Paul said:
The part you're missing, is the impedance of a motor is not a constant.
A normal resistor on the other hand, is a fixed value.
If you account for the power in the circuit, one fan draws P, and
two fans in series (based on your observation they are at full speed),
are drawing 2P. Since V is fixed, the current through the series
circuit must be 2I. V * 2I = 2P. All the power is accounted for.
The impedance of each motor has magically cut itself in half.
If the fans did not achieve full speed, then not as much additional
current would be needed to account for the power used by the fans.
And the degree of impedance change would not be as great.
In any case, the impedance of the fan is a function of its operating
conditions, with a stalled motor representing the lowest impedance
condition. (Barring any funny behaviors caused by the commutation
function used in brushless motors.)
If you put enough fans in series, it is just possible that
one won't spin, and several of the others will do well. The
stalled fan would have the lowest impedance of the lot. I'd
try the experiment here, but I don't have enough identical fans.
Paul
I tried my own experiment here, with a couple Sunon KDE1208PTB1-6
12V 2.6 watt fans. A 12V supply was used (my computer). I tried
series connecting the fans, then tried Fan #2 by itself. This is
what I got.
Test_Case Fan #1 Fan #2 Current
Series Connected 2300RPM 1663RPM 0.11 amps
Single Fan --- 3200RPM 0.16 amps (1.92W, less than the rated value)
So no violation of common sense was needed. The fans in my case, did
not run at full speed. Neither did the current do anything crazy.
Perhaps if you know the appropriate equation for power needed versus
fan speed, you can check whether the input power and output power
(air movement) agree or not.
The fan speed was measured with a home made optical tachometer. When the
tachometer is pointed at an incandescent light bulb, the readout said
"120" (and we're on 60Hz electricity here). I expect a light bulb uses
both "bumps" of the sine wave power, and that is why a 120Hz component
was detected.
The fans have seven blades, so I take the readout off the meter, divide
by the number of blades (seven), then multiply by 60 to get RPM. The
tachometer operates in transmission mode - a LED flashlight is held on
one side of the fan, and the phototransistor is held on the other
side, to get the chopped signal.
So the series connection did reduce the current flow. My fans didn't
run at full speed for me, and the current flow reflects that to some
degree.
Paul
