JTag Saver?

[TheBiGW]

Thought I'd start a thread here to ask what some of the experts think about the idea of a JTag saver. I've been reading here for a while and have never heard of anything close to this before so is this legit, or is this a waste of time?

Reasoning behind the design:
We wanted a simple, easy, and effective way of shutting down the jtag when it wasnt needed any longer.
Trying a 555 timer circuit, the circuit worked well, but was too expensive for people to get behind. This revision
actually costs about 2 dollars USD to implement and works well, without all of the extra components that were needed for
our first revision.

What will Jtag Saver do for me?
It will, in essence, allow you to safely use your jtagged xbox 360 without worry due to extra voltage or current being applied
to areas that weren't meant to have them applied. This circuit will shut the jtag circuit off after 15 seconds to try and limit
the damage that could be caused after long periods of use.

We developed this circuit after a few boards died soon after jtagging. It was too close together to be coincidental. Most boxes
that died did so within a week of jtagging. THIS is the reason for our pursuit.

The circuit has been tested thoroughly. Many on/off cycles were gone through to make sure that it worked every single time. It did.
Optionally, you can add a resistor and LED to pin 7 to show you when pin 6 is shut off (goes high). A 220 ohm resistor should suffice, but
this part is up to the end user.

Saw it posted elsewhere

http://forums.xbox-scene.com/index.php?showtopic=709835

[keropi]

yeah, what gives? I would like to see some of the knowledgeable people here commenting on this since their opinion is valued higher in my book 

[tommyv]

yeah I didn't know there was a problem with the JTAG circuitry, I don't think there is one, I haven't read about people having this problem at all, so it really calls my attention, I mean if it was a problem, a lot of people would have been making tons of posts on the last few months reporting a heck of a lot dead 360s, nevertheless if this is a safer solution why not use it.

[maximilian0017]

BS

[ReverseAffect]

well if the user did it correct in the first place,
and nothing was wrong with the unit before hand..

we wouldn't have this stupid post......(not yours TheBiGW..the link you posted)...
but seeing though 95% of errors are usererrors..I guess some noob has to think of some way to f*** it up more...

weird i have 5 running everyday on and off from xenon to falcon and not 1 problem beside the normal reboot glitch (rare but happens)
all been running for up to a year straight....

[Lush]

I have 3 wired and running since the inception of the JTAG hack with no problems... They include a xenon, falcon, and jasper 512... I did upgrade my xenon from the 3 x 100ohm to the diode method though when it was released... 

[Gazcoigne]

same here done about 20 odd and no issues reported yet whatsoever

bit of a waste of time IMHO

[dtrmad2004]

Same here, loads done and all the ones that died were due to HW errors not relating to Jtag points. Orig nand dump didnt work neither.

[Blackaddr]

If using the JTAG diodes, there is no electrical risk whatsoever.  The "issue' is the SMC is driving 3.3V while the GPU JTAG is 1.8V.

The way the diodes work is they allow the SMC to drive logic LO, but when the SMC drives logic HI (3.3V) the diode effectively disconnects the SMC and a pullup resistor on the board already will pull the line to 1.8V which is the GPU voltage.

The connection with the jumper is simply tying an active low JTAG reset pin to 1.8V, so reset is not asserted.

This method of JTAG hack is electrically SAFE.

The LPT is a different story.  Your LPT port is +5V, your SMC is 3.3V.  Putting resistors in series helps provide a current limit but can still damage the NAND.  There is a definite risk here (even if a small one).  But since most people access the NAND in this way only a few times, then later updates can be done wtih software tools running on the console, this isn't a big issue.

However, I still DO recommend people only use a buffered, level converting interface if they want to be 100% safe.  I posted a schematic for such a circuit in the Hardware section.  I have seen enough flash chips get fried with just resistors from my satellite experience to know not to do it on an Xbox.  So I never use anything other than an electrically safe circuit, but I'm anal, and to my knowledge NO ONE HAS DEFINITELY FRIED THEIR NAND *edit* CONTROLLER *edit* FROM USING ONLY RESISTORS YET.

But this JTAG Saver above is pure snake oil.

[Tag234]

I have the JTAG installed on my Xenon but im still using the 3x 330ohm resistors... i havent had a problem with it since it was done about 2 months ago but is it worth upgrading to this method or the other newer method if i havent had any problems???

[n00bpwner360]

I have the JTAG installed on my Xenon but im still using the 3x 330ohm resistors... i havent had a problem with it since it was done about 2 months ago but is it worth upgrading to this method or the other newer method if i havent had any problems???

2 diodes wouldn't cost you much, and they're safer, so why not upgrade?

[ccfman2004]

If using the JTAG diodes, there is no electrical risk whatsoever.  The "issue' is the SMC is driving 3.3V while the GPU JTAG is 1.8V.

The way the diodes work is they allow the SMC to drive logic LO, but when the SMC drives logic HI (3.3V) the diode effectively disconnects the SMC and a pullup resistor on the board already will pull the line to 1.8V which is the GPU voltage.

The connection with the jumper is simply tying an active low JTAG reset pin to 1.8V, so reset is not asserted.

This method of JTAG hack is electrically SAFE.

The LPT is a different story.  Your LPT port is +5V, your SMC is 3.3V.  Putting resistors in series helps provide a current limit but can still damage the NAND.  There is a definite risk here (even if a small one).  But since most people access the NAND in this way only a few times, then later updates can be done wtih software tools running on the console, this isn't a big issue.

However, I still DO recommend people only use a buffered, level converting interface if they want to be 100% safe.  I posted a schematic for such a circuit in the Hardware section.  I have seen enough flash chips get fried with just resistors from my satellite experience to know not to do it on an Xbox.  So I never use anything other than an electrically safe circuit, but I'm anal, and to my knowledge NO ONE HAS DEFINITELY FRIED THEIR NAND FROM USING ONLY RESISTORS YET.

But this JTAG Saver above is pure snake oil.

If I use the USB Nand dumper/Flasher found here http://www.sparkfun.com/commerce/product_info.php?products_id=676, will this pose any electrical risk to the xbox.

[trasixes]

I have the JTAG installed on my Xenon but im still using the 3x 330ohm resistors... i havent had a problem with it since it was done about 2 months ago but is it worth upgrading to this method or the other newer method if i havent had any problems???

2 diodes wouldn't cost you much, and they're safer, so why not upgrade?

Can't speak for him, but in at least my case, I have one Xenon that won't boot anything except the original dash with diodes, but works fine with the 330 ohm resistors. The same diodes worked fine in other 360s though. /shrug

[X-QlusioN]

I had a Xenon that would occasionally E79 on me with 3 resistor method (and yes they were soldered fine) changed it to 2 diodes and haven't had an E.. since...

[trasixes]

However, I still DO recommend people only use a buffered, level converting interface if they want to be 100% safe.  I posted a schematic for such a circuit in the Hardware section.  I have seen enough flash chips get fried with just resistors from my satellite experience to know not to do it on an Xbox. 
But this JTAG Saver above is pure snake oil.

True. Was very silly to jtag an IRD with the resistors/DB25m, and many did it anyway. I'm not sold on the diodes though, at least not until I'm able to discover why this particular Xenon won't work with diodes. Every other Xenon works fine with the same diodes (from the same pack even).

[jelle2503]

just a noob wanting to get rep off releasing something "useful"

[maximilian0017]

NO ONE HAS DEFINITELY FRIED THEIR NAND FROM USING ONLY RESISTORS YET.

Correct me if i'm wrong..

They wouldn't blow the Nand, they would blow the SB SPI port, and some people have blown the SPI port.(ofcourse without resistors or other mistakes)

[Blackaddr]

NO ONE HAS DEFINITELY FRIED THEIR NAND FROM USING ONLY RESISTORS YET.

Correct me if i'm wrong..

They wouldn't blow the Nand, they would blow the SB SPI port, and some people have blown the SPI port.(ofcourse without resistors or other mistakes)

You're absolutely right, I've gotten into the bad habit of saying NAND like many others here when we're really talking about the NAND Controller i.e.  SMC via SPI interface.

If I ever refer to the SPI interface to the SB as "jtagging it" you can put me out of my misery.

[Blackaddr]

I'm not sold on the diodes though, at least not until I'm able to discover why this particular Xenon won't work with diodes. Every other Xenon works fine with the same diodes (from the same pack even).

Overvolting an input doesn't have to instantly fry it, it is often a cummulative damage that usually increases input capacitance of the input receiver.  Increasing the input capacitance will cause signal edges to rise and fall slower.  You can increase the slew rate by increasing the voltage.

Just a theory...

It is possible that you've already damaged that particular GPU with resistors in the first place.  The signal rise rate is a function primarily of the pull up resistor on the board and the input impedance of the pin.  With the diode configuration, the pullup and the input pin might be rising too slow.

When you put the resistors back on, you are now driving a full 3.3V into the GPU, which helps overcome the input pin capacitance and slews fast enough to work.

- Blackaddr

[trasixes]

I'm not sold on the diodes though, at least not until I'm able to discover why this particular Xenon won't work with diodes. Every other Xenon works fine with the same diodes (from the same pack even).

Overvolting an input doesn't have to instantly fry it, it is often a cummulative damage that usually increases input capacitance of the input receiver.  Increasing the input capacitance will cause signal edges to rise and fall slower.  You can increase the slew rate by increasing the voltage.

Just a theory...

It is possible that you've already damaged that particular GPU with resistors in the first place.  The signal rise rate is a function primarily of the pull up resistor on the board and the input impedance of the pin.  With the diode configuration, the pullup and the input pin might be rising too slow.

When you put the resistors back on, you are now driving a full 3.3V into the GPU, which helps overcome the input pin capacitance and slews fast enough to work.

- Blackaddr

That's an excellent theory, but alas, I tried diodes first. I soldered/resoldered, quadruple checked my work with magnified images, and still no love from that 360. Diodes from the same bulk pack work great on other Xenons, so quite some time passed before I realized why nothing would boot except the original nand dump. Once I switched to resistors, everything worked beautifully. It is worth noting that it was a repaired RRoD unit (GPU xclamped), so that may or may not have bearing on the matter.