VMCH Mod Has Me Nervous. P5B-Deluxe.

[Kunaak]

I did every mod available to this board today.
the Vcore and Vdroop turned out perfect.
now when I select 1.5 volts, it's 1.5 volts, not 1.38 volts...
and the droop, is virtually zero.

the Vdimm, I did with a 5K fixed resistor for a .2 offset, and that turned out perfect too. set 2.45 in bios, get 2.65 in windows.
worked great.

but the Chipset Mod (VMCH) has me very nervous as I wonder if the mod is correct...

I did the mod, exactly as shown here... but regaurdless of how much I tweak the VR, the voltages never match bios, or even surpass it.

I set 1.65 volts in bios, and measure it, even with minimum resistance at 0K or close, and I still only see 1.45 volt or close.

right now, I have it set at 1.45 volts in bios, and it only shows 1.28 volts when measured from the measure point shown.
now, I've done this mod twice today, trying to see if the VR was no good or something, and they are both fine, and the solder point is correct and theres no problem with the mod itself.

I also took readings of the VMCH when there was no mod, and the voltages again, didnt even come close to what bios says...

so I am worried that this may not be the proper read point for this mod...

and if it's not, then thats definatly a bad idea to be adjusting the VR, if the read point isn't right.

thing is... when I adjust the voltages, I am seeing a voltage change I can read with my meter...

however, even at absolute least resistance, the voltage is still far lower then I would expect...
like I said, set 1.65 volts in bios, and then measure it... it shows 1.45 volts max.

I can't think of whats going on here...

[persivore]

Try checking Vmch at the inductor leg that I've marked. There should be some exposed solder that you can connect to.

The voltage seen at the leg marked on the MOSFET will probably be a high frequency PWM signal switching between 5v and 0v, not the DC voltage which is used as the output of the voltage regulator.

[freecableguy]

Why does that not look like the right IC to me? Can you show us where on your board that you have installed the pot?

FCG

[Kunaak]

FCG the place this mod is done, is right above the first PCIE slot, just below where the WIFI card is.
it's a very tight fit, so I did a alternate ground point, but that shouldnt make any difference, right?

Persivore, I think your onto something.
I measured the voltages again, from the point shown in the pic, and it reads 1.2 volts.

but from the base of the coil you asked me to read from, I get 1.33 volts.

I adjusted the voltage with the VR, and yep, the voltage does change when read from the base of the coil.

[crotale]

FCG the place this mod is done, is right above the first PCIE slot, just below where the WIFI card is.
it's a very tight fit, so I did a alternate ground point, but that shouldnt make any difference, right?

Persivore, I think your onto something.
I measured the voltages again, from the point shown in the pic, and it reads 1.2 volts.

but from the base of the coil you asked me to read from, I get 1.33 volts.

I adjusted the voltage with the VR, and yep, the voltage does change when read from the base of the coil.

While you're at it, have you conformed that the other voltages are shown correctly in the BIOS? I had a rude awakening on a P5W64 the other day, BIOS was showing 0.1V (100mV) less vcore compared to measured voltage.

[kiwi]

Did you use 50K? It works for me You can also check voltage at the points I marked and on the other side of that coil

[bachus_anonym]

This is how I did mine, Eric... Works perfect and gives at least 1.95v (that's as far as I went so far, that is).



Measuring point is correct as well (mosfet's leg), trimmer used - 50k, Pin6 to Ground. For more workspace, WiFi card has been removed.

EDIT: nvm, I just read you had already picked an alternative ground Don't mind me

[Kunaak]

ok, I tried all read points advised.
I got the VR to its least resistance, and again, the voltages reported are still no where near expected.

Kiwis read points are pretty much exact as the mosftet leg read points.
reading from the coil still gives about a .1 voltage reading higher then the mosfet leg, which is alot closer to the expected voltage, then all other read points...

the VR is 50K.

my only option left, is to redo the mod, and give it another shot with another VR. this will be the 3rd time, but I honestly dont know what to do next. VMCH voltage is not where it should be, not even close. regaurdless of if theres a VR there or not...

first thing, I think I will clip the wires to the mod and see what voltages get reported, at what setting.
cause I did check that yesterday, between doing the mod a second time, but I want to be more thorough now, with reading from various read point.

this mods honestly got me scratching my head.
it's not a hard mod, and its pretty straight forward, its just not working right.

[Kunaak]

ok, well, I went ahead and tested it all again, and just plain old redid the mod, and changed the VR and changed the ground point.
now it appears to be working...
however, the voltage read point still seems to be abit if a mystery as to what read point is correct.

heres what I mean.
set 1.25 volts in bios, and read from the coil and you get 1.28.
read from the mosfet leg, and you get 1.20.

at every voltage, the readings are very different, but neither are close enough to be more accurate then the other.
both are .03 to .04 volts off, regaurdless of what one you read from.

so set 1.25 volts in bios, and you dont get 1.25 from either read point, but both read points are close enough to be nearly the right spot, and both read points do change voltage with the bios changes....

however, having a .08 to .1 voltage difference between the two read points is pretty signifigant, so I think it's important to try and figure out what voltage read point is the right one...

but how????

cause if I go by the mosfet leg and thats the wrong voltage, then I am really running 1.8 volts, when the leg shows 1.7...

or if the coils the right spot, then I can be running what I think is 1.8 volts, and its really 1.7...

this questionable read point can be disasterous if someone starts pushing the voltage, but does so from the wrong read point...

[snatchmonkey]

My MCH mod seemd to go a bit awry also. After the mod the readpoints showed 1.77v straight off the bat at 50k Ohms. Rebooted and changed the bios voltage and still the same reading, took the mod off and got the same reading again! Anyway, left the battery out for a bit and all is now well.

Think I'll leave this mod until I find out what is holding back my FSB, turns out it wasn't MCH voltage as 1.77v gave no improvement.