Transformerless power supply improved-to Secuture

I must say that this linear regulator is impressive trully a advanced one. Can u try use first two series pass transistors as scr equivalent replacing bythem two diodes?? U pass linear regulator dissipate around 6 watts in 230 volts mode and those losses simply should be impossible to reduce when it is resistive regulator. Resistance must dissipate power when act as linear reductor. If it will not dissipate then simply it is not a resistor at all SIC! To reuce this six watts u can try to suplement those two transistors in rectifier and drive them as thyristors to introduce a synchronous regulated rectifier acting as preregulator in the same manner as thyristor one. It chop ac so on cap a lower voltage will be developed and as voltage whose linear transistor must take on chest will be reduced a power dissipation will drop to almost zero level. Another option is use a voltage dividing rectifier (oposite to voltage multiplier) those exist even as most ppl don't even think this is so easy to achieve and even possible to reverse operation of multiplying passive diode charge pumps.

Why not a capacitor instead 500 ohm resistor?

I could be wrong here, but I'm thinking that replacing the two 500 ohm resistors with 1mF capacitors would net VERY similar results to the resistors. You could debate advantages for each, but I think it would work fine either way. However, decoupling after each of tree two resistors with something like 4.7uF (or maybe even 10uF, I'm making rough estimate) capacitors would probably smooth the ripple a little. How much? I'm not sure, but like I said, "a little." I doubt it would be much. It looks like voltage is from 8 to 12v coming out of the rectifier. I doubt decoupling caps could bring up that lower end voltage up much. Increasing capacitance in them to much would smooth out the ripple moreso, but it would also bring down the higher voltage beyond acceptable limits. So, all in all, you could replace those 500ohm resistors with capacitors, but I don't see a real advantage or need there; you could keep the resistors and use capacitors to decouple at the resistors, but there's still no practical advantage. Now, you could increase the capacitance of the decoupling cap already in place after the Darlington Transistor (or add a second, larger one), and that will, for sure, significantly smooth out the ripple. You could probably get the ripple down to just a few mV with just a 1uF cap. This comes at a price tough. Increased capacitance at that point in the circuit will also drastically increase the current being drawn. With that same 1uF cap, you would probably double the current. And that, would entirely negate the efficiency of this particular circuit. You might have a little room to increase capacitance there before you start losing a significant amount of efficiency, but without playing with the circuit a little, I can't say how much you'd stand to gain from it. So I would say don't bother to replace those resistors for capacitors, and if you do play with the caps at all, maybe, maybe increase the capacitance after the Darlington to try to smooth the ripple some, but only a SMALL amount.

By the way... Very nice circuit here, @Voltransistor. Thanks for sharing! I'd have to say that unless they're an expert with electronics, people are going to have a hard time improving this design.

Meh. I don't think a larger decoupling cap would be too bad there at higher loads, but I still think it's more trouble than it's worth.

Something called voltage dividing charge pump is required here. This is simple charge pump but works backwards. Means not multiply but divide voltage. Its looks like ordinary diode capactor pump. None inductances are in use or similiar. Modern ones implement inductance in form of reactance diodes (in hi frequencies) or inductively armed gyrators. But use of them is only need if u worry about reactive currents whose do nothing but stuff the wires with unwanted Amperes but not affect any opertion of rectifier at all. But as we don't pay for reactive currents unles we have some factory so we no need to worry about them.. this divider is easy i tuning output voltages more i have a full wave one whose are rare cuz most of them is half wave. I even say that they overall are damn horrible rare devices. Most do not even take in mind that those exist at all or thinks as usual " it's impossible" and that's wrong cuz they are extremely handy when power supply is in urgent need. Some design allow to use electrolytic caps but i personally will never use those alluma barrels filled with yuck ...a goo like corrosive thingy. As they ripple factor not exceed in most of them a 40% so any pulsing dc whose lod them more han 40%, of their capacity simply dry those out within a two weeks or even faster if used frequently making from them a power wating resistor. A one cool advantage is possibility to implement a eay controll scheme by replacing one of diodes by controlled switch and use it like thyristor. Caps will nicely approximate chopped AC wave to nice sine again and we can get bythis way a extraordinary good preregulator and voltage reducer to set voltage as close possible tonseries pass stabilisers make them almost lossless devices. Aripple regulated power supplies is name for such very rare alternatives for switcher. I post in just a moment one votage reducer without controll implemented.

Thanks! Lol. Wouldn't capacitors blow up like a volcano, when it's used more than 100v and more than 1uF?

Yeah, I didn't really pay attention to the voltages you had set on the power sources lol.

Ok one is posted by now. A little late i know. I post one topology whose patent protection time ends by now so it can be freely used. I chose such it can have possibility use transistor or bettr a thyristor to even more fine tuning output voltage where main drop is on dividing network and fine tune is performed by phase angle tuninng of whole divider operation. It not introduce insane ammount of elctric trash thwt thyristors produce wh used as bare hard switchers. Post a half of circuit cuz for full wave opertion it's simply double itself to maintain second half wave. As for caps i say one thing on market those can be easily obtaiin l in any electro-install related stores. A whole bunch of directly operated from mains AC caps are on offert. Those are widely used in any single phase squirrel caged motors whose power excess more than 400 watts typically up to 3,5 kw a larges single phase are build. Without them those engines are totally inoperatable. Those caps are found also in most modern metal-halogen electric arc light sources most widely used in outdor light assemblies or hi pressure sodium vapor arc lamps also most wide used for light up roads and cosssroads due to orange light they emmits whose are best anti fog light that can ever invented. Also a popular cfl-tubes also use them when operted with electromagnetic ballast to catch up reactive currents. Due to this continous usage a very small "Power Capacitors" are still widely produced due to constant demand. They are typicaly constructed from ranges so ultra small like 2,5 or even a 1,5 uF up to around 40 uF and in countries whose use 110-120V power up to 100 -120 uF take in mind that those are true capacitors not a electrolytic . Means a metalized polypropylene or polyethylene foil based (MKP, MKT) and with foil grade used that allow 100% ripple toleration under continous use over whole lifecycle counted i tens of years continous direct mains operation without any current limiting and break in opertion. But they still are prone to higher harmonics (causing self resonance and overheating) and high voltage spikes. Even as they upper limit is set around even a one kV overvoltage toleration during short time. But in trashy mains such spikes are frequently happens. I must say that 10uF cap usd with 230V mains can easily conduct currents capable to power up two 50watts bulbs without noticeable voltage drop so it will surely poer up even three of them. So even if it is a small vsalue at first glance u must be known that capcitor efficiency is not linear function when voltage rises but a expotential one so typically a few microfarads is evn too much for our power needs typically