| T O P I C R E V I E W |
| tamemin |
Posted - May 25 2007 : 09:35:05 AM
Hi there,
The electric radiator fans in my car use a resistor to give them a low speed setting for use by the air-conditioning. Just like many other cars of the same type, the resistor burns out leaving only high speed. Many people have replaced these fans only to find that the resistor burns out again after about 20k miles.
I investigated mine, and decided to not replace the fans (which contain the resistor inside them) but to connect to the control relay so that I now have the switched low-speed supply and the load connection to the fan. (If I were to connect these, the fans would run at full speed)
What I want to do is build a small device that will use pulse width modulation to drive a power mosfet, so that the fans get a 30% power supply when the low speed feed runs.
I can get a power mosfet IRF540A that can handle 30A, and 100v across drain and source. Do you think I need to add a shunt diode to protect the FET from the motor? I noticed on your PWM circuit, there is no diode.
I want to do this with as few components as possible, and was thinking about using a two transistor flip-flop, like in your inverter. What do you think? could I adjust the R-Cs to get 30% on 70% off ?
Thanks
Tamer
PS
Excellent website.
|
| 7 L A T E S T R E P L I E S (Newest First) |
| cyclopsitis |
Posted - Jun 18 2007 : 10:34:30 PM
Where did you get the circuit drawing schematic? I've been wanting to give guru a schematic for a while now but mine don't seem to be recoginized by this site.
Thanks
K |
| tamemin |
Posted - Jun 02 2007 : 7:25:37 PM
Ok here are the pics:-
Inside revision 2 http://www.nime.co.uk/downloads/fan-controller-inside.jpg
Revision 2 running hot: http://www.nime.co.uk/downloads/fan-controller-running-hot.jpg
Revision 3 almost no heatsink: http://www.nime.co.uk/downloads/fan-controller-cool-running.jpg
Circuit Diagram: http://www.nime.co.uk/downloads/fan-controller-diagram.jpg
Tamer
|
| tamemin |
Posted - Jun 02 2007 : 6:02:25 PM
Yea,on paper though. Is there a free CAD program for doing diagrams on a PC? last last time I draw a diagram on a computer, it was on a BBC model B.
I have a fixed R1 between 555 pin 7 and Vcc. R2 is between pin 7 and 6, and C1 between 6 and ground. There is a small signal diode from pin 7 to 6. The diode ensures that the cap charge time is approximate to R1 x C, and discharge time to R2 x C. By replacing R2 with a variable res in series with a 4K7, I can swing the resistance from well below R1 to well above. This translates to fully adjustable mark to space ratio. I believe it's the mark space ratio which is what PWM is about. For sure, the frequency will deviate a lot; but as the as this circuit is to be used to maintain a single speed, I'm not worried about it.
I spent quite a bit of time with the circuit today. Switched to using a single MOSFET on a Celeron heat sink and fan. I had it running my cooling fans at about 30% speed. The current draw was about 10A at idle 14v DC. Heat sink was still getting pretty hot (estimate about 70 deg C)
I really didn't feel happy about this. It was almost as if it was burning up the the power like a resistor (as you said in your last post). I couldn't imagine how this was happening unless something was stopping the MOSFET from fully switching on.
Anyway, cut a long story short, I thought I'd add a clamping diode to short out the self inductance of the motors to see what happens. I added the heaviest thing I could find reverse bias across the motor, and the heat from the MOSFET disappeared, the motor sounded quieter and best of all, the current draw at the same fan speed, dropped to 4A.
Given that is wasn't heating so much anymore, I first removed the fan, and re-tested. The Heat sink was not even warm, so went back and switched to a much smaller heat sink and this runs fine. In fact the diode seems to be absorbing more energy than the MOSFET, as it gets quite hot. I replaced it with the power rectifier type of diode you find in PC power supplies (they look like power transistors).
I learned that this clamping diode is known as a 'freewheeling' diode in motor control circles, great name eh?
Anyway, I've boxed the device and wired it into my car with a 13A fuse in the line, just in case something goes wrong and the MOSFET shorts. I hope to order some more devices and other bits, to make a much smaller and rugged version of the original.
Pictures and circuit diagrams to follow.
Cheers
Tamer
|
| Aaron Cake |
Posted - Jun 02 2007 : 10:45:14 AM
Do you have a schematic of the circuit you created? If you just used one 555, I can't see how you accomplished PWM. The circuit is likely similar to the one on this website.
Pulse frequency modulation doesn't really modify the effective voltage to the motor like PWM does, so you are sort of using the MOSFETs as a big resistor. Even wtih PWM you will get some heating. |
| tamemin |
Posted - Jun 01 2007 : 7:43:26 PM
Oh, forgot, picture of prototype:-
http://www.nime.co.uk/downloads/fan-pwm.jpg |
| tamemin |
Posted - Jun 01 2007 : 7:41:36 PM
Hi Arron,
Went for a NE555 timer in the end, as I wasn't sure about how square the output of my 2-transistor astable was, noticed quite a lot of heat disipation from the MOSFET, and thought that is was not switching on properly/fully.
The two fan motors will drag over 30A from the battery without any control. With the PWM circuit, I can adjust it from 3.5A to 15A. Still quite a bit of heat generated. I was hoping for cool/warm operation.
The current heatsink is pathetic, I'll probably try a Pentium/Celeron/AMD heatsink FAN combo. I don't know about you, but I seem to have accumulated a few of these over the years. Never had the heart to throw them out after the CPUs became obsolete, as they seemed like they could be useful one day ;-).
I added a 2200uf reservoir capacitor to the supply rails to help reduce spikes on the supply. I also experimented with a cap across the motor, and noticed that it silences the acoustic effects of PWM on the motor, but does not reduce overall current consumption, so I ditched it.
I also experimented with operation frequency too, and found that the MOSFET generated less heat at lower frequencies. I will probably settle on about 300Hz.
Can you think of any reason why the MOSFET is burning hotter at higher frequencies?
Tamer |
| Aaron Cake |
Posted - May 27 2007 : 10:32:45 AM
The PWM circuit on this website is not true PWM as the notes indicate, it's pulse-frequency-modulation. While it does work, it doesn't work as well as a true PWM circuit. You can find some better circuits at http://www.discovercircuits.com .
The diode is not necessary, but it's not going to cause problems if you use it. For the minimal extra cost it's probably a good idea.
Another option would be to use an external series resistor. |
|
|
