Op-amp project question

Hi everyone,

I'm new to the forum and am pleased to be a member. I have been working on a project where I keep a light bulb at a constant temperature using op-amps. Essentially I take the output of a thermocouple, feed it into an in-amp to get a decent gain, and feed the output of that into a comparator. Depending on polarity of that output either a) light a green LED or b) light a red LED and turn on a fan which points at the bulb to cool it down. Pretty simple idea, I'm just encountering lots of problems when building it.

The output of my last stage won't power the fan. Before connecting the load I get a saturated output of 10V. Connecting the fan (which has a load resistance of 47k) drops the output voltage to 2V and produces a strange beeping sound from the fan. The fan works fine when tested by connecting to a DC source. Why is the op-amp losing output voltage to such a large load and refusing to power the fan? Thank you very much for any suggestions.

 

What opamp and fan are you using? 47kOhm seems high for a fan.

 

Using a 741 for the output stage. The fan is a 4" 12VDC cooling fan from radioshack.

 

No way can the 741 power a fan.

 

Is there a better choice for a different op amp in this situation?

 

Thank you for looking at the circuit.

 

The 12 inch cooling fan from Radioshack requires 4.2 watts to power from 12 volts, that is 350 mA.

 

Would a power transistor do the trick do you think?

 

A power transistor would work. Just be sure to get one that can support the required current with a bit extra.

 

I was under the impression the op-amps current output was dependant on the source powering it. No current flows in or out of the op-amp inputs so any current requirements come from the power supply.

Am I wrong on that?

 

Look for a MOSFET, a special type of transistor. You'll use it like a switch. With 10V or so from the output of your op-amp applied to the "gate", it will be fully on, allowing current to flow from the "drain" to the "source", which is typically connected to ground. When "on", a MOSFET has very low internal resistance and it probably won't even be warm. And there's virtually no current required to drive the gate, so the little op-amp will love it. A normal BJT requires roughly 1/10 of the load current at the base. If that's over 30mA for your fan using >300mA, that would be a lot for that op-amp.

 

on the side - why only one of LEDs is connected to current limiting resistor?

 

Since 741's current limit is about 20mA in practice, it seems OK to do that I think.

 

Thank you for that suggestion wayneh. Is there a way to implement that to achieve a certain output current based on the input current? Or is the (Beta) value just fixed? Also: do you think I can get 300mA out of a single MOSFET?

 

and much more than 350 mA... pick power mosfet with low on-resistance.
the reason i mentioned led is that you already have resistor (you could simply connect both LEDs in antiparallel) and if you don't, that LED will clip the output.

 

The first opamp has an input offset voltage that could be as high as plus or minus 6mV which is much more than your signal. 6mV times a gain of 1000 produces an output of 6V.
You need to add the input offset voltage nulling circuit shown in its datasheet.

 

Not in the switch (comparator, on/off) configuration, which is what you want, IMHO. Of course you could build a current amplifier. I think what you're asking about is actually a voltage (at the thermistor) to current (at the fan) amplifier. I wouldn't bother unless the thermostat doesn't meet your needs. I bet it will.

If not, a better way to control the fan proportionally, instead of on/off like a thermostat, would be pulse width modulation (PWM). The PWM approach would use the same on/off MOSFET switch but you would control the duty cycle of a square wave fed to the gate. Much better than trying to control the fan's current.

Heavens yes, sky's the limit. Just choose a MOSFET with a peak current spec maybe 4X your application.