Why is my wein bridge gain setting supply voltage dependant

[Dan Moos]

I have a wien bridge breadboarded with a split supply.

My filter network is 10k resistors and 470 nf caps. I get ~36 hz (as expected)

My gain is set with a 20k resistor from out to inverting in, and a 7371 lamp in series with my resistor subst. box to ground. I buffer the output withe the other half of the tl072 I'm using.

I have the supply set at + and - 2.5v

So if i increment the supply by 100's of millivolts, I have to adjust the gain of the circuit with my resistor box. If I raise the supply. I have to lower the resistor (raise the gain) to maintain oscillation.


Why? Shouldn't gain be independent of supply, at least at such low variation in supply? and shouldn't the lamp be dealing with this anyway?   

[Benta]

The 7371 is rated at 12 V, 40 mA. I doubt that a TL072 will drive that. Also, a TL072 will be unhappy operating at just +/- 2.5 V.

[Dan Moos]

The 7371 is rated at 12 V, 40 mA. I doubt that a TL072 will drive that. Also, a TL072 will be unhappy operating at just +/- 2.5 V.

I can't find any detailed data on the lamp as to its low voltage characteristics. Certainly I shouldn't run it at its rated voltage, right?

[Dan Moos]

ok, I just raised my supply to +- 5 volts. The thing is now way more stable withs adjusting the supply. Guess I was just running the opamp  to low.

[Benta]

[I can't find any detailed data on the lamp as to its low voltage characteristics. Certainly I shouldn't run it at its rated voltage, right?

No, but it should at least be heated. And incendescant lamps have a much lower resistance when cold (5...7 times lower), so it'll draw more current than you might think.