Print

[catdog2]

Diagram here

Extending the life expectancy of a modchipped v1.6 xbox

In a v1.6 xbox, one of the data signals on the motherboard (LFRAME) is short circuited to ground to let a modchip boot.  This puts a strain on the xbox electronics whenever the modchip is operating.  Although the modchips work, the life expectancy of the xbox may be shortened by doing this - much like driving your car around with the engine always running at 6000rpm (the car still goes, but the engine will get warn out much more quickly).  A resistor can be used to prevent the problem of installing a modchip, and extend the life of the xbox.


A more detailed  description:
The bios on a v1.6 xbox is loaded from the LPC bus, the same bus that the modchip uses.  Therefore there will be two different bios (modchip and normal MS bios) both trying to load at the same time onto the same wires.  v1.0-1.5 xboxs used separate wires to load the MS bios compared to the modchip and so this wasn’t a problem.  

So how do we stop the MS bios from loading onto the wires we want to use for the modchip in a v1.6?
LFRAME is used to signal the start/end of a data transfer on the LPC bus.  In v1.3-1.5 xboxs MS removed this signal from the motherboard to make "cheapmod" modchips no longer work - so modchip makers generated this missing signal with some extra electronics on each modchip. On v1.6 xboxs we short out the LFRAME signal on the motherboard so that the bios chip on the motherboard doesn't know it is being sent data (it can't see the start signal) and so it doesn't do anything.  The modchips now generate the missing signal themselves, so they see the data correctly and can load up without any problem.

So what is the effect of shorting out LFRAME?
Shorting out a signal causes a higher current to flow than is normal.  In v1.0-1.5 xboxs we short out D0.  This causes a current to flow during the time the onboard bios is trying to load - which is a maximum of a few hundredths of a second.  After this it is disabled so no current flows, and the modchip loads up.

On the v1.6 we are shorting a different signal, LFRAME, which as far as the xbox is concerned is required to load the bios in any circumstance.  This causes a problem - the signal does not turn off when the modchip loads like the v1.0-1.5 xboxs do.  We have a continuous current flowing whenever the modchip is being used - a current which is coming from a data line driver which is not designed to do this!  

For those interested, I measured 62mA and 68mA on two different 1.6 xboxs - to put this into perspective a standard 3.3V 74LV series IC output normally operates at a few mA, and has an absolute maximum rating of 25mA (I am comparing with other generic electronic components as I obviously don't have the actual xbox component datasheets - anyone who is feeling generous feel free to send me a copy :-)  Some new modchips, such as the Aladdin XT, have had to add an extra point ‘X’ to try and deal with this high current as the original Aladdin modchips cannot cope.

One fix for this is quite simple, cut the data track on the motherboard and insert a resistor.  With the modchip off, the resistor has very little effect on the data line and so the xbox boots normally.  With the modchip on, the resistor limits the current to less than 5mA (less than one-tenth of the original value) and so prevents any long term damage occuring to the xbox.


If I have made any errors in this description let me know and I will edit it appropriately.

UPDATE: Modchips which pulse LFRAME and therefore do not suffer from this problem:
Xenium ICE
Duox2
(These results are from posts in this thread - I have not confirmed them)

This post has been edited by catdog2: Nov 23 2004, 08:55 PM

[ferrari_rulz_02]

QUOTE (catdog2 @ Nov 13 2004, 03:25 PM)

One fix for this is quite simple, cut the data track on the motherboard and insert a resistor.  With the modchip off, the resistor has very little effect on the data line and so the xbox boots normally.  With the modchip on, the resistor limits the current to less than 5mA (less than one-tenth of the original value) and so prevents any long term damage occuring to the xbox.

thats a great idea, but how much would it shorten the life of the xbox? i woudl have thought that with ic's they would eitehr work, or get blown and not work

[catdog2]

It appears an earlier post about this has prompted one modchip designer to do something about it, read here.

To my knowledge all other modchips hold D0 (LFRAME) low continuously.
Edit: Xenium ICE does not do this and is therefore this is not a problem

Also, cheapmod installs are not effected by this since the LFRAME track is cut and connected to a switch instead of grounding it.

This post has been edited by catdog2: Nov 19 2004, 09:31 PM

[catdog2]

QUOTE (ferrari_rulz_02 @ Nov 13 2004, 06:33 AM)

thats a great idea, but how much would it shorten the life of the xbox? i woudl have thought that with ic's they would eitehr work, or get blown and not work

Wait for a year and I will tell you how many modded 1.6 xboxs I have seen just die one day for no reason...

Until then I cant give an exact number - only a warning about it.

[ferrari_rulz_02]

ok then, os its not somethign i have to worry about then. its just that i install mod chips for people, and i dont want them blowing coz of too much current through an ic.

[catdog2]

QUOTE (ferrari_rulz_02 @ Nov 13 2004, 07:32 AM)

i dont want them blowing coz of too much current through an ic.

That is what will happen - but not necessarily immediately.

As an installer you have the choice of adding a 3 cent resistor, or having people bring their xboxs back to you complaining when they stop working.

[catdog2]

Edit: Double post

This post has been edited by catdog2: Nov 13 2004, 07:56 PM

[alladdin]

catdog, do you have a diagram/pic  with res. value. thx.  this worries me  also,

[catdog2]

QUOTE (alladdin @ Nov 13 2004, 11:45 PM)

catdog, do you have a diagram/pic  with res. value. thx.  this worries me  also,

There is a link to the diagram at the very top of the first post.

[Chancer]

this is an excellent post. I shall certainly be following your advice. Thanks catdog2 it is appreciated.
This should be pinned by one of the mods

[Perplexer]

Definitely!  Everyone, including (and especially) the modchip manufacturers, needs to see this.  I will start doing this as well.

[Morglum]

Hmm this is interesting, i seem to recall that both Xenium and SmartXX modchips only pulsed the D0 & LFRAME anyway to reduce stress such as this. Will have to see if any of the members from either team will comment on this matter.

Untill then though, do you have a link catdog2 of where we can order some of these needed resistors from online?

[XOR]

I think it's this one...

RS Components : 348-0443
www.rswww.com

Please correct me if I'm wrong.

This post has been edited by XOR: Nov 14 2004, 03:21 PM

[Chancer]

Thats the wrong part.
From RS the part number is 348-0566
680R SMD resistor.
A standard carbon 0.125watt 680R also fits in here easily and is very easily obtained all over.

[hippo]

Testing on my new AFRAME signal is going well. Although designed for the 1.6 I was suprised to see this working for all xboxes extremely reliably. I might leave a pad with the old D0 signal but I think the new signal will work for everyone.