Simple FM Transmitter

https://makezine.com/projects/super-simple-fm-transmitter/

Close the switch and wait some seconds for the oscillation. I found having a switch is the best way to make oscillations work in EveryCircuit.

Um... the oscillations die out. (Using latest Android EC)

Eekee, you are right. It oscillates stable for a minute around 89 MHz and then slowly declines. I have built this and made the coil from copper wire (4-5 turns). It transmits a audio signal fine at a range about 30 meters without antenna.

I think everything is OK, only the methode to start this thing is a bit tricky. start sinulation with closed switch, so the DC operating points are alread stable, e.g. if you start with open switch the 10nF at the base is uncharged and starts at 0V and it takes endless in this tramsient simsetup 5ns/s to get at its final DC voltage around 1.8V. So again start with switch closed and then shake to inject some noise. It will start and if you monitor the voltage at output which is suppose to be the 470 Ohm resistor you will see the voltage with every periode gets peak to peak higher! Voila circuit works, its just a matter to setup correctly the simulation environment which is more then 30% of all cases on EC the main issue. Specialy RF circuit are sometimes tricky like this one. Hope his helps

here another trick to start this circuit, read description http://everycircuit.com/circuit/6683823220457472

Yes. That is working. Thank you, hurz.

@schmobi2001, you are welcome!

The simulation shows a oscillation around 89 MHz. In real life it is around 120 MHz. Doing the math shows, that it is right. How can I fine tune this in EveryCircuit?

then show me the math. If i set 120MHz and take 170nH as exact given i calculate according to w² a capacitance of 10.35pF, 10pF we can see easily but where are exactly 350fF left over?? Show me your math, but dont forget we have to many variables we do not know its exact values, its starts with 170nH, the parasitics around the transistor, f_transist of the transistor, e.g. set the hFe down to 10 from 100 and you get 92MHz and comeing closer to 120MHz. I wont expect 120MHz from math is more precise then what EC does with its information, i think your math is just lucky enough to come closer to 120MHz, and with a little change in your formulas we will see a totaly different result. So show me your math and lets see what we might change or take differently into account. Again im quite happy with this delta 92 to 120 (measurements are not precise) everycircuit in RF would need H parameter to have a change to simulate more precise like Spice.

Of course you are right. There are too many parasitics. It even changes frequency when I come close to the coil with my hand. A crystal can solve this better. Do you think they will implement in EC?