Warlords - Rom F/H6 needed. Mame doesn't have this one. THanks

hmmmmmmmmmmmmm

This would be strange. The same program written two times for 2 different roms?
Is it possible? I'll give it a shot.
 
This would be strange. The same program written two times for 2 different roms?
Is it possible? I'll give it a shot.

I haven't checked but makes sense. It is a very symmetrical graphics layout.

Bill
 
I thought it was strange also. I have 2 boards here with original roms. All 4 are labeled with the 37159 part number. Manual shows they are the same at both locations. Checksum on all 4 roms match. I attached a couple pictures.

This would be strange. The same program written two times for 2 different roms?
Is it possible? I'll give it a shot.
 

Attachments

  • Warlords.jpg
    Warlords.jpg
    94.6 KB · Views: 22
  • Warlords 006.jpg
    Warlords 006.jpg
    94.9 KB · Views: 23
yep verified

Yep.. F H/6 and E6 are the same eprom. Now to find out which ram is showing as bad.
Anyone know of any quick way to test 2101 rams besides placing one on top of the other to see if there is a difference in game play? that didn't work for me .
 
Yep.. F H/6 and E6 are the same eprom. Now to find out which ram is showing as bad.
Anyone know of any quick way to test 2101 rams besides placing one on top of the other to see if there is a difference in game play? that didn't work for me .

Did you just piggyback just one? Try a stack of a few. They are in pairs and odd/even banks so you may have to PB more than one. Scope trick mentioned is also good if you have a scope.

If you are going to keep the game, the 2101s are always dying on Ataris. I'd bite the bullet and replace and socket all of them if two or more are bad. These games are around 30 years old give or take and the original target for logic ICs back in the 80s was a 10 year useful life. Pretty amazing that they are still working. The ones that die prematurely are the ones that run hotter or at higher operation cycles like counters, memory and video signal multiplexers like 74LS157s.

Also make sure you replace all of the EPROM sockets as well and clean the legs of the Eproms before you re-insert them. Those sockets Atari used tarnish and lose retention. Try holding onto something for 30 years. Depending upon if the board saw high moisture or not, I'd also consider replacing the 40 pin sockets as well. This will "bullet proof" your board for years of trouble-free service.

Bill
 
Post a screenshot of graphics corruption and I may be able to tell you which RAM it is (if it is in fact graphics corruption.)
 
It's either K5 or K3 since it's the 3rd "block". Half of a row I'll call a block. The chips are banked such that each takes care of 4-bits for one of four blocks. I'll fire up Mame and figure it out. Here are my notes on which RAM chips are which addresses from when I fixed mine:

L5 is upper nibble of 4th set
K5 is upper nibble of 3rd set
J5 is upper nibble of 2nd set
H5 is upper nibble of 1st set
L3 is lower nibble of 4th set
K3 is lower nibble of 3rd set
J3 is lower nibble of 2nd set
H3 is lower nibble of 1st set
 
It's either K5 or K3 since it's the 3rd "block". Half of a row I'll call a block. The chips are banked such that each takes care of 4-bits for one of four blocks. I'll fire up Mame and figure it out. Here are my notes on which RAM chips are which addresses from when I fixed mine:

L5 is upper nibble of 4th set
K5 is upper nibble of 3rd set
J5 is upper nibble of 2nd set
H5 is upper nibble of 1st set
L3 is lower nibble of 4th set
K3 is lower nibble of 3rd set
J3 is lower nibble of 2nd set
H3 is lower nibble of 1st set

Actually, is definitely the high nibble, and I'm pretty sure it's the 4th set. I was a little bit confused since I'm used to look at the cocktail mode and some things are flipped around differently between the two modes.

I'm pretty sure that L5 is your bad chip. You can validate this by connecting either 5V to the chip output or ground to pull the results high or low. To protect the circuit, you should stick a 1K resistor between ground (to force to 0) or 5V (to force to 1) to keep from burning out the good chips. If 1K makes it not work, progressively go to lower resistance until it does work. You'll want to look up the pinout for the chip to be sure that you are shorting the correct pin and be very careful that you don't short something that you don't mean to when you do it.

Here's a video I put up to show this sort of debugging:

 
Also, if you click on the "arcade notes..." link in my sig, I have a section in there for Warlords with a little bit more info on debugging this.
 
Results to be posted

Thanks for the explanation and write up. I will socket and exchange L5 either tonight or
tomorrow and will post results. Absolutely valuable info here for the warlords fans.
 
Thanks for the explanation and write up. I will socket and exchange L5 either tonight or
tomorrow and will post results. Absolutely valuable info here for the warlords fans.

If you'd like, I can verify on my board that that is the correct ram before you do the swap. Would be good to double check because of the cocktail versus upright confusion.
 
Is definitely L5. I had to replace that one on mine, so it was socketed. Pulled it and I get the same image as you.

The confusing bit with using Mame (mame -debug warlords) to figure out which bank it is is that I forgot that the upright uses a mirror and Mame quietly flips it for you. :) Mame is how I figure out what was wrong with mine in the first place. . you can pop up a memory window and start modifying memory to figure out what bits are messed up and at what address.
 
interesting!

Very interesting.. I didn't know you cold do this. How do you bring up the memory window to alter the memory bits. Very cool!!! Thanks for the information. I'll change L5 as well.
 
Back
Top Bottom