Availability and upcoming changes to the PCB

We are still in the process of moving production to China and is not going very well :slight_smile:
Covid has made our life far more complex making everything take double the time and effort to accomplish. We don’t expect realistically to resume manufacturing and sales in 2020.
On the bright side, speaking with some factories we were faced with some realities we need to address (or accept):

  1. Our non-SMD approach makes the project expensive. For those not familiar these are SMD components:

    They are optimised for “robot” factories to solder them in great quantities cutting down the costs in the long run but can be very intimidating to people wanting to make HestiaPi themselves. This is the reason we have avoided them so far. Not sure if this will possible in the future.
  2. The OMRON relays (G3MB-202P). These very popular solid state relays (SSR) were cheap and widely used but are now EOL/discontinued. Well they have been for some time really but suppliers stocked them enough. Moving forward we need to swap these crucial component with something we will be able to source for a few years from now. The problem is that the replacement part from OMRON (G3MC-202P) is not compatible with Pi’s pins and will require additional components to connect them. So if we are going to modify the board it is a good chance to bring up a few things we have kept at the back of our head for some time now…
    Standard relays (the magnetic ones that click)… we think about bringing them back.

Pros:

  • They click - easy troubleshooting without seeing logs or using multimeters
  • They operate on any AC or DC current (SSR only support one, usually AC)
  • They provide both NO and NC contacts (SSR only have on, usually NO)
  • Their contacts can be tested easier with a multimeter

Cons:

  • They click - may be considered old-fashioned by some
  • Because they incorporate moving parts they “should” have a more limited lifespan although we are sure many models will sustain the needs of a thermostat for (too) many years
  • They require additional circuitry hence more PCB space

This post is just to get your thoughts on this subject.

Some additional possible changes we may throw:

  • Tiny LEDs for each relay for easier troubleshooting (by viewing the board directly - not really visible during normal operation)
  • Double up the terminal blocks to offer both the NC and NO contacts of each relay for easier hacking. In that case we may move to terminal blocks with smaller pitch in order to fit all those wires and maybe use spring loaded for tool free changes:

Clicking doesn’t bother me one bit. But I’m luckily (unlucky in other ways) to have my thermostat placed in a room where we wouldn’t hear the clicking normally anyway. I agree that the clicking can be very useful for trouble shooting and such. I control my garage doors with them I was able to figure out I had a bad relay once because the RPi said it should be on, the light on the relay said it should be on, but the relay itself never clicked.

The more PCB space is a little concerning as the HestiaPi is already significantly larger than most commercial thermostats. I’d hate for it to need to grow even larger.

Personally, I’m not all that concerned with the SMD components. Were I to build a HestaiPi myself, I’d get the PCB from you anyway so I wouldn’t be dealing with those bits anyway. But I’m a software guy so my opinion on the matter shouldn’t count for much.

I like the idea of tool free terminal blocks.

I’m sure everybody is familiar with the sound, but just to make it clear for the rest, it is a relatively quiet sound and it usually doesn’t occur more than, say, couple of times per hour during activated heating/cooling depending on your home conditions.

SMD components are a solution for this…

Actually dropping the SSR for mechanical ones would most likely reduce the height (off the wall) of the unit. Adding more circuitry may increase the PCB a little but not the outer case, so no need to worry :slight_smile:

1 Like