large axial fans: power used is less at higher voltages?

[80251]

I'm here referring specifically to the San Ace 172 9E type, 172mm x 51mm axial fan. If you look at the datasheet for this fan and compare the power drawn at the same peak RPM for the +12VDC models of the fan vs. the +48VDC models of the fan the +48VDC models of the fan use from 0.5Watts to >1.5 Watts less power for the same CFM, static pressure and RPM. Even greater power savings are seen in the +24VDC models of the fan where for the exact same CFM, static pressure and RPM, the +24VDC models save anywhere from 0.5Watts to 3.5 Watts over the +12VDC fans.

Is this generally true for large axial fans (i.e. that the higher the operating DC voltage the more efficient the fan)?

 

[80-watt Hamster]

It's generally true of many electrical systems. More volts equals fewer amps, which means less transmission loss.

 

[80251]

It's generally true of many electrical systems. More volts equals fewer amps, which means less transmission loss.

I^2 * R losses. That makes sense, the less amperage the less resistivity losses. Is this also why PSU's that draw from 220VAC mains exhibit better efficiency than when using 110VAC US power?

 

[80-watt Hamster]

I^2 * R losses. That makes sense, the less amperage the less resistivity losses. Is this also why PSU's that draw from 220VAC mains exhibit better efficiency than when using 110VAC US power?

I don't know enough about electrical circuits to definitively say "yes", but one would presume it at least contributes.

 

[Shrek]

I^2 loss suggests the 12V fan might have 16 times the loss of the 48V fan, so I suspect the dominant loss is actually from the 0.7V diode drop, which would be proportionally more significant for the 12V fan and would also mean the 12V fan would only have 4 times the loss of the 48V fan.

 

[80251]

I^2 loss suggests the 12V fan might have 16 times the loss of the 48V fan, so I suspect the dominant loss is actually from the 0.7V diode drop, which would be proportionally more significant for the 12V fan and would also mean the 12V fan would only have 4 times the loss of the 48V fan.

The diode models I've used also include a lumped resistance which would also exhibit more I^2 * R losses, but it does seem like the difference in power drawn should be greater if I^2 * R losses were the reason.

The least power drawn is by the +24VDC model for the same RPM(all the other stats as to static pressure, CFM, dB-A are the same):

RPM/power drawn/input voltage/rated current

4100/34.8W/+12VDC/2.9A
4100/31.2W/+24VDC/1.3A
4100/33.6W/+48VDC/0.7A

2500/10.8W/+12VDC/0.68A
2500/8.4W/+24VDC/0.35A
2500/9.12W/+48VDC/0.19A

2000/5.76W/+12VDC/0.48A
2000/4.8W/+24VDC/0.2A
2000/5.28W/+48VDC/0.11A

The San Ace 140L (140x51mm) axial fan does exhibit similar characteristics for models that run at 3100 RPM, but not necessarily for other rated RPM's.

 

[eidairaman1]

PIE

 

[80251]

But why is the power drawn for the +24VDC model consistently less than the other models? 2 to 3 Watts isn't much unless you have dozens of these fans running 24-7 in a datacenter at which point at the end of the year it might make a measurable difference in power drawn.

 

[Shrek]

0.7A and 0.2A are accurate to only one significant figure; do you have a link to the raw data?

1W for a year costs about $1, so I would not worry too much about savings.

 

[80251]

There are other San Ace fans that exhibit this reduced power draw @ 24VDC vs. 48VDC, for example, the San Ace 200mm DC axial fans 9EC2024H001 and 9EC2048H001 which have the same stats except for rated input voltage (+24V vs +48V) and the power drawn (48W vs. 57.6W):
https://products.sanyodenki.com/en/sanace/search/?c16=on&s16=1

 

[Steevo]

All else being the same the power conversion factor will be equal, 1W at 12V is the same as .5W at 24V. Resistance losses only matter when passing the bend of the knee, where resistance is causing thermal output high enough to create higher resistance and more thermal output (runaway), on large motors higher voltage is used to overcome thermal issues and allow for higher RPM in brushed motors. Since these are PWM driven brushless their power consumption is related to fan efficiency far more than voltage input.

 

[Shrek]

Resistance losses only matter when passing the bend of the knee

?

 

[80251]

1W is equal to 0.5W? Huh? I'd like to see the physics that supports that assertion.

Even those these fans are all from the same model could it be that the actual control circuits and motors are slightly different because they're working with different input voltages? If they were to use the same, exact coil circuitry (as to inductors/capacitors/resistance in the models with different input voltages the inductor charging/discharging curves would be different at the different input voltages.

 

[Steevo]

?

"In a pure resistance the relationship between voltage and current is linear and constant at a constant temperature, such that the current ( i ) is proportional to the potential difference V times the constant of proportionality 1/R giving i = (1/R) x V. Then the current through the resistor is a function of the applied voltage and we can demonstrate this visually using an I-V characteristics curve.


In this simple example, the current i against the potential difference V, is a straight line with constant slope 1/R as the relation is linear and ohmic. However, practical resistors may exhibit non-linear behaviour under certain conditions for example, when exposed to high temperatures."

1W is equal to 0.5W? Huh? I'd like to see the physics that supports that assertion.

Even those these fans are all from the same model could it be that the actual control circuits and motors are slightly different because they're working with different input voltages? If they were to use the same, exact coil circuitry (as to inductors/capacitors/resistance in the models with different input voltages the inductor charging/discharging curves would be different at the different input voltages.

I x V = P
Current Times Voltage equals Watts.

So 1Amp times 12 volts = 12 Watts
So .5Amp times 24 volts = 12Watts

The characteristics of the fan windings and other nuance is the variation in fan speed at the difference in watts used at X volts, it may have been engineered to perform better due to a higher magnetic flux with a specific voltage(24V), but there will be a curve for each voltage applied, and since its a PWM driven fan unless its driven at 100% duty cycle the apparent voltage is variable making the discussion between voltages moot. The table you put together shows virtually the same watts used at each voltage.

 

[80251]

How is 48W vs. 57.6W drawn "virtually the same"?

If they're virtually the same how about I give you $100 and you give me $196 in return?

 

[Steevo]

How is 48W vs. 57.6W drawn "virtually the same"?

If they're virtually the same how about I give you $100 and you give me $196 in return?

4.8 vs 5.76 at 2100 RPM. Unless you are running 10

1 watt more is at that level is 20% more power.

what voltage do motherboard PWM headers put out?

 

[80-watt Hamster]

what voltage do motherboard PWM headers put out?

12V

 

[Shrek]

4.8 vs 5.76 at 2100 RPM.

I still suspect the problem is with the 1 significant figure for the 24V, 2000 RPM fan

If the 0.2 A was actually 0.23 A one would get 5.52 W and everything would make sense

 

[80251]

4.8 vs 5.76 at 2100 RPM. Unless you are running 10

1 watt more is at that level is 20% more power.

what voltage do motherboard PWM headers put out?

I'm referring here to the San Ace 200 model fans that also exhibit this reduced power draw @ 24VDC vs. 48VDC. Specifically, the San Ace 200mm DC axial fans 9EC2024H001 and 9EC2048H001 which have the same stats, dimensions and frame except for rated input voltage (+24V vs +48V) and the power drawn (48W vs. 57.6W). The difference here isn't 1 Watt.

 

[LabRat 891]

Without reading more than the OP:
Makes sense. Fans are (primarily) magnetic inductive loads, not resistive.

Lower voltage,
less magnetic flux,
slower rotation,
less mA draw,
less Wattage draw.

If it was a 'thermally operated' motivator, like a wax motor; then the application of ohms law is simple.

 

[80251]

What I'm curious about is why it seems like the +24V models of these axial fans use the least amount of Watts? It's not even restricted to just San Ace DC fans either.

The Delta FFB1424VHG (+24VDC 140x50.8mm fan) uses less watts (21.84 watts) than the FFB1448VHG (+48VDC 140x50.8mm fan, 24.00 watts) even though all the other characteristics are the same.

For a delta 140x38mm fan the difference in power drawn between the +24VDC and +48VDC models is even larger:

common characteristics of both models (aside from frame and dimensions): RPM: 5500 CFM: 367.62 mm-H2O: 41.75 in-H2O: .644 db-A: 74.0

FFB1424UHE -M -R00 / -F00 operating voltage: 24V 3.35A 80.40watts
FFB1448UHE -M -R00 / -F00 operating voltage: 48V 1.80A 86.40watts

 

[Shrek]

If I could choose voltage, I would get a fan running on mains voltage; probably the most efficient route.

 

[Steevo]

I still suspect the problem is with the 1 significant figure for the 24V, 2000 RPM fan

If the 0.2 A was actually 0.23 A one would get 5.52 W and everything would make sense

For sure they were probably designed with a target voltage and RPM in mind, and that seems to be the 24V at peak efficiency for magnetic flux and duration at angle at RPM.

I'm referring here to the San Ace 200 model fans that also exhibit this reduced power draw @ 24VDC vs. 48VDC. Specifically, the San Ace 200mm DC axial fans 9EC2024H001 and 9EC2048H001 which have the same stats, dimensions and frame except for rated input voltage (+24V vs +48V) and the power drawn (48W vs. 57.6W). The difference here isn't 1 Watt.

They are the same fan hardware at different voltages and coefficients as stated above, they probably change the winding counts for different voltages is all which changes the field strength and back feed loss. I have learned a LOT about voltage and amperage draw with monitoring my two phase with run cap Copeland compressor and ground loop pumps. Its more dynamic than PWM can give since the design is for a RPM at voltage and instead they are expected to tolerate a range of PWM input for startup and run.

If they gave more data points it would be easy to create a graph showing the peak efficiency for each voltage model, they probably have it and would be willing to share it if asked.

I read your chart instead of the datasheet which is why I believed it was 1 watt, my bad there, but each will exhibit the same behavior at a set voltage and RPM. Depending on manufacturing tolerances you might get a batch that all want to run at one specific RPM, my current fans vary by 200RPM with the same duty cycle, could be bearings, balance, windings, etc....

 

[tpu7887]

If you need to use a boost converter to make your 12V into 24, its efficiency will probably be somewhere around 93%, narrowing the gap to:
leave your fans on 12v!!!!
lol

 

[DeathtoGnomes]

PIE

CAKE > PIE

Its my understanding that, when it comes to fans, it takes more to bring the fan up to speed than to hold it there. There is also a bit of physics involved, this assumes there is no resistance from the fan blade spin. There is also a difference between static pressure fans and normal fan bladed fans. All factors in efficiency, not just PIE or I2R math. 24v fans that are not meant to run at 24v likely will take more to keep at desired speed, lifetime is severely reduced too.

Depending on manufacturing tolerances you might get a batch that all want to run at one specific RPM, my current fans vary by 200RPM with the same duty cycle, could be bearings, balance, windings, etc....

^^^^ This will never be perfect.