I've been tasked with switching a number of 115 VAC, 60 Hz supplies to various pieces of equipment, and monitoring these output supplies on a PIC18. I'm confortable with the PIC and ADC side, but where I'm struggling is the analog signal conditioning of the 115 V.

Playing around with a simple circuit simulator I've come up with the following which looks to be doing the right thing, but trying to interpret transformer datasheets is proving challenging.

I don't need very much power, I just want to isolate, scale, add a DC offset, and a little bit of low pass filtering. I've allowed some overhead for up to 200 V peak (~140 VAC RMS). The aim is to get an output between 0 V - 4 V, or something close to that.

falstad circuit simulation Link to circuit simulation

Questions I have:

  1. Am I approaching this the right way? Is there something fundamental I'm missing that could cause problems?
  2. Any suggestions for suitable transformers for this application, and what would be the impact to my proposed circuit? My simulated circuit has a 1:1 coil ratio and a primary inductance of 100 H (which seems high!). Most transformers I've looked at don't seem to have this specified though. It's all very confusing!

I don't have any specific accuracy requirements, but I'd hope to be able to be within a few (2-3V) volts. Isolation wise this will be within an enclosure that's then fitted within a larger sealed rack, but I'm more worried about protecting the PIC microcontroller, obviously I can't just connect neutral to the 0V.

Any advice greatly appreciated.

  • 1
    \$\begingroup\$ If they aren't, please edit your question with the accuracy you need and your isolation requirements. If I didn't do this with an off-the-shelf isolation amplifier I'd probably do it with some roll-my-own equivalent if my accuracy needs were modest and I could make it more cheaply. \$\endgroup\$
    – TimWescott
    Commented Jan 27 at 16:55
  • 1
    \$\begingroup\$ "Whatever you can get without over-engineering a solution" means "put effort into designing it then I'll shoot it down if I don't like it". Go to whoever told you that and give them examples and see how they react -- i.e., tell them you're just going to sense the presence of any AC over 50V or so, without knowing what it is other than that. If they like it -- great. If they don't, tell them you're going to get a reading that's accurate to \$\pm 20V\$ -- check for their reaction. Keep that up until you have a specification in engineering units that you can write down. Then go design. \$\endgroup\$
    – TimWescott
    Commented Jan 28 at 16:50
  • 1
    \$\begingroup\$ TRying to scale down a voltage by adding a series resistor to a transformer primary (let alone one with unspecified inductance and resistance) seems like an EXTREMELY bad idea. Why not instead use an N:1 transformer with output voltage essentially guaranteed. These are readily available in low wattages. After that you can resistively scale the output down slightly if you must. As shown it is undesignable. \$\endgroup\$
    – Russell McMahon
    Commented Jan 29 at 0:23
  • 1
    \$\begingroup\$ CLOSE VOTERS - if you do not say why I am liable to reopen the question (and others similar) as the system (stupidly) almost always just shows the reason as "not enough detail" and this and other questions often are easily understood and adequately detailed. As the system's false-flag reason does not help the OP then closing does no good. \$\endgroup\$
    – Russell McMahon
    Commented Jan 29 at 0:26
  • 1
    \$\begingroup\$ Here are links to many small mains stepdown transformers. This is not meant to be a sourcing suggestion per se - more just showing that these are common. Choose one with an output / secondary voltage that suits. \$\endgroup\$
    – Russell McMahon
    Commented Feb 1 at 7:48


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.