help me !!!

[Anonymous]

Hello to all:
I am working in my walk-throug of weapons, the detector basically, they are two detectors of induction of pulses synchronized by a quartz clock. The transmitters carry out them with two mosfet IRF740 that next attack at two operational LM318, then it goes the sign by two floodgates CD4066 and next to two diodes 1N4148, starting from there, an only operational amplifier uA741 that it acts as integrative and that to the exit I obtain an almost continuous tension that adjusting the work point with a potenciometer, I maintain this exit halfway the feeding tension. Until here everything it works perfectly...
I hope some minutes waiting he warms the resistance after the coil (the one that leads the signal to the two diodes in opposition before the entrance of the LM318. It is here where I have problems.... I have a landslide from the tension to the exit of the uA741 due to this resistance that stays with certain temperature. When this tension begins to descend him I blow to the resistance and this it increases again, I thought of putting a fan to maintain refrigerated the resistance that places an of 470 ohm 7wats. but I appeal to this forum to know some alternative different to which I thought, some termistor, etc.
If I forget something, I request they make me arrive their suggestions.
Thank you Carlos

 

[Anonymous]

Hi Carlos,
I'm having a little difficulty with your English but I think I get the gist of the problem. Basically there is a drift of the integrator output as the transmitter warms up. This is normal and there are two main causes for this. The primary cause is the temperature coefficient of the Mosfet which is positive. This means that as the IRF740 warms up, its resistance increases, with the result that the transmitter current falls by a small amount until the temperature stabilizes. With the size of coils you need for a walkthrough, you are bound to be picking up the signal from some surrounding metal. You can initially zero this out with the pot. on the 741 but as the transistor warms up and the current drops, the signal from this metal will also fall, with the result that the integrator output will drift negative. I suspect that when you blow on the damping resistor, you are in fact, cooling the IRF740, which must be nearby. The temperature of the damping resistor itself should not have any effect. The other cause of drift is if you are sampling too close to the recovery edge of the waveform on the LM318 output. Again, as the TX current falls on warm up, the switch off time of the TX is reduced and the circuit will gradually recover quicker taking the edge away from the sampling point. This again is a negative drift which can add to the drift previously described.
There are two ways of overcoming this problem and for critical applications you can use both. One is to use an auto-zeroing circuit in the receiver, which constantly corrects for any drift. This can have a long time constant so that movement of metal through the coil is not affected. Also you can feed the transmitter from a constant current power supply which has the effect of raising the TX supply voltage as the circuit resistance increases due to temperature. Also, mount the IRF740 on a generous size heat sink to reduce the temperature change.
By the way, throw out the 741 and use something more modern. These were prone to drift in themselves. Something like a TL071 at least.
Eric.