Jtagging using a PIC

[plaguereign]

Ok.  Ive been doing a bit of probing on this idea.  What I would like to do is come up with a jtag hack that sends the signals needed without resorting to jumpers of any sort.  I really would like to get this ball rolling as I have 2 boxes I have jtagged and would like to try something along these lines.  I have the time to work on it if someone else would like to work on something like this or have any ideas on it, lemme know
Thanks

[ReverseAffect]

the reason why these (you call) jumpers is for the jtagging...
kind of a stupid question,don't you think?
if there was a way to do this we wouldn't need the jumpers...but seeing though them paths are not valid without it(duh)..

so 2-3 wires diodes and 1 jumper.....settle with that! a PIC would be useless for doing this,
think about the question you asked..
and if your talking about a pic for nand reading...well then you asked about it all wrong...

[plaguereign]

Why is it stupid?  There was a similar project by DarkstarTM.  I just got in contact with him.  Thanks for the reply, even though the tone pissed me off.

[littlestevie360]

Why is it stupid?  There was a similar project by DarkstarTM.  I just got in contact with him.  Thanks for the reply, even though the tone pissed me off.

well thats not jtagging, thats the SPI programmer, something totally different to what your first post said

[plaguereign]

Quote from: sandungas on January 12, 2010, 05:15:38 PM
There is more code/functions for the PIC pins that are not used in my diagram ?
DarkStarTM said:
Not in the current version of the code. I have some code that does the JTAG hack, but it is easier to do this with the SMC.

There is the direct quote from DarkStarTM answering a question about the Jtag code. 

[neonpolaris]

It sounds like he's trying to make a homebrew cygnos minus the extra NAND.  I'm not sure that I follow the advantage of having the JTAG performed by one versus the other.

[plaguereign]

The advantage, at least to me is having the signals there only when needed.  Ive had an xenon fry on me after only 1 week only one week jtagged.  Ive heard quite a few of these stories and wanted to try and figure out how to stop there from being extra current/voltage on these pins all the time.  Yeah,  it does sound kind of like a cygnos, minus the extra nand.  But, I havent heard of many people who have been using the cygnos having issues with systems dying.  If anyone wants to join me in this project,  you are welcome to.  Even those who mad an a$$ out of me cause THEY didnt understand or are narrow minded.  PICs can do ALOT.  Atmel chips can do alot simply by knowing how to program correctly. 
An only on when needed jtag hack could help save systems.  Im just asking for help to get it rolling. 

[maximilian0017]

The problem with the pic's and the AVR's is the same as with the old Xenon resistor JTAG hack with the SB, you need a propper level converter for it.

With the diode design JTAG hack this should not be a problem.

Most likely your xenon died of something else, removing it from the case will put a lot of strain on it and can certainly cause RROD etc.

[Tiros]

Why is it stupid? 

Because the smc (using diodes) already does jtag at the correct voltage levels?
Isn't adding another chip to do the exact same thing redundant and pointless?

I dont see how you will design newer and better without an understanding of the current situation.

[Blackaddr]

The only advantage I can see to improve upon the diode based JTAG circuit is create one that does not suffer from the non-zero turn on voltage of the diode.  Basically, use a proper level converter chip (which you are not going to find at  your local components store).  While this might fix the boot problems with Zephyr boards (we don't know for sure what causes it but let's say it's the logic LO isn't capture right) it is unnecessary from a safety perspective.  The SMC never drives a 3.3 voltage into the GPU, it drives approx zero volts, or lets the line get pulled to 1.8V.

- Blackaddr

[plaguereign]

Even the jumper wire, which has nothing on it?  And, as I stated above, there have been too many consoles that died for it just to be a coincidence.  It hasn't just been mine.  I have talked to at least 20 people, who, within a week of jtagging, lost video totally or got the rrod. And, I know that isnt everyone.  The boards do not have to be removed to jtag the xenon.  It seems like too high of numbers to be coincidence. 

[plaguereign]

if consoles are dying for unexplained reasons and we assume the issue is not software based then, removing the only hardware change after it has served its purpose can do no harm and may alleviate the issue.

[plaguereign]

hence , the addition of a jtag switcher can do no harm and could do some good.  the only other logical reasons for these unexplained failures is that either they are normal failures that any stock 360 could experience or, that the user cause motherboard flexing or some other stresses to cause the issue during exploitation.  not removing the motherboard removes the flexing aspect, unless their xclamps or xclamp replacements are not functioning properly.  as I stated above, the numbers seem too high for a coincidence.
I posted here (which I never forum post) to get some constructive input on the project to improve the design of a 'transparent' jtag.  All that seems to be happening is flaming.  Blackaddr was the only one kind enough to explain his POV.

[Blackaddr]

Even the jumper wire, which has nothing on it?

The jumper wire is simply tying the active low JTAG reset pin to a 1.8V supply rail (thus ensuring the JTAG port is not held in reset).

There is need for protection on this wire on the Xenon's for sure, and I don't think there's been evidence that PCB net had changed on the new consoles.

Even if the JTAG wires did cause damage, it would damage the GPU JTAG port, which would prevent XBR from booting, but would not cause any other permanent damage since the JTAG port on the GPU is not used for normal operation with the original NAND.

[maximilian0017]

While this might fix the boot problems with Zephyr boards (we don't know for sure what causes it but let's say it's the logic LO isn't capture right)

Would suspect this is timing/software or STIO related and not the wiring/diodes.

hence , the addition of a jtag switcher can do no harm and could do some good.

With every extra option you create more problems arise.
Take a look at the Cygnos v2 users with their 3 install methods (resistor/diodes and wires JTAG) with 3 firmwares, and all the images have to be modified with their toolbox that is ^%$^%$# because the user has no control....asking questions here creating yet more ways for it to fail.

With the level converters for 1.8v being hard to get you get people who make their own designs.....problem problems problems.

K.I.S. Keep It Simple

And most of the fails are due to the users not modding the xbox correctly/without respect or the known problems with the ROHS solderballs/heat.
Removing a MB from its case can easely create problems, why have i had 0% troubles? even with the JTAG resistor mod , even with most repairs from other people you can easely guess what the problem is, search for "FrankenX"

I'm just a simple hacker with basic knowledge but if the people that created the original hack say its bogus...well then its just bogus.