I thought it might be a neat idea to use a capacitor instead of a transformer to save a couple of buckos. I plan to place some extra inductors in the circuit to compensate for the removed isolation. The load will be two voltage regulators powering 3 optocouplers, one op amp chip and a adc. Except from the isolation, Is there any reason i should stick to the transformer?
RippoZero:
I thought it might be a neat idea to use a capacitor instead of a transformer to save a couple of buckos. I plan to place some extra inductors in the circuit to compensate for the removed isolation. The load will be two voltage regulators powering 3 optocouplers, one op amp chip and a adc. Except from the isolation, Is there any reason i should stick to the transformer?
The Capacitor will act like a resistor [an AC resistor]. So, like a resistor, it will drop different voltages depending on the current being drawn. If the current will be fairly steady, this can be a viable solution. BUT, as you pointed out, there will be no isolation, and thus the circuit will be dangerous. Make sure it is well insulated, with no conductive paths to anything living! You said you planned to add some extra inductors to compensate for the removed isolation. Unless those inductors will function as a transformer, there will be no mitigation of the lack of isolation!
Also, this is only a viable solution for low current demand--which it sounds like your case is. The greater the current, the larger the capacitor needs to be, in order to achieve the same voltage drop.
I've done this sort of thing to power a string of LEDs, which is a good case because the LEDs draw a steady current. Attempting to do this with a motor, for instance, is probably NOT a good idea, because a motor draws a wildly varying current during its operation. A huge spike to get it started, and then if there is a varying load, the current will increase or decrease as the load changes. And, such varying current, will cause a corresponding voltage change across the capacitor, and thus across the motor. In other words, poor regulation.
Also, there is a special kind of capacitor that you should use. I forget what it's called, but it's guaranteed to fail open, thus making the failure mode less hazardous. Maybe someone else, here, will fill in that blank.
Thank you for the input! I think they are called class y capacitors.
If there are NO exposed conductive parts on your circuit then the use of capacitive droppers might be ok. However if ANY part of the circuit is capable of being touched then the use of capacitive droppers is extremely dangerous. It is not unusual for such circuits to be connected to the power line by an unpolarised 2-pin plug and hence there is a 50:50 chance that the line potential (110 or 230 volts AC) will appear on the circuit 'ground' or 0-volts line.
Yes, class X and class Y safety caps.
The class X is much cheaper (still expensive as capacitors go) and smaller, these caps are designed to fail closed. You're supposed to use those with a fuse, so the cap failing will blow the fuse. For most uses this should be good enough.
Class Y is designed to fail open, and considered higher level of safety.
As jackray says, this is DANGEROUS and it would be illegal to sell such a product unless it's completely enclosed & isolated. For example your multimeter is completely enclosed in a plastic so it's OK to measure AC line voltage (as long as you don't touch the probes).
And, while you are developing and debugging your design it will have to be accessible so it's a dangerous thing to develop, and if it's an Arduino and you connect the USB & AC at the same time, you could fry your computer, and the Arduino, and everything else.
I'm not sure what your device function will be but by the time you fuss with the capacitors and add inductors might you not be better off using a plug in adapter (wall wart)? If you expecting to sell these I suspect this is the wrong forum.
I would be interested to know how the addition of some inductors could (at least partially) compensate for the lack in ohmic isolation.
John
RippoZero:
I plan to place some extra inductors in the circuit to compensate for the removed isolation.
That does seem to indicate you have no idea what you are doing! 
If you are actually going to build this type of device you MUST have an isolation transformer if you want to debug and test it, and they are far more pricey than what you will save by using capacitors rather than transformers.
Maybe we should just let Darwin's theory of evolution take its course 
Maybe we should just let Darwin's theory of evolution take its course
But there must be points added for at least asking 