# dc fan motor speed control

Discussion in 'Electronic Design' started by leon, Dec 16, 2005.

1. ### leonGuest

Can anybody help with a simple 12volt auto fan speed controller. The signal
from the aircon controller is zero to 5Volts positive and the desired output
to the fan zero to 10V or higher at about 7 Amps. Plenty of heatsink area
available, so preferably not PWM due to complexity.
Leon.

2. ### Rich GriseGuest

Well, lessee if I can come up with something sane:

+12V +12V
| |
C [M] Fan Motor
0-5V ---B Q1 |
E |
| C
+----------------B Q2
E
|
[R]
|
12V ret.

For the transistors, you could use a 2N3055 although there are probably
better transistors on the market for Q2, and Q1 has to be able to handle
whatever base current Q2 needs to get 7 amps. The R value should be
whatever resistance passes 7 amps when it's got (5 - Vbe1 - Vbe2) volts
across it. I leave the arithmetic as an exercise for the reader.

This is a "constant-current sink", even though I like to call it a
"source", since it's sourcing electrons and I'm "only" a tech. ;-)

Just rechecked the 2N3055 data sheet, and holy crap! Vbe can go up to 1.8V
at 4A Ic on the 2N3055! Find a better transistor - maybe a darlington,
then you'll only need the one.

And, yes, since it's analog, you'll need quite a substantial heat
sink. I'd even go for putting some of the big ones in parallel,
and beefing up the driver some - just remember that equation,
and if you use three in a "trilington", you'll have to subtract
Vbe3 as well.

Good Luck!
Rich

3. ### Guest

This circuit cant meet the spec. With 5v in, youve got 4.4v at the e
then 3.8v at the next e. Allow 1.5v Vce and that leaves you only
V_battery - 5.3v on the fan. Also any circuit needs to cope with a very
dirty supply.

I'd likely use an auto relay with 2 speed setting. Cheap, simple, and
crap proof.

NT

4. ### Bob MonsenGuest

12v----------o-----o---------------.
| | |
[22k] | |
| | s
o-----)--------------g
| | d
| | |
c c |
in---[470]- b b--o---[1.2k]--o---.
e e | | |
| | | F |
'--o--' | A -
| [1k] N ^
[1k] | | |
| | | |
GND-------------o------o-----------o---'

This might work. This is an amplifying buffer, with a gain of 2.2. As in
increases, the output is near in*2.2.

The output won't start tracking the input until about the NPNs start to
conduct, so it doesn't really meet the spec. Note that at 2.5V, the
P-MOSFET will be dissipating quite a bit, so it'll need a big heat sink.

Do an ac loop analysis, and watch for oscillations. One possible issue
would be the non-constant fan load. I wouldn't trust it without a lot of
testing.

--
Regards,
Bob Monsen

Love is that condition in which the happiness of another person is

5. ### ehsjrGuest

I haven't tried it, but it might be this simple:

+12 -----------------------+
|
[FAN]
aircon 0-5 v ---+ |
| |
/ |
\ /c
500 /<------| BU941ZT
\ \e
| |
+--------+
|
Gnd

I don't know what current the aircon 0-5 v line
can provide. The transistor is a darlington,
used in ignition coil circuits, so it can take
the automotive environment.

Ed