trap.

Bus now tests Okay. Yay!

Found a listing of the ROM SIGs. All came back Failed. All passed compared in the Burner, so something is going on in between.

Can read a location, can write to a location, but re-reading it comes back incorrect. Error repeats upon power down and with same Trap disabled. Looking at the re-read, looks to be a dead data line. Verified it and MD1 and MD3 look to be stuck low. Checked the high ones, and all are correct.

Next step is to check the activity out of the processor, after the 8216 and to the ICs. Will also check each bank of RAM to see if it's specific or all.
 
Last edited:
SUCCESS!

One RAM was completely dead and dragging down a bunch more. Found dead line on a chip select.

A little clean up, a little screen adjust, and voila! It'll be finished!

Pic attached. It's a Omega Race Mini.

Now to see if I can fix the next worst board....
 

Attachments

  • OmegaRaceS.jpg
    OmegaRaceS.jpg
    85.9 KB · Views: 25
Well done. I love it when a classic roars to life! When I said disable the reset, I was refering to the control line thing. Sorry, it's been a few months since I fired it up.
It is also well known that the FLUKE sig is different than anything else. Write down the sig that you get so you can compare it in the future. It seems to be easy to find eprom checksums, but the fluke sig is different.

Mike
 
Yes, I figured that out. I'm still learning how to use the Fluke, and it's proving to be a very useful piece of equipment!!

Very cool game, and pretty tough!!

The screen is a little squished and the Vert Adjust isn't doing enough to fill the screen, so figure I'll put in a cap kit and new transistors. Can't hurt the 15V2000.

Thanks for the help!!
 
I checked my notes for the Fluke ROM sigs, and found (if I can be trusted to read my own notes)

0000-0fff 2AA1
1000-1fff 57EE
2000-2fff 62EE
3000-3fff 5613

and if you do it in one check

0000-3fff DEC1

Mike
 
here are some other values - curious to hear what you learn
http://tech.quarterarcade.com/tech/VM.aspx?g=1976&set=omegrace


ROM Regions:

Region ROM File Start - End CRC 9010A SHA1
CPU1 omega.m7 0000 - 0FFF 424D46E 9B58 cc1ac6c06ba6f6e8466fa08286a0c70b5335af33
CPU1 omega.l7 1000 - 1FFF EDCD7A7D 57EE 5d142de2f48b01d563578a54fd5540e5d0ac8f4c
CPU1 omega.k7 2000 - 2FFF 6D10F197 3F07 9609a0cbeeef2efa10d49cde9f0afdca96e9c2f8
CPU1 omega.j7 3000 - 3FFF 8E8D4B54 4B6C 944192c0f6f0cdb25d492ee9f33959d38a1062f2
CPU1 omega.e1 9000 - 97FF 1D0FDF3A 286F 3333397a9745874cea1dd6a1bda783cc59393b55
CPU1 omega.f1 9800 - 9FFF D44C0814 A9F1 2f216ee6de88bbe09775619003aee2d5aa8c554d
CPU2 sound.k5 0000 - 7FF 7D426017 7D72 370f0fb5608819de873c845f6010cbde75a9818e
 
The CRC routines used in MAME are different from the checksum algorithm used by the fluke...

Would be useful to make an app that does the fluke checksum (and signature) routines, but I don't know if they're well documented.
 
6502 pods seem hard to come by of late....back in May of this year, Art had lowered his price to $145 shipped. His web site still lists them at $309 (ouch)
 
Back
Top Bottom