Previously Unnoticed Test Points and Slide Switch Behavior?

[Dr. Vegetable]

Ok, this might be nothing, or it could be huge...

Flip over your PSP and pop out the battery. In the upper-left corner of the battery compartment, just above the Sony logo is an easily-available test point. Toward the bottom of the compartment, between the two "warranty void if removed" stickers is a row of additional test points. These all seem to be connected together, ohming out to zero, and looking like parts of the ground plane on the underside of the button board. (I have not measured the impedance at higher frequencies...)

So why would they have made special cutouts in the case to provide access to a ground plane?

Looking at this board some more, I notice that the Square button is a bit different than all the others. The rubber contact pad is oval, where the other seven buttons have round pads, and it has three vias at one spot where the other buttons only have two. And then I remembered that they had problems with the Square button working properly on early production models. "Fascinating, Captain."

One other point of interest - the power switch (also on this board) is a single pole, four position switch. We all know that you slide it hard up to toggle power and (if held long enough) reset the machine. And you slide it down to put it in Hold mode. But did you know that there is a third output contact on that switch that breaks contact if you slide the switch partway up? In the normal rest position (spring-loaded) this contact is closed. Anyone who is trying various button presses while powering up the unit should be aware that this switch has four positions: Hold, Off, Halfway On, and On. I have been told by an authorized Sony service rep that there is a button-press combination at power-up that will reset all custom user settings, such as wallpaper. (He wouldn't tell me what that combination was.) There are probably other button combinations as well.

As I said, it's probably nothing. But if Sony wanted to have important access to the innards of the PSP, why not put it on an innocuous peripheral I/O board instead of on the main board which they know everyone would scrutinize? I would imagine that the technicians at Sony are told to take the customer's battery out and pop it onto some kind of rig to provide a reliable power source when servicing the unit. And we already know that the middle pin on the battery is a bidirectional COM port...

[pyrosama]

Holly PSP's batman!

Well we will get a better look once I get that dboard from dot and I can take some more sexy photos.

But it is interesting that those points are there and they are far larger points than those on the main board witch beggs the question what the hell are they there for?

With my half dead board I will see about this button combonation thing. Just rigg it so it's always half on and press buttons lots and lots of buttons see if it acts diff at any point. No chance of it but thats 10 minutes of filling blank space in my day?

PyroSama

[Dr. Vegetable]

To be clear, this is not the Wifi/SIRCS/Memory Stick daughterboard that I am talking about. This is yet another board which sits above the battery compartment and beneath the four AOX[] buttons.

I have determined that these buttons are pressure sensitive (resistance across the contacts decreases as you press harder) which may explain some of the over-engineering that initially made me suspicious about this board. However, it is possible that the buttons also have a special mode of being pressed that could be detected by the PSP: Lightly pressing the upper-left corner of the AOX buttons or the lower-right corner of the [] (square) could connect the button sense path with the bottom ground plane on this board, while pressing lightly in the opposite direction could connect it with the top plane only. (The planes seem to be connected, so again this is probably nothing, but the external access is still curious.)

I got really excited at one point when I noticed that the Square button is different from the others - this is the only one that you can get your fingernail underneath from the left side to easily rock the button toward the right. (Notice how easily the square button can be rocked toward the left if you press it off-center.) The conductive rubber contact on this button is oval, while the pad on the other three is round. However, it appears that the cutout in the plastic of this button was made to prevent it from touching the edge of the display, and may simply be the "cure" for the problem reported to occur with this button on early production units. (Anyone have access to an early-model Japanese PSP that they can verify this on?)

So this is probably nothing. However, it does mean that the ground plane "test points" that are available through the battery compartment (vias, really) can be connected to any combination of four analog input lines into the system simply by pressing/holding a combination of AOX[] buttons. It is possible that the PSP might recognize a signal that gets injected into the system through these contacts when the appropriate button combination is held...

[dot_blank]

i have early model japanese
and the rubber is not oval ...it is missing
a bit from left cuz of the screen ...
infact square pad rest right on screen
metal covering ;P

pyro: i sent my daughter shes all yours now ;)
take good pics of EVERYTHING
i have removed metal shielding beforehand
to make it more simpler for you....and
so you wont mess things up :P

Dr.V: i was curious of what you were talking
about and i looked yet cannot detect this
elusive forth "half on" switch with power
switch

i have one psp setup as "blown out" ala psppet :P
...meaning no shell just internals and it runs
nice as dev unit ...and i have tried numerous
times with this unit and no fourth power state for me :P
and numerous combinations on real psp....and nothing
also the vias where battery lies is most interesting
indeed ...in fact sony requires you to remove
your battery before sending so ...maybe they need
this for techs to have simpler reflash operation...who knows
side note: also battery must be remove to prevent
against any shocks or possible leeks from battery
so theres common sense in this action of removal of battery

well like you said Dr.V it could be nothing or it could
be everything ...time will tell ;)

cheers

[digihoe]

I think the answer to the keypressing at bootup is in the IPL as it is the first to boot, if one would know how it works I'm sure one could find the combination or even an softmode to reflash... Too bad there is not much information on the IPL...

Best regards!

[pyrosama]

Further investigation of these test points I discovered that there are infact 8 points yet numbered 1 right 10 to left. They are quite strange there is a tab off this dboard that apears to be used to hold the dboard in place in psp yet intentionaly placed over a point of access. Quite strange. There are several other testpoints that are all grounded to each other on surface of board but all apear to be connected to the ribbon connectiong the dboard to mobo. (I will photograph shortly) But it is quite interesting that the testpoints on this dboard are all grounded =/

Comparing the number of traces in the ribbon vs the number of used points is equil. 10 points go to the buttons under the LCD the 5 are used for the buttons then the 2 are used for the led and 3 for the power switch as such all are accounted for.

More later when I have had a chance to photograph these parts.


Canti_

[HaQue]

I was thinking about this 4 position switch.. Are you sure it isnt really 3 positions with 4 contacts? refer to the ghetto code to see what I mean.

Code: Select all

|_| | |

| |_| |

| | |_|

I cant see that there would be a halfway, when it would take some "skill" to get it in the right position.

Id love to be proved wrong though, and see a whole new light to the little bugger.

HaQue

[Dr. Vegetable]

Right - the switch is a three-postion switch with four contacts. (You always need a common pin.) However, there are three positions where the common pin is tied to one of the other pins: Hold, Off, or On, and a fourth position where the common is not connected to any of these pins: Half On. To get there, one need simply apply slight upward pressure on the power switch without pressing hard against the stop. Since the switch has small tabs on it, it would be possible for the technician's service station to guide this switch into the Half On position when the PSP is inserted for service.

Remember that, like many modern devices, this is not a real power switch - it doesn't physically interrupt power to the device. Instead, the processor (or some peripheral) monitors the state of the switch to see what state it is in, and then decides whether to turn on or off or... The RESET function that occurs when the switch is held On for ~20 seconds is probably implemented in hardware, with a slowly-charging capacitor tied to a reset line on the processor. (Again, speculation so far, but this is how these are usually implemented.)

The point is that there are four discrete states that this switch can be in, and the processor is capable of detecting all of them.

EDIT:
Clarification: Reset lines are usually active low so that the device starts in reset mode when initially powered up. Instead of "slowly-charging capacitor" I should have said "slowly discharging capacitor. Same idea, though.

The other point to remember is that, except when the battery is completely dead, the PSP is always "on", even when you turn it off. You cannot turn it on unless the processor occasionally wakes up to examine the state of the power switch. And even then, I have found cases where the PSP will refuse to turn on even when you ask it nicely. (When a battery is inserted, the PSP interrogates it and will refuse to use it unless it can answer me these questions three... to borrow from Monty Python.)

And even with a "dead" battery, remember that there is some kind of button-cell battery inside the PSP which could continue to power the device even when the battery is removed. I don't know if this is a simple lithium cell that will eventually die after some years, or whether it is rechargeable.

[pyrosama]

I'm messing with some things right now and this is what I have so far:

The bar across the bottom's connector to the dboard is as follows
|G|G|SELECT|MUTE|Voluem+|SCREEN|START|Volume-|HOME|G|

AOX[] buttons as follows

G|+
+|G

Powerswitch:

|-Poweron
|-On
|-Ground
|-Hold

Canti_

[pyrosama]

New photos uploading now they should be done in about 10 minutes.


Canti_

[demologik]

maybe you're supposed to hold the power switch in the half on position, and then implement the special code combination..?

[Erant]

I don't think the switch has a fourth state, I don't have my multimeter here, so I can't confirm this, but what I see is that the switch has four pins, two for each state. One pin isn't connected, which is probably because it's an extra GND pin, so each state (aside from center) has it's own GND.