Dangers of the OCZ Booster

[Kamerat]

Cut'n'paste from another thread I posted in: http://www.xtremesystems.org/forums/...290#post647290

No need to mod it, the voltage of the booster varies with the 12V line (this by using an OCZ PSU). Tried again with 4,32V, display on booster only showing 3,9V. One problem is that the booster uses the 5V line as reference, if the 5V line isn't 5V the display on the Booster shows wrong voltage.

This is dangerous for people not knowing that, the 5V line must be exactly 5V and the 12V line must be 12V for the booster working the right way. With higher voltage on the 5V line than 5V the Booster shows lower voltage than it gives and lower than 5 it shows higher voltage.

Tested again with 12V line at ~13,9V, 5V line at ~4,6V, display showing 4,8V, actual voltage is 4,32V. With 5V line at ~5,8V, the booster only showed 3,8V, boster still giving 4,32V.

Pic of one measuring point:

[Juspriss]

Mhmmm... interesting

[mrlobber]

Care to translate that text on the pic? I can pretty much guess what's meant there, but guessing-trial-error method is not the most applicable here .

[Trixxy]

I've just done some testing on mine (rig in sig).

OCZ Powerstream rails as measured with DMM:

12v = 13.041v
5v = 5.432v
3.3v = 3.881v

I set bios vdimm to default 2.5v:

with bios vdimm showing 2.5v, the booster showed 2.5v on both display & read point (using above pic), the DMM showed 2.5v at the mosfet on the dfi (to measure vdimm).

I then increased the booster knob until the display read 2.8, and the bios/booster read point and dfi mosfet scaled accordingly (all showing 2.8v).

I tested up to 3.6v, and all 4 voltage readings were in line with each other (other than usual .05v discrepancy between bios/actual). Everything seems to be fine using my normal rails.


I then decreased my 5v rail to 4.8v (had trouble booting any lower):

with vdimm set to default in bios (2.5v):

bios showed 2.5, dfi mosfet showed 2.5, booster read point showed 2.5 but the booster display showed 2.8.

turning the booster knob until the bios/mosfet/booster read point showed 3.6, the booster display showed 3.9, so it does appear that something is weird...

I haven't tried any more rail configurations (different 12v, higher 5v etc), but I will if anyone wants me to.


btw, Kamerat, wtf is that green thing in your avatar? I've actually been using that as my msn messenger pic for a couple of weeks but don't know what it is, lol (looks really cool tho).

[Pejsen]

Care to translate that text on the pic?

The actual voltage can be measured here, by using ground also.

[Trixxy]

just did another test by maxing out my 5v rail (5.62v).

using bios default of 2.5v:

with a reading of 2.5v actual @ bios/mosfet/booster read point - booster display = 2.4.

with a reading of 3.7v actual @ bios/mosfet/booster read point - booster display = 3.5.

too scared to try higher vdimm just now.

[STEvil]

green thing is called "bundles"

its a genmay thing

/thread hijack

[I_Am_ToO_SeXy]

where can i measure the 3,3 Rail??

System in Sig

[Trixxy]

where can i measure the 3,3 Rail??

System in Sig

use one of the orange wires from the psu.

On the OCZ, I have an unused 4-pin cable (looks like the 4-pin 12v ones that plug into motherboards), it has yellow, red, orange & black wires.

thanks, STEvil, that's been bugging me for ages.

[Kamerat]

mrlobber: What my Danish friend Pejsen said.

Trixxy: Bunchies. OCZ should be aware of this problem, weird thing is that the manual doesn't say anything about it.

[STEvil]

er bunchies... was close enough

Kamerat - Does the manual not say anything about calibration? Its not much of a problem if you calibrate the unit to your system

[IvanAndreevich]

Kamerat
OMG man.. that green thing is hilarious!

So the voltage isn't that much off, as long as your 5V rail is KINDA close to spec?

[Kamerat]

er bunchies... was close enough

Kamerat - Does the manual not say anything about calibration? Its not much of a problem if you calibrate the unit to your system

The manual mention calibration, but it doesn't say the readout is dependent of the voltage of the 5V rail, nor does it say that the voltage increases with the voltage of the 12V line.

Adjusted my 12V rail to max, which is 13,8V and adjusted the voltage of the Booster to 3,9V. Then i adjusted my 12V rail to min, which is 11,0V, and the Booster only gave 2,8V. Imagine someone whos not aware of this and adjusted his Booster to 3,4V when running 11V... then maxing the rail, that will give about 4,7V to the RAM. I'm lucky that I have BH-5...

[STEvil]

This is why it mentions calibration

[Kamerat]

This is why it mentions calibration

Not from what's written in the manual, doesn't mention anything about the 5V rail:

Q: I set a voltage of 2.6V (or some other value) in my BIOS, but the DDR Booster is not reading the correct value! Whar's wrong??
A: The DDR Booster may have been knocked out of calibration during the shipping process. In order to recalibrate the DDR Booster, locater the small screw adjust on the front of the DDR Booster, just under the LEDs. To adjust the readin upward, turn the screw very slightly clockwise. To adjust the reading downward, turn the screw very slightly counter-clockwise.

[ziberia]

The manual mention calibration, but it doesn't say the readout is dependent of the voltage of the 5V rail, nor does it say that the voltage increases with the voltage of the 12V line.

So:

Readout: Varies with 5V rail

Voltage: Varies with 12V rail

I'm running 5V: 5.4V 12V: 12.64V

Booster reads 3.5V.

What is your educated guess about the actual VDimm?

zib.

EDIT: MB is Neo2 Plat, which has the crappiest BIOS readouts in the world anyway...

[Holst]

Its a good idea to recalibrate the booster if you make any system changes.

Setting it to minimum then reajusting is the best way.

Dont trust the bios as thats allways wrong, use a multimeter.

The way the booster regulates voltage is a bit shoddy IMO, if I could get my system to turn on with it fitted (which I cant) I can think of a few mods that might make booster better.