I have a question for electronics/electricity wiseguys out there.
Let's take a my system that has a serious current draw from the PSU and uses +5V line for the MB (no +12V connector separately for the MB).
Now, if the PSU +5V line is measured from a 4-pin Molex it show 4.99 Volts.
However, all motherboard voltage regulator pins and bios reading show that MB is registering the +5V line at 4.73V (maximum, it may dip below that).
I know that the common wisdom is not to trust the reading from the Bios sensors.
However, if on such a system, changing the PSU changes the BIOS +5V line reading from the 4.73V back to 4.95V AND IF the PSU is NOT giving overvoltage (measured with a true RMS meter from Molex under load), then isn't the voltage loss then happening between the MB and the PSU as something that is due to an increased current demand and related resistance?
That is, should a good PSU not only give decent reading from the 4-ping molex, but also decent readings from the motherboard BIOS sensors?
How else is it possible to explain a situation where changing the PSU raises the Bios +5V reading from 4.73V to 4.94V while both PSUs give a nominal +4.98-4.99V output and do not give overvoltage?
regards,
Halcyon
Q:Motherboard voltage losses and PSU voltage line regulation
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee, Devonavar
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
It's not unusual. When you measure voltage at the 4-pin output connector from the PSU, you're measuring the voltage output directly from the PSU with only one really dubious contact point -- that between the multimeter probes and the pins in the connector. The motherboard reading is AFTER the contact between the 20-pin ATX connector and the socket on the mobo, and we don't know exactly where the voltage reading is being taken on the motherboard. The 20pin ATX connection could be not fitting quite right with the mobo, causing some voltage loss. This is probably the case bcause it doesn't happen with one of your PSU -- which maybe has a connector that mates better.
Here's a detailed relevant article from Overclockers.com on this very topic - "What's Up With My +5V?" http://www.overclockers.com/tips845/
Here's a detailed relevant article from Overclockers.com on this very topic - "What's Up With My +5V?" http://www.overclockers.com/tips845/
Jan,
I couldn't measure straight from the ATX-connector, because then I couldn't measure it under load (I have to have it connected to MB to be able to stress load it). But you are right, that's how I should measure it, if I was skilled enough (or brave enough to cut my atx connector wires).
MikeC,
that is a probable explanation.
How about the PSUs ability to feed steady voltage under a high current demand situation?
PSU 1 on Mobo 1 (using mobo sensors) gives 4.73V under load
PSU 2 on Mobo 1 (using mobo sensors) give 4.98V under load
Current demand is high and the cross section of wires seem to be different between PSU 1 and PSU 2.
Could the loss in voltage be caused by increased current demand, which raises the resistance on the wires and causes a voltage drop?
regards,
Halcyon
PS I can't get my FPS300-60PN(PF) mobo sensor voltages go above 4.75V, although my Enermax and Seasonic are able to pull this off. None of the PSUs in question is giving overvoltage on the +5V line.
I couldn't measure straight from the ATX-connector, because then I couldn't measure it under load (I have to have it connected to MB to be able to stress load it). But you are right, that's how I should measure it, if I was skilled enough (or brave enough to cut my atx connector wires).
MikeC,
that is a probable explanation.
How about the PSUs ability to feed steady voltage under a high current demand situation?
PSU 1 on Mobo 1 (using mobo sensors) gives 4.73V under load
PSU 2 on Mobo 1 (using mobo sensors) give 4.98V under load
Current demand is high and the cross section of wires seem to be different between PSU 1 and PSU 2.
Could the loss in voltage be caused by increased current demand, which raises the resistance on the wires and causes a voltage drop?
regards,
Halcyon
PS I can't get my FPS300-60PN(PF) mobo sensor voltages go above 4.75V, although my Enermax and Seasonic are able to pull this off. None of the PSUs in question is giving overvoltage on the +5V line.