Well... It lost settings again. I was getting cocky (poor star wars reference there) and was trying all sorts of power cycling and switching to ESB. It did eventually lose settings. I think it was after switching back from ESB, but because I was going crazy with my testing I lost track of where I was.
Just to try and get more info, I'm not going to switch to ESB at all for a few days and see if I lose settings again. FWIW, after losing the settings mentioned above, the SW is holding settings again --- even after several plain power cycles. I want to be able to confidentially say that it happens only when switching back and forth or it just happens sometimes with SW.
Interesting tid-bits... I'm been looking at the 2212 chip and Clay's mod. The Clay mod does have all 2212 paired pins tied together except for CS. What's interesting, however, is that the 2212 allows moving it's RAM into NOVRAM or its NOVRAM into RAM even when the chip is not selected. This means that regardless of which games you are playing, when the game writes to NOVRAM, BOTH 2212s are commanded to write into NOVRAM.
I wonder if this is somehow related. When the game boots, the "recall" line is low which causes the chip to read from NOVRAM into RAM. Both 2212's should read in their settings. With this in mind, if while playing ESB some event occurs that cases ESB to write to NOVRAM, the SW 2212 will also be told to write to NOVRAM. SW should just have the cached settings in RAM, so telling it to write shouldn't clobber anything. This is because after the SW 2212 was told to recall NOVRAM --> RAM at boot, since the SW 2212 never got CS, it will never have performed an operation to change those values. But perhaps for some timing reason the initial recall from NOVRAM -> RAM wasn't successful and then a future store operation wrote the random garbage into NOVRAM.
Just a thought with no real word facts behind it.
... Altan
On Feb 22, 2011, at 10:59 PM, Altan (GAPAS) wrote:
> Quick Update...
>
> First, I found out I cannot socket the 2212 chips as they do not fit below the sound board with the extra height! At least I noticed this before soldering on the sockets.
>
> I decided to take the 2212 from the ESB position, which always seemed to work, and move it to the SW position. I also decided to leave the other 2212 location empty. Specifically, the ESB spot now does not have a 2212.
>
> Recall prior to this, I had determined that the SW 2212 could clobber itself even without switching to ESB. I saw that happen yesterday. But it seemed to always clobber itself when switching from ESB to SW (with power off).
>
> So with the this setup...
>
> 1) the EAROM voltage should be fine as there is only 1 chip.
> 2) Any capacitance should be fine as there is only 1 chip (assuming the cap isn't totally bad) [*see note below]
> 3) SW is now using the 2212 that always seemed to work (from ESB)
>
> And the verdict is....
>
> So far so good! SW is holding settings. SW is holding the #3 score I got during play testing. Switching to ESB doesn't clobber SW.
>
> It's interesting, however, ESB doesn't complain at boot that the 2212 is bad. It's not even there! Is this behavior expected?
>
> My plan is to play test SW under this configuration for several days and see if the 2212 drops its settings. If it all turns out well after a few days, it probably makes sense to consider the EAROM voltage when 2 chips are installed. Or... maybe I'll just be happy with a SW that holds high scores and settings and an ESB that always uses DIP switches.
>
> Also, for what it's worth, my Star Wars PCB doesn't have a cap at c94. Looks like there never was one. Anyone have a cap there?
>
> Thanks again guys.
>
> ... Altan
---------------------------------------------------------------------------
** Unsubscribe, subscribe, or view the archives at http://www.vectorlist.org
** Please direct other questions, comments, or problems to chris@westnet.com
Received on Wed Feb 23 21:11:08 2011
This archive was generated by hypermail 2.1.8 : Wed Feb 23 2011 - 21:50:00 EST