A7N8X High Vcore Shutoff fix

**YoupY** · 10-17-2003, 11:15 AM

I'm having my third A7N8X now and with all three I'm having the same problem, I can't raise the Vcore as much as I would like. I have tested this with three different procs, but every time I come in the neighbourhood of the 2.1V my system shuts down.

So I started thinking when I saw the Epox mod. This one inspired me to go looking for a datasheet of the voltage regulating chip near the cpu.

Studying the datasheet I saw that there is an overcurrent protection that prevents the the Vcore from getting over 2.1V and that the maximum that the chip can deliver is 2.25V.

OVERCURRENT PROTECTION REALIZED USING THE LOWER MOSFET’S RdsON OR A SENSE RESISTOR

So I know this protection is real

Output Voltage Protection and Power Good
The output voltage is monitored by pin VSEN. If it is not within +12/-10% (typ.) of the programmed value, the powergood output is forced low. Power good is an open drain output and it is enabled only after the soft start is finished (2048 clock cycles after start-up).
The device provides over voltage protection; when the voltage sensed by the VSEN pin reaches 2.1V (typ.), the controller permanently switches on both the low-side mosfets and switches off both the high-side mosfets in order to protect the CPU. The OSC/INH/FAULT pin is driven high (5V) and power supply (Vcc) turn off and on is required to restart operations. The over Voltage percentage is set by the ratio between the OVP threshold (set at 2.1V) and the reference programmed by VID.

and is giving me some trouble.

I have talked this over with some other tweakers and were pretty sure that pin 10 causes the computer not to boot at voltages higher than 2.1V. The solution seems simple, just cut the pin and see what happens. As fas as we think 2 things can happen:

The mod works and the overcurrent protection is disabled or the mod doesn't work and automatic shutdown will occur even at voltages of 0.1V or higher.

What do you guys think that will happen when cutting pin 10. Or are there some other idea's / theories over how to get a higher Vcore on this board.

**YoupY** · 10-18-2003, 01:08 PM

^^^^^^

**kokoro_kara** · 10-22-2003, 06:09 PM

Any Luck out there??? I wish I could help but this is way behond me. I wish someone would answer.

**YoupY** · 10-23-2003, 08:48 AM

I still haven't had time to test the mod. Since this board was going dead on me I have installed another mobo, but I will test this soon. Probaly sometime this weekend or next.

**Bra!nFreeZe^** · 10-29-2003, 03:06 AM

Hey nice info there!

ill try the mod when i get home today,im currenctly at work

ill post the results tonight !

**YoupY** · 10-29-2003, 09:57 AM

I'm not sure if all this info is right. I'm nog an elektro guy, but when I saw the datasheet it was pretty sure to me there is a way to disable this protection.

When cutting pin 10 doesn't help pin 13 is also an option. This should disabble the proctection for high amps. If the mod doesn't work soldering the connections back together should revive the mobo again if the chip survives.

Until now I didn't find the time to try this mod myself, but I intend to do this mod somewhere in the future.

**xDUCK** · 10-29-2003, 11:11 AM

Good work! Looks like an open and shut case, IMHO. As you said, just cut the leg. If the board freaks, you can always solder back the connection.

After cutting the leg, you might want to run the board without the CPU in. Just power it up and measure vcore (with a multimeter). Just a precaution, so you don't fry your CPU.

**bias_hjorth** · 10-30-2003, 06:36 AM

where do you measure the vcore with a multimeter on the A7n8x?

**Bra!nFreeZe^** · 10-30-2003, 07:09 AM

Thats what i would like to know!

**xDUCK** · 10-30-2003, 08:35 AM

Originally posted by Bra!nFreeZe^
Thats what i would like to know!

If nowhere else, the caps on the back of the socket should have a VERY accurate reading of vcore (mind you, it's a little inconvenient - I would recommend soldering a little extension line to it so you don't have to flip the board to read vcore

).

**Bra!nFreeZe^** · 10-30-2003, 09:20 AM

hehe thx for the info,but would you mind taking a pic of the place where u monitor the vcore?

**YoupY** · 10-30-2003, 11:57 AM

Vcore can also be measured on the mosfets. I'm not sure which one it was, but is is one in the bottom row. I thought it was on the left legs was for Vcore

**xDUCK** · 10-30-2003, 12:45 PM

Originally posted by Bra!nFreeZe^
hehe thx for the info,but would you mind taking a pic of the place where u monitor the vcore?

I have an NF7-S (although I've had 2 A7N8X's in the past)... But any of the caps on the back of the socket will give you a vcore measurement.

**Bra!nFreeZe^** · 10-30-2003, 01:34 PM

hehe okay thx!

**YoupY** · 11-05-2003, 01:05 PM

Bra!nFreeZe^ have you tried the mod.

Since I'm currently working 6 days a week I have not found the time to test the board, because I only have one system. I'm currious if you could tell me if it works

**Bra!nFreeZe^** · 11-05-2003, 02:38 PM

Ill do it this week,the reason i havent done the mod yet is because i wont go above 1.85 on the vcore with my currenct aircooling but this week i will get my peltier cooling up and running,and then ill try out the mod

**kokoro_kara** · 11-07-2003, 07:41 PM

Originally posted by Bra!nFreeZe^
Ill do it this week,the reason i havent done the mod yet is because i wont go above 1.85 on the vcore with my currenct aircooling but this week i will get my peltier cooling up and running,and then ill try out the mod

:banana4: :banana4:

**Holst** · 11-08-2003, 05:10 AM

Hi all,

I took a look at this weeks aggo when I was thinking about buyin an A7n8X, but as I got an NF7 instead I never tried it.

When I look at the PDF its pins 13 and 16 that stand out.

They are the cirrent sense pins that controll the overcurent protection, as this is a 2 channel controller each channel is sensed seperately.

Cutting pin 10 and connecting the pin to Vcore out may also work, but you need to be carefull what you attach to. Idealy you need to get onto the Vcore plane once its combined... IE AFTER the output caps, as I dont have a board I cant tell you exactly where but it will be very close to the socket. Pin10 also affects other things.. .so he result of such a mod is a bit unknown.

When the overcurrent is activated the vreg will not shut down as in other designs but will enter "quasi constant current operation" basically it will drop the Vcore by skipping cycles untill the current drops below the threshold... this I imagine would cause very poor stability... is this what you guys experience?

If we look at figure 4 inthe PDF we can see a brief diagram of how the overcurrent operates, basically its all set by one resistor between the ISEN and PGNDS pins...

If you look on your boards between pins 15 and 16 and pins 13 and 14 you should see a resistor thats the same size on both sides... this is the resistor "Rg" by altering its size we can change the OCP setting.

If we look at the equasion for "the current information produced internally"

This is Iinfox = (Rsense Rphase) / Rg

So we want the reported current Iinfox to appear to be less than it is... as Rg is on the bottom of the equasion we make Rg larger to make IinfoX smaller.

I "think" this relationship will be linear, add 50% to Rg and add 50% to the overcurrent.

So the mod is simple find the two resistors between pins 15-16 ans 13-14 and increase there size by 50% or so.
You may be able to remove them all together, although id prefer to keep them intact.

Using a pair of fixed size resistors (on sort wires if space is cramped) would be the best.

**Bra!nFreeZe^** · 11-08-2003, 05:55 AM

You may be able to remove them all together

isnt that alot easier than just adding new resistors ?

**Holst** · 11-09-2003, 12:11 PM

If it works yes, but im not sure if it will, that seemed a bit too easy to me...

**Bra!nFreeZe^** · 11-09-2003, 02:04 PM

hmmm i think i will try just to cut the measurement pin first, and if that doesnt work then i will try your suggestion

**Holst** · 11-09-2003, 02:28 PM

Cutting pin 10 almost certainly wont work.

You will have to solder it to the Vcore output.

**YoupY** · 11-09-2003, 02:30 PM

I think the measurement pin is still the best option to try first, because "the problem" only occurs while pushing Vcore over the 2.1V. When doing this in de bios it will result in an no startup. When I do this in Windows with a VR Windows will freeze.

Because of this I think it has more to do with the Volts that the chips has to produce than with the amps that is uses.

But If cutting Pin 10 doesn't work then modding Pin 13 - 14, 15 - 16 is also a good option. Maybe it is possible going over the 2.1V when both mods are done.

I still want to try this, but time is a §§§§§. I have way to much work to do lately

and since I have only 1 sys which is all build in, it isn't easy for me to test this in a few minutes and than build in my wc setup, prommy etc. in an hour or less sometimes

**Bra!nFreeZe^** · 11-09-2003, 02:34 PM

to test this in a few minutes and than build in my wc setup, prommy etc. in an hour or less sometimes

Of curse not i will test this tuesday or wednesday because im going to finish my pelt setup there

**sysfailur** · 11-09-2003, 11:44 PM

So no firm solution yet eh? Damn

It would be rather nice to get some higher volts.

Thread: A7N8X High Vcore Shutoff fix

Thread Tools

Search Thread

Rate This Thread

Display

Hybrid View

A7N8X High Vcore Shutoff fix

Bookmarks

Bookmarks

Posting Permissions