[PS2 TWO] Romeo2 Fix y Otra posible solucion

Hey,
yes that last link was in italian and they were describing a "romeo2 mod":
http://www.hardstore.com/default.asp?cmd=getNews&cmdID=2225
Hooking up the "silent PWMx2 5-channel motor and tilt driver" to a 5 volt power supply from the PSTwo didn't work, the driver IC would just not function at all (I tried that too).
But in this site they describe using a 7805 voltage regulator that uses the ground plain of the PSTwo as a cooling ribbon. And placing two diodes in series in between the ground input of the 7805 and the ground of the PSTwo. This way they would elevate the 7805's output voltage with 1.2 volts above ground. The 7805's input is soldered to the same powersuplpy point as the tilt-driver IC, but then on the "other side" of fuse PS11 (on the bottom). Then the output of the 7805, which will be at 5 + 2*0.60 = 6.2 volts is soldered to the VCC2 of the tilt-driver.
So what they did is hooking the tracking/focussing power supply of the driver IC to 6.2 volts... one VERY BAD thing about this fix is that they soldered the input and the output to the same voltage node with fuse PS11 in between.
The driver IC is connected to both 8.5 volts via fuse PS11 and 6.2 volts via the 7805. In result fuse PS11 would burn out to seperate the two circuits and all is fine.... but what if you want to restore the PSTwo to default??? then you would have to replace the fuse.
I would say... nice fix to get the maximum current down with just a few parts. And a much more elegant solution than clipping the driver IC's input voltages to +/- 1.2 volts... that would seriously degenerate the PSTwo's speed and capability to re-focus and re-track it's discs.
Unsolder fuse PS11 when you do this (dangerous too btw)... or even better... solder the 7805's input pin to the bottom of the PSTwo on the right (protected) side of fuse PS11 and lift the pin of the driver IC off the motherboard and solder the output of the 7805 to this pin.
Here's a link to the 7805 datasheet just for reference:
http://www.national.com/ds/LM/LM7512C.pdf
ok... so this was another one of my semi-scientific analysis.. hehe... I hope it's of any use. If I wouldn't have any better ideas to fix the "PSTwo Syndrome" I would take out my camera, do this fix, make some better tutorials of this and post it on this forum. But no, you can try this if you like and in a week or two I'll introduce my own fix when it's tested and proven ... ow.. it won't be a "quick and easy" fix... but it will be a very good one. There are some serious design flaws in the PSTwo... and the only way to fix it... is to fix all flaws sufficiently.

(P)So...is this fix better than the matrix fix?

(R)The driver IC was not really designed to run on low voltages, it's response to higher frequencies input will not be as good as they would be when supplied at the normal 8.5 volts. But it does have the "side effect" of limiting the maximum output currents and this is what prevents the coils's coating from melting.
The matrix-fix does the same, but then by hard-limiting the input voltages by clipping them. This doesn't particularly affect the response on higher frequencies, but the overall response is made worse. Discs will read slower and a wobble in the disc or a misalignment in the disc "tray" will make discs unreadable because the lense can not be positioned over it's full dynamic range.
Neither of these solutions are really perfect. They're just temporary patches that will hopefully prevent your coils from shorting your IC's and burning them all out. I would suggest applying this "romeo2 mod" wih the 7805 and the two diodes in combination with the "matrix fix", but then with 3 diodes in series. This will widen the range of your lense's motion compared to the normal "matrix fix" and the 7805 will help to do some current limiting by "underfeeding" the driver IC.
None of these solutions will ever prevent DC current going through your coils and none of these actually solves the problem: noise coupling into the tracking and focussing feedback circuits!

(P) So basically what we are looking at here is that DMS4 should not be installed on a PSTwo yet if you don't want problems. If we ordered it already what should we do?

(R)It's not DMS4, it's all modchips!!!
And if you already opened up your console, you should just solder it in and apply the matrix fix and also hook the tilt driver IC up to 6.2 volts with a 7805 and 2 diodes. And later, when a better fix is around, you should apply that one too. Or when these fixes (or a combination of both) are proven to be ok, you can leave things the way they are.

If your console is still closed and under warranty, I would wait a week or two to see the feedback on these fixes or wait for new ones. You could do some tests too in the meantime:
* Let your PSTwo play a movie on a DVD+-R for a whole day and night
* Let your PSTwo play the intro of GTA Vice City, San Andreas etc. for a long long time.
* If it breaks -> return to sony
* If it survives -> your PSTwo will probably do fine with the matrix fix and the altered power supply, at least for a long enough time.

I'm working on a fix that's quite "extreme" but will very likely get rid of all problems at once. I'm still designing and testing them, so there's not much I can tell about it besides this: a filter between the CD/DVD DSP and the actuator to get rid of high frequency noise, directly limiting coil currents and a shutdown circuit for the tilt driver in case there's an overcurrent. And of course some rerouting and redesign of the powersupply and some better cooling, like the pupazzo fix

I hope my experiments will turn out to be succesful and easier to apply than the current status of my designs, some simplification is needed. I'll post results here when I'm done.

Hey guys (and girls)...

KaISeR SoZEi handed over some other document he got from some other (of course inferior :-p) modchip maker... and it probably desctribes very well the same cause for the PSTwo laser burning that I independently found:

http://www.ripper3.com/v12_DEAD_BEEF_067.pdf

Read the document and you'll find this out:
* DVD DSP locks up during boot or during play, most probably due to glitches and noise in the power lines.
* During lock up the DSP will no longer send adjustments to the focussing and tracking actuator, but send a DC signal instead.
* This DC signal will be amplified by the Actuator (the RS2004FS) and this huge current will burn your coil.
* Your RS2004FS will probably survive this huge power drain, because it shuts down at high temperatures, where the BA and LA chips would just burn out and die.

Up to now I have built a few "better safe than sorry" fixes for the PSTwo. I built a 3.3 volt power supply for the DMS chip to help keep the 3.3 volt rail clean. A 6.6 volt power supply for the coil actuator VSS of the RS2004FS chip (like the romeo2 mod). Two current limiters built from standard opamps that are inserted in between the RS2004FS and the coils with a red warning led that will turn on when the current through the opamps is on it's maximum allowed. Also when the red light turns on, the 6.6 volt power supply to the RS2004FS chip is disabled.

When I would operate my PSTwo with all these patches applied I would hardly ever get the red warning led to turn on, except when I would continuously reset the console. Then the red light would turn on and any disc in the drive would just not boot (disc read errors) because my circuit just disabled the supply to the coils.

The only thing I could do to get it to function again was completely turning off the console (unplug power) and turn it on again... resetting would not work. I had no idea why, could not find any cause, but for sure my circuit saved my coils more than once. Now this document I think described exactely what I found but did not understand.

Those people said they would build and test a circuit that would detect this lockup and apply a reset to the DVD controller... in the same time I will try to find this reset function of the DVD controller (where the h#ck is it!) and hook up my red-light-warning circuit to this reset function. Or at first.. I will try to check if a reset of the DVD DSP will enable me to play again without having to turn off and on the system again. If it works, I can make a simple circuit that will protect your coils and reset the DSP during operation when it would ever lock up.

Ok.. it's easier said than done... I';; go to sleep now (it's 5:15 am here) and I'll post back to you guys when I find out more.
Ow and one more thing...

* The Matrix Fix is NOT a good solution...
* The Romeo2 mod is more likely to save your coils than the matrix fix!
(romeo2 mod is the one with a 7805 and 2 diodes)
* The pupazzu mod is not a good solution either on itsself... it's a good addon to any other fix but will not help to save your coils
* putting a resistance in between the coils and the actuator will not help either, you'll only get DRE's

Good night

Hey...

I am currently (yes right now) making a simplified version of my current-limiter that will for 100% sure prevent your coils from melting because of overcurrent, DSP/Mechacon lockup etc. etc.

There will be a usefull solution, I actually already have it and soon it will be simple enough for all you people to implement.

What my fix will not do as of now is reset the DSP or Mechacon so you can continue playing as if nothing happened after such a lockup. I only was able to "recreate" this lockup after lots of resets and plugging/unplugging the power suply and some TL light switching. I haven't played long enough to encounter one during play... so I think a simple current limiter will do for now.

Just have patience...

What the matrix team showed was a graph of the voltage between one coil and ground. And yes.. those voltages can make a huge swing... but that's not how it works.

Both coils have 2 connection points that are connected to the actuator and none are connected to ground. The offset voltage of both coil connections is 3.5 volts and it has a swing of +/- 3.5 volts making it possible to have a voltage of 0 to 7 volts in both possible polarities on the coils so it can move up and down. Sony did this, because they operated the actuator with only 1 positive power supply.

The Ripper team showed a different graph... the way it was supposed to be measured. They measured the coil-ground voltage of BOTH connections of the coil and substracted those in their osciliscope program... and then you can see the actual voltage that is over the coil itsself.

When you apply the diode "fix" the voltage of each input connection of the actuator is clipped to +/- 1.2 volts. So you could have a maximum of 2.4 volts difference over the actuators input... after amplification by the actuator and putting in the 3.5 volts offset your coils would still get almost the full possible voltage swing with no current limitation.

It might looks nice in the graph that before the "fix" you see a "big signal" and after the "fix" its a "small signal"... but it doesn't mean anything! It does take down the maximum output voltage a little bit, but by far not enough to save your coils and it does degenerate your tracking and focussing abilities.

The romeo2 mod that puts the RS2004FS on a 6.2 volt power suply works better. But it puts the entire chip on 6.2 volts, including the charge pumps and other components that really need a higher voltage power supply. It's also not current limited so if you have bad luck it could still burn your coils, although it will take significantly longer than wihtout this mod or the diode "fix".

The only thing that is know to cause the laser coils to burn is a lock-up of your Mechanics controller or DVD Digital Signal Processor. This "lock up" is a strange condition in which your PSTwo can get during boot and maybe also during play in which it does not respond and does not read from the disc anymore.

It just hangs, and outputs a DC voltage and noise to the actuator (the chip that amplifies the signal and drives the coils and motots -> RS42004FS, LA6508, BA5815). This DC voltage and noise will kill your coils after amplification. I'm trying really hard to get my PSTwo in this locked up state to see if there's someway to reset is without having to turn the PSTwo entirely off and on again... and to see if my own fix will really prevent the coils from burning, also if you leave the console in this locked up state for a longer time.

But as I said... it's hard to get it to lock up. You actually don't want to ever have it lock up when you just modded it and are playing your game, but it could happen. Sometimes it happens after the first boot, sometimes after some serious TL light switching and a reset or two and sometimes it takes me ~100 resets to even get a little bit of DC or noise... You probably have one of those consoles + adapters + home electricity circuits that will make it almost impossible for the thing to lock up.

The ripper team released another document about the V12 "DEAD BEEF" problem:

http://www.ripper3.com/v12_DEAD_BEEF_070.pdf

They are describing a possible cause for Mechacon and DVD DSP lockup. It's not THE cause but one of the possible causes. They describe that bad patching of the DVD bus can cause the chip to lock and show this strange behaviour I talked about. They were also telling that the way "their chip" patches the bus should not cause any interfernce at all because of they way they patch it... and also one other modchip uses this advanced patching... they didn't say any name... but I verified and it's the DMS4 modchip they're talking about!!! (on their diagrams wires B, G, H and I which are on DMS4's diagrams wires P, U, V and W are all connected to two pins of the Actel FPGA to drive the bus with increased current).

I also verified one cause of this locku pthat has nothing to do with modchips... just ripples on the power supply voltage and random errots... they're hard to pinpoint and recreate.. but they're there.

I'm not working on a fix that will prevent this lockup, because that would probably mean a redisign of the entire PSTwo .... Just a fix that will prevent your coils from burning and will maybe later reset the mechacon or DSP too after a lockup occurred.

But this forum is taking up too much time that I should use on testing and publishing my fix