Fan speed control

[remzy]

Basically what's happening is that the voltage stays continious but the current drops for some reason. When i connect the fan it spins at full rpm at 24 V and draws about 180mA, and after a few seconds the curret drops to about 70mA at which point the fan barely spins. The delay between the current differences is random from what I've noticed. Sometimes it takes 10 seconds and sometimes 5. Here is the schematic. I drew with hand because I was too lazy to wait for my Pentium 4 to turn on.

 

[Chemelec]

Your 2K Trim Pot, Should be a Fixed resistor of 240 Ohms. (or 220 Ohms is OK)

 

[Alec_t]

Which version/package of the LM317T, and what size heatsink are you using? The IC may be going into thermal shutdown.

 

[remzy]

Your 2K Trim Pot, Should be a Fixed resistor of 240 Ohms. (or 220 Ohms is OK)

Does it matter if it's a fixed resistor or a trim pot fixed at that same resistance?

 

[remzy]

Which version/package of the LM317T, and what size heatsink are you using? The IC may be going into thermal shutdown.

It's a TO-220 package, the LM317T doesnt get hot at all. I didn't put a heatsink on it because I'm still testing the circuit but I'm assuring you that it's not the thermal problem. Actually it doesn't even manage to get past room temperature because it drops the current too quickly to even get warm.

 

[duke37]

I have never seen a capacitor connected across the LM317T. A capacitor should be connected across the LM317T input.
Is the circuit going into oscillation?

 

[remzy]

I have never seen a capacitor connected across the LM317T. A capacitor should be connected across the LM317T input.
Is the circuit going into oscillation?

That was my falut. It's actually connected between input pin and adjust pin. Not really, I just need it to slow the fan speed a little bit when I need to.

 

[duke37]

Connect the capacitor as in the data sheets.
Check whether oscillation is occuring with a fast oscilloscope.
If all you want to do is to slow the fan speed a little, then chose a suitable series resistor.

 

[Chemelec]

Does it matter if it's a fixed resistor or a trim pot fixed at that same resistance?

That would be OK.

 

[Audioguru]

A capacitor from input to output on the LM317 is WRONG!
A capacitor from input to the ADJ pin is also WRONG!

The heating in the LM317 is the voltage across it times the current through it. The voltage is 37V - 14V= 23V and I have never seen a fan that draws less than 100mA so the LM317 chip overheats and shuts down long before its case gets warm.

 

[remzy]

A capacitor from input to output on the LM317 is WRONG!
A capacitor from input to the ADJ pin is also WRONG!

The heating in the LM317 is the voltage across it times the current through it. The voltage is 37V - 14V= 23V and I have never seen a fan that draws less than 100mA so the LM317 chip overheats and shuts down long before its case gets warm.

Thank you sir for your reply. I don't get it why are then capacitors between VIN-adj, and adj-Vout pins on the lm317 datasheet

 

[duke37]

Thank you sir for your reply. I don't get it why are then capacitors between VIN-adj, and adj-Vout pins on the lm317 datasheet

I have never seen capacitors connected that way. Show us a schematic from an official web site.

A capacitor connected across the input close to the LM317 is used to stop oscillation.
A capacitor connected to the adj pin to ground can be used but I have never used one, the performance is good enough without it.

 

[remzy]

I have never seen capacitors connected that way. Show us a schematic from an official web site.

A capacitor connected across the input close to the LM317 is used to stop oscillation.
A capacitor connected to the adj pin to ground can be used but I have never used one, the performance is good enough without it.

Here is the datasheet. http://pdf1.alldatasheet.com/datasheet-pdf/view/11665/ONSEMI/LM317T.html
And the capacitory were actually the problem. After removing them the circuit works fine. Thank you all for help!

 

[Chemelec]

Here is the datasheet. http://pdf1.alldatasheet.com/datasheet-pdf/view/11665/ONSEMI/LM317T.html
And the capacitory were actually the problem. After removing them the circuit works fine. Thank you all for help!

Look AGAIN.
One Capacitor is Actually Between the Positive Input to Negative Ground Rail.
The Other Capacitor is between the Output to The Negative Ground Rail.
Not to the Adjust Pin.

 

[Audioguru]

There is a HUGE difference between what is shown on the datasheet and what you said you did:

 

[Audioguru]

We Canadians say the same things, Eh!