Partially working Robotron MPU Board...almost there!

Muel

Well-known member
Joined
Feb 4, 2011
Messages
1,536
Reaction score
40
Location
Half Moon Bay, California
Partially working Robotron MPU Board...almost there!

So I've been trying to bring back to life: a Williams "C" board from a Robotron.

It was dead giving a 1-3-1 error and it wasn't the RAM chips.

I replaced ICs 3D and 3E (both 74153) and the board came to life - no 1-3-1 error and now posts a "0"....Woohoo!!

So now that the board lives, I've discovered another problem: it won't hold vertical (can't adjust out on a known working monitor) and the horizontal is shifted to the right 50%.

The decoder ROMs (ROM 6 at 3C & ROM 4 at 3G) are working fine as I pulled them from a known working board.

Thinking that it might be something with the sync, I replaced ICs 3A (7411), 4A (7486) and 6C (7474) and no change whatsoever.

I can see video on the screen - it goes thru the rug pattern, goes to the High Score screen and when the screen tries to refresh to go to the attract screen, it re-boots - back to the rug pattern, then to the scores (this may be a separate issue from the vert & horiz video problem)

ROM board is known working when paired with another MPU board.

Any idea what IC that I should be looking at from here??

Thanks,
-Muel
 
The last time I had one with similar issues, it was the clock signal that is filtered through the 9316s (5C, 5D, 5E and 5F). You need to check the signal coming into pin 3 of chip 6C. That drives the horizontal sync. The 9316s are driven off the 4MHz clock output on pin 9 of chip 7G.

ken
 
The last time I had one with similar issues, it was the clock signal that is filtered through the 9316s (5C, 5D, 5E and 5F). You need to check the signal coming into pin 3 of chip 6C. That drives the horizontal sync. The 9316s are driven off the 4MHz clock output on pin 9 of chip 7G.

ken

Okay, I checked pin 3 of chip 6C and it pulses LOW. I also replaced the (7474) at 7G and pin 9 also pulses LOW.

The board is no longer reseting - it goes thru the attract, coins up and all the good stuff - not sure exactally what I did to fix that, but I'll take it! ...another WooHoo!

However, the vertical is still scrolling and wont hold. Horizontal is still in the middle of the screen as well.

What is curious is both the Vertical and Horizontal are not working properly....
If the 12mhz crystal were bad, wouldn't that also cause both to fail at the same time? ...or would the board not even produce video at all?

On another note, I noticed that on 5C and 5D both pin 2 hardly pulse at all - maybe the problem lies with the 9316's?
 
Last edited:
The 9316s cascade the clock signal. If one of them was bad, that would cause the timing to be off. The down side is that they are extinct chips. It is nearly impossible to find replacements. Of the last 2 that I have seen that went bad, I had to pull replacements off of a donor board.

The best thing to do is to look at the signals on a known working board and compare them to the signals you are seeing on the bad board. That may give you a clue as to what is going on.

Otherwise if you have a logic comparator you can put a known good 9316 in it and see which if any of the other chips are bad.

I have used both techniques successfully.

ken
 
The 9316s cascade the clock signal. If one of them was bad, that would cause the timing to be off. The down side is that they are extinct chips. It is nearly impossible to find replacements. Of the last 2 that I have seen that went bad, I had to pull replacements off of a donor board.

The best thing to do is to look at the signals on a known working board and compare them to the signals you are seeing on the bad board. That may give you a clue as to what is going on.

Otherwise if you have a logic comparator you can put a known good 9316 in it and see which if any of the other chips are bad.

I have used both techniques successfully.

ken

You're right... the 9316's are tough to find...
However, I'm seeing that some people have used a 74161 as a direct replacement for the 9316 on Williams Defender boards.

Have you ever had any experience subbing this particular IC with the 74161?
 
You're right... the 9316's are tough to find...
However, I'm seeing that some people have used a 74161 as a direct replacement for the 9316 on Williams Defender boards.

Have you ever had any experience subbing this particular IC with the 74161?


The pinouts look the same for the 74161. I haven't tried dropping one in, so I don't know how it would react.

If I get a chance, I'll look to see if I have any and try running one in the logic comparator and see if there are any differences.

ken
 
The pinouts look the same for the 74161. I haven't tried dropping one in, so I don't know how it would react.

If I get a chance, I'll look to see if I have any and try running one in the logic comparator and see if there are any differences.

ken

Thanks for offering to check!

If all looks well on the logic comparator, I will pick some up at Jameco on Monday and post the results.
 
Well, I spoke too soon. I have some 74163s which definitely don't work (tried it in the comparator), but no 74161s. So unless Fry's has any, I won't be able to try that until I can get an order from Jameco or somebody else that has them.

When I tried the 74163 in the comparator the vertical hold went all to hell, so that it looks more possible that it is one of the 9316s.

ken
 
Ken,

Thanks for checking.
I have no idea how close a 74163 is to a 74161 - but it sounds like you've duplicated my problem. Must be on the right track!
I'll be picking up some 74161's at Jameco on Monday to see if they are a good sub.

I'll report back with the results.

Thanks,
-Muel
 
Frys of course didn't have any 74161s but they did have a 74HC161. That is the Hi power CMOS version. When I plugged it into the comparator the outputs matched so it may be a valid cross for the 9316s. Because the internal state of the chip is based on its initialized values, to verify the output I needed to cold boot each time I checked a 9316. When I did that the outputs of the 74HC161 matched the 9316. I tried all 4 chips and got the same result.

ken
 
Okay,

More progress!

I fixed the horizontal shift problem where the left edge was in the center of the screen.

I found this just by sheer luck.

I can't even find this trace on the drawing set, but when I measured it with an ohmmeter, I was getting a reading of about 5 ohms? I checked this against a known working board - it read 0 Ohms... Hmmm....

I noticed a little scuff in the trace, not even a deep scratch mind you - but enough to cause resistance. I de-soldered the pads.

I tried a 0 Ohm resistor and - bam! Fixed! Pull the resistor, back to the shifted state.
(See the attached pics)

Now I'm still trying to fix the vertical scroll. That's getting better too. I replaced one of the extinct 9316 ICs with a 74161. Improvement! I get a good (no vertical scroll) rug pattern, then a good "Test OK" message (also no vertical scroll), then as soon as it goes to the high score screen and attract, the vertical scroll starts.

Well, only 3 other 9316's that it could be right?

I'm so close!
 

Attachments

  • C-board Bad Trace.jpg
    C-board Bad Trace.jpg
    55.3 KB · Views: 27
  • C-board 0-Ohm Resistor.jpg
    C-board 0-Ohm Resistor.jpg
    55.7 KB · Views: 26
Last edited:
Good catch on that trace. According to the schematics, that is W1:

attachment.php


The picture makes it look like someone deliberately tried to cut that trace.

When I was looking up something on the Mystic Marathon schematics I noticed something else that bodes well for the 74161s replacing the 9316s. Mystic Marathon uses 74161s for the same function as the 9316s.

ken
 

Attachments

  • jumper w1.JPG
    jumper w1.JPG
    46.1 KB · Views: 126
Yea, I'm thinking that the trace may have been cut intentionally - for what reason tho??

The reason that I'm thinking that the trace was cut intentionally is that I bought a box lot of assorted "untested" Williams MPU (C & D) boards - 2 more of the boards in the lot have similar horizontal problems as well as the 1-3-1 error. The traces on those boards also look like this one that I posted and read also readd 5 Ohms. Some of the other boards in the lot were just fine and the traces were untouched - new memory and the boards booted and played fine.

I don't understand exactally why that trace is there in the 1st place, -and- why someone would bother to cut them!

Anyhow, this discovery just cut down a whole lot of diagnosing time on the 2 other boards.

I'm gonna change out the other two 74161's tonight and that board should be running fine.
 
Last edited:
Cool. That sounds like you are making some progress in getting those boards under control.

ken
 
They have finally given up then, the last time I checked 9316s were xrefed to NTE8613s. It was a bitch but I found a small electronics shop that still had a few. 74161s are way, way cheaper.

ken
 
To invert the sync signals. This would be useful if you run the boards in a JAMMA set up that require the composite sync to be inverted.


Doh! I knew that. :mad: It has been so long since I've had to do that I forgot about it. Damn tequila rotting my brain cells.....

ken
 
Ken,

I have forgotten that 2 or 3 times since I first saw it documented (maybe in RGVAC). So you need to drink more to match me. ;-)

Franklin
 
Back
Top Bottom