Thanks everyone for your comments... On 25/10/06, Harold Hallikainen wrote: > These sort of supplies have a tremendous shock hazard, so be careful! > Also, I'd add a resistor in series with the capacitor to limit the current > when it's plugged in at a line voltage peak. I realise that - but they have a very low part count and are efficient. On 25/10/06, Mike Harrison wrote: > obviously you'd use a 5V1 zener instead of 11V. > You'll want to add a little bit of series R (~100R fusible) to limit inrush current if it's plugged > in at the top of the cycle. What do you mean fusible? (in the context of a resistor)? On 25/10/06, Bob Axtell wrote: > I have posted it before, but I will post it again. Gotta find it... > I can only tell you that it is driving a small 120VAC 60hz > gearmotor. Thanks for the post bob On 25/10/06, Wouter van Ooijen wrote: > The other problem with this type of circuit is the voltage on the mains > plug after uplugging... This is going to be used for fixed equipment (if I actually get round to making it). On 26/10/06, Charles Rogers wrote: > Why don't you try microchip.com. AN954 might be just what you > are looking for. Taking a look... On 26/10/06, Bob Axtell wrote: > Let me make it very clear: these transformerless designs are > dangerous because they operate with exposed lethal voltages. > The only reason I did the gig was because it was to be potted > in plastic.. and it was. The design will be used in plastic cased fixed equipment, and an earth is present so I can design the innards in a safe fashion. On 26/10/06, peter green wrote: > > Unexperienced developers might risk their lives, and even design > > hazardous equipment that can risk the end users' lives > also doesn't transformerless imply not isolated from the mains? mains > neutral cannot be depended on to stay at near eath potential! In what context do you mean "rely" ? In my houses electricity supply (designated TN-S in UK terminology), the neutral and earth are joined at the substation (ie the star point of the REC's transformer where the neutral originates is grounded). So only in major fault conditions can the potential differ much - only due to a break in neutral or earth, or a major fault current on the earth causing a voltage drop on the earth. Anyway, for everyone's information, the project i'm considering is making an intelligent, anticipating room thermostat for my central heating. My current electromechanical stat is rubbish in it's performance. It's supplied by a 3 core + earth cable - phase, neutral, switched live. The neutral is only present to power the anticipator resistor. Of course, I could use a battery to power the pic, and relay to switch the mains volts. But that'd annoy me. Batteries need to be changed, and unless I put a bunch of effort into circuit and firmware design to reduce power consumption then the battery could go flat real quick. And when there's mains voltage in the thermostat, why not use it? I'd plan on using my current thermostat housing, gutting it and fitting a new PCB with a VR for the control knob. thanks jon -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist