Sticker shock

I started making a bill of materials (BOM) for the power monitor and realized that each node was likely to cost $30 or more.  With 30 nodes to make, I am looking at a $900 in parts. Perhaps I need to rethink the project. I was originaly going to just mux the 30 sensors to one ADE7763 chip to get Vrms and Irms and I was going to compute power and integrate energy usage myself. But after playing with the ADE7763, I fell in love with its ability to do the energy integrations. But I am not sure it is worth the $870 dollars for a unique node per branch circuit. Perhaps I need to go back and look at the mux and doing the computations myself.

This entry was posted in Home Energy Monitor, Projects. Bookmark the permalink.

4 Responses to Sticker shock

  1. ril3y says:

    I would say that 30 nodes for 900 is a good deal. Perhaps a kickstarter ? 🙂

    • Skye Sweeney says:

      It may be a good value, but it is still higher than my budget. Since I am not the government, I do need to spend less than I bring in.
      My current plan is to place a 30 to 1 analog mux in front of a single ADE7763 chip. I would lose the ability to integrate power to get energy, but that is not my primary goal. I am running experiments right now on how to make the mux. Analog muxes exist with 8 to 1 about as big as they come. I can gang several of these together to get my 30 inputs. The only problem is dealing with the negative voltage from the CT and voltage transducers. Need to find a mux that can go negative or learn to bias the sensors up so they always remain positive.

  2. Felix Rusu says:

    How about an analog signal matrix switch? That would enable you to select which signals go where (if you had multiple ADEs), so you could read all the parameter integration from the ADE, without having the overhead of doing it yourself in the micro.
    Analog has a few such chips:

    • Skye Sweeney says:

      I have indeed gone the route of a series of analog muxes. This allows me to feed one of 15+ current sensor signals to one ADE7763 and another 15+ current sensor signals to the other ADE7763. The control of these muxes (two pole four throw) is done using a very small PIC chip using a common data/clock bus.
      With this solution I am not able to allow the ADE to perform the energy accumulation (integration) as the selected current sensor is changing every few seconds. But if I choose, I could “park” on one sensor and integrate energy over an extented period of time. This software feature is certainly on the table.

Comments are closed.