Food Fight NVram

[MikeyDee]

My Food Fight board wouldn't update the NVRAM so I removed the chip and put a socket in. I put a new X2212 chip in and it also won't update

I did some searching on RGVAC and someone suggested that the chip is only written to at power down. They suggested that some of the voltages last a little longer than others when turned off and that is when it is written.

I'm using a switching PS and was wondering if that is my problem.

Can anyone confirm that their settings are saved using a switching PS. It is possible that the chip I have is damaged, but I'm not sure.

Thanks,
Mikey

 

[mspaeth]

I did some searching on RGVAC and someone suggested that the chip is only written to at power down. They suggested that some of the voltages last a little longer than others when turned off and that is when it is written.

As usual, people on RGVAC have no idea what they're talking about. The x2212 as explicit /store and /recall pins to transfer data between the SRAM and NOVRAM arrays. It only uses +5v, so there are no power sequencing issues.

In my star wars/ESB kit, I replaced the 2212 with a modern non-volatile RAM, which will appear to fail the NOVRAM test, since there isn't a separate RAM array, but it does save the scores and settings w/ no issues.

 

[MikeyDee]

Mark,

Specifically, what newer NVram did you use?

 

[mspaeth]

I used the Simtek STK25c48, but simtek has discontinued all of their 2kx8 and most of their 8kx8 parts, so you're probably not going to be able to find them anyway. They're nowhere near pinout compatiable with the x2212 anyway, so you'd need an adapter board to use them anyway.

 

[MikeyDee]

Well, I decided to try an experiment. I took a working A/R II board out of my Centipede and hooked it up in my Food Fight cabinet and disconnected the switching power supply.

With the A/R II board supplying power, the NVram works fine. It initialized and saved my settings.

I also metered the +5v on the A/R II board as I turned the game off (with the FF pcb disconnected), and the voltage took about 3-5 seconds to discharge down to zero.

So, I'm wondering if the nincompoops on RGVAC might have been right.

Anyway, I still want to know if anyone has their FF powered by a switcher and the game saves the setting. I'm skeptical after my experience.

I'm going to dig through my stuff and find a working A/R II or fix one and replace the switcher.

 

[mspaeth]

Anyway, I still want to know if anyone has their FF powered by a switcher and the game saves the setting. I'm skeptical after my experience.

I've bench tested Foodfights with a jamma adapter and switcher with no problems, but I never played them enough that new data would need to be written to the NOVRAM.

 

[MikeyDee]

I've bench tested FF's with a JAMMA adaptor and it's played fine too. The game plays perfectly in my FF cabinet with a switcher -- except for the NVRAM update.

One way to check if the NVRAM updates is in the statistics. It keeps a running total of "game running" time. So, you can check the time then turn it off and then back on again and check the time on to see if it saved it.

 

[MikeyDee]

I'm overthinking -- To test the NVRAM all you have to do is change any of the settings (attract mode sound, # of lives, etc.) and see if it remembers.

 

[Aniraf]

Mikey, did you ever figure this out? I have a problem where my NVRAM doesn't save high scores after I score over 200k. Not sure how the actual score could be relevant, but it seems repeatable.

 

[MikeyDee]

Wow, old thread.

I stopped using a switching power supply and am using the original A/R II.

With the switching power supply, the NV ram didn't update, but the game played.

With the A/R II, everything works fine... even beyond 200k score.


It's possible that you messed something else up when you changed the NVram.

 

[brzezicki]

bringing back an old thread but was working on someones food fight board using a switcher and noticed that when I started it I got the NVRAM statistics failure.

I figured the NVRAM was dead so I put in a new one that I know works and reset the NVRAM and turned it back on. still had a failure.

looking at the schematics it does seem to wait for the power to turn off before forcing the writes.

I manually grounded the necessary pin to cause it to store back the data immediately.
grounding pin #3 manually on 9P does the trick, this won't help people who are running the games permanently off the switcher without making a modified circuit but if your fixing a game on a switcher and need to verify the NVRAM is good. you can use the options to reset the VNRAM and then force the write.

looks like it's a NAND on "update" (pin #1 on LS00 @ 9P) and "power-" (pin #2 on LS00 @ 9P) will cause NAND output (pin #3 on 9P) to go low when when both update and power- are high. pin #3 feeds !STORE (pin #9 on the x2212)

when you reset any NVRAM stats "UPDATE" goes and stays high, and when the POWER- goes high the write will happen normally.

just updating in case anyone else is wondering about this behavior.

-brian

Wow, old thread.

I stopped using a switching power supply and am using the original A/R II.

With the switching power supply, the NV ram didn't update, but the game played.

With the A/R II, everything works fine... even beyond 200k score.


It's possible that you messed something else up when you changed the NVram.

 

[Aniraf]

looking at the schematics it does seem to wait for the power to turn off before forcing the writes.

Could this write be corrupted if the voltage was incorrect? I've been fighting NVRAM failures for years! https://forums.arcade-museum.com/showthread.php?t=367586

 

[brzezicki]

Not sure if/how out of range voltage would effect that chip. I don't recall but I think it just uses +5 so more than likely you'd see other issues. If I recall it just has a wierd circuit to detect power off so it can send a signal to do the write before stored power (from the cap) is actually gone. If I recall it detects when the voltage on the +10 unreg line drops below 8v I think then grounds the write back line. keep in mind it's been like a year since I last looked so some or all this info could be completely wrong.

 

[Aniraf]

Not sure if/how out of range voltage would effect that chip. I don't recall but I think it just uses +5 so more than likely you'd see other issues. If I recall it just has a wierd circuit to detect power off so it can send a signal to do the write before stored power (from the cap) is actually gone. If I recall it detects when the voltage on the +10 unreg line drops below 8v I think then grounds the write back line. keep in mind it's been like a year since I last looked so some or all this info could be completely wrong.

This is good info, I'm certainly not versed in reading the schematics, but I am learning. I'm up to two non-working, and one working, but nvram issue board . One way or another I am going to get this thing back on track!

 

[gamefixer]

More old thread revival!!!!

Working on a FF board and I think there might be something to this switcher/NVRAM failure relationship.

The board I'm working on had a bad NVRAM, tested in a cab with an ARII. On the bench (switcher powered) it also fails, as expected.

So I pull it, replace it and the new one fails as well. Hmm... Start reading this thread and it talks about the supply voltages falling slower on the ARII than a switcher so I put a large cap on my +12 supply and what do you know... The new NVRAM that I installed now works.

So theres something to that +10 dropping away too fast and causing problems with the writes to the NVRAM.

 

[Aniraf]

Now that is ingenious! I never thought about actually keeping the 10v hot longer than the board thinks it is powered on. This must be something in the code that is depending on very perfect timing of the 10v dropping away.

Keep testing and post back if you would. That would be an awesome and easy fix to solder a cap to the 10v line.

 

[gamefixer]

Now that is ingenious! I never thought about actually keeping the 10v hot longer than the board thinks it is powered on. This must be something in the code that is depending on very perfect timing of the 10v dropping away.

Keep testing and post back if you would. That would be an awesome and easy fix to solder a cap to the 10v line.

If you are using the factory ARII you shouldn't need to add any caps to your cab unless you have one that's failing.

as a matter of fact, I think you would notice if the cap that holds the +10 up was failing because it's your main filter cap (aka Big Blue (shudder)).

 

[Aniraf]

If you are using the factory ARII you shouldn't need to add any caps to your cab unless you have one that's failing.

as a matter of fact, I think you would notice if the cap that holds the +10 up was failing because it's your main filter cap (aka Big Blue (shudder)).

Oh, maybe I misunderstood. I replaced big blue almost immediately after starting to understand the cause. That didn't help my issue.

 

[gamefixer]

Oh, maybe I misunderstood. I replaced big blue almost immediately after starting to understand the cause. That didn't help my issue.

The only reason I tried putting a large cap on the +12 on my bench supply was that the slowly falling voltage was mentioned in this thread. It seemed to fix the issue I was having.

That problem you are having is a strange one, for sure.

It's not out of the realm of normal. WMS used the slowly degrading +12UNREG to write protect the NVRAM on their boards. So if one uses a switcher the 12 and 5 fall together and the RAM can get corrupted.

 

[Aniraf]

Well, if Jrok can't fix it at my local arcade party this weekend, I'm sending it to you .