Here is one for the PSU experts

[Strangelove]

OK I'm building a fan control which switches between 12 and 7V see picture for rough diagram (sorry about my paint job ) Here is what is very strange. I don't have a spare PSU at the moment so I tap into my server's PSU, when testing. If the computer is running it works great I can switch between speeds (voltages) no problem. If I take out the ATX connector and "start" the PSU by connecting pin 14 & 16 (I'm not sure those are the correct pins the computer is running right now so I can't check) I can run the fans at 12V but if I switch to 7V the PSU dies. After unpluging the PSU and the circut and putting it back together again it runs just fine at 12V but dies every time I try to switch to 7V. Anyone know why this is??

 

[PsycoPhreak]

an ATX power supply has to see a sufficient load on the +5 volt rail in order to operate properly, whent the PSU is not connected to anything and the "jump" trick is done to power it up, there isn't any load on it and the PSU will do exactly what your seeing....


I forget the resistor value at the moment, but if you put a resistor between the +5 and ground while doing the 'jump" trick, the PSU will stay on...

 

[Hoot]

A 4 ohm 10W power resistor will suffice. Radio Shack should have them. BTW, they get hot.

Hoot

 

[Darrenct]

Hmmm... Here's one I just thought of that I'm going to need a little help with. If I had 4 LEDs lined up, would there be a way to make them light up from first to last as I turned up the Rheostat? I know it would be tough, but I'm sure It's possible.

 

[Hoot]

National Semiconductor makes a Bar LED driver that turns on 1 though 10 LEDs as the voltage it sees rises. I haven't worked with one since the late 80s, but I think it is something like an LM3915. The chip probably costs a buck. You add a few programming resistors and voila. It can be set as a linear response or logrithmic response. Check the National Semiconductor Data Sheet page at their Web Site.

Hoot