Hardware layout For new Pi installation from meeting 1/8/17

[hamishmb]

Hi,

As we discussed in our meeting about installing a mini test-system and another Raspberry Pi for the water butts, I'm making two new posts, one about the hardware layout I shall be writing the software for, and one about the software design. This is the post about the hardware layout.

As always, if I get something wrong, please correct me

So, what I'm envisioning for the hardware layout is:

• We will have 2 raspberry pi systems, and:
  • One of them is a new pi to be installed by the sump.
  • This pi will be responsible for:
    • Taking measurements from the existing resistance probe in the sump.
    • Turn the relay on and off for controlling the auxiliary motor to pump water into the butts.
    • Providing and managing communications with the secondary pi.
    • Load balancing the sump with the water butts to try to keep the whole system in balance and avoid wasting water, using some (at least at first) simple decision-making logic.
  • A new Pi will be installed at/near the water butts behind the wendyhouse.
  • The second pi will be moved from the sump, to the water butts, and is responsible for:
    • Using a float switch (will later be another resistance probe) to measure the levels on the water butts.
    • Saving these readings to a file on its SD card (for debugging).
    • Sending each reading over the network to the master pi by the sump.
• Communications will be performed using an ethernet cable running through the pipes from the water butts to the sump.
• I will test the ethernet cable again with my tester just to be pedantic.
• *** In this installation we will _NEED_ a better way of water proofing than we have done by the sump pi. *** This is because:
  • There is only a minimal amount of protection under the water butts (if we decide to install the pi there).
  • It has as of yet been impossible to be absolutely certain the water butts have no leaks, indeed they appear to have 1 or 2 very, very, very slow leaks, but it could also be dew, so it cannot be guaranteed that a Pi will be safe underneath the butts.
  • Installing a plastic box on the side of the wendyhouse might be a good idea as Penri suggests.
• The buttspi will be powered via PoE and a DC-DC converter. It will also have an interface board as described by Terry below.

Okay, so this is a long list of things, and I've probably made some mistakes, forgotten some things, and there are probably also more things that we need to consider, so please post below so we can try and make sure we agree on all this and we aren't forgetting anything important

Hamish

[Penri]

Hamish

For the Sump Pi:
If you mean by "controlling the auxiliary motor" "turn the Solid State Relay On and Off" then that's perfect.

For the Butt Pi:
In the short term there will only be a simple "Butts Full" signal, provided by a switch of some kind, the details may emerge tomorrow, when I get something working. A level probe will follow but it may be some time so level logging will not be required yet, but it would be good to know if the Butts Full signal is triggered.

I did do a continuity check on the ethernet cable after crimping on the connectors but I'm more than happy for you to try it with your tester. Better to be sure we have a cable we can trust rather than have to debug a dodgy connection later.

Suggest we do not put the Butt Pi under the butts, I would like to see it in the right sort of box screwed onto the Wendy house, near the standing pipes, at a comfortable working level. The bushes in that area could do with a bit of a trim so we can access it without being attacked!


Hwyl

Penri

[hamishmb]

Okay, I have updated the list I will bring my cable tester just so I can be pedantic

[TerryJC]

So, what I'm envisioning for the hardware layout is:

• We will have 2 raspberry pi systems, and:
  • One of them is the existing pi installed by the sump.
  • A new Pi will be installed at/near the water butts behind the wendyhouse.

Some minor corrections:

1. The existing Pi installation will remain pretty much as it is and be moved to the butts. This will be re-christened buttspi and the hardware will be enhanced by the addition of a DC-DC Converter for providing power for the Pi (and eventually to the Buffer / Translators for the Level Probe). This 5 V supply will will be derived from the voltage picked off at the Splitter end of the PoE pair.
2. At the sump end there will be the new Pi installation, christened sumppi and this will include the new Interfacing Board which will carry:
   1. The Buffer / Translators for the Level Probe.
   2. The Real Time Clock pinouts (for when we decide to use it).
   3. The 3.3 V to 5 V Level Shifter for driving the SSR.
   4. A DC-DC Converter for power to the Pi and probe, etc.

In addition to the new Interfacing Board there will be the Injector end of the PoE pair; this will take the 19V supply from the DC Power Brick.

I am making provision on the new Interfacing Board for an additional 3.3 V to 5 V Level Shifter for driving the PWM into the Motor Drive Board, but we won't need that yet.

[hamishmb]

Okay, I have added some bits from that into the above list.

[Penri]

Hello

Plastic enclosures for electronics ordered, one off of each:

200*155*100
300*250*130
255*250*120

Hwyl

Penri

[hamishmb]

Excellent

[Penri]

Hello

Just a note for reference, I was at WMT and caught the tail end of a meeting between Stuart and the electrician (Rev) who's got the job of trying to clean up and document WMT's mains electrical system. He has made a recommendation that as far as possible, particularly outside in the model town itself, that the use of mains is discontinued and that instead we use 12V DC sourced from a battery that would be charged from solar panel. To my mind that is not a complete solution but it does point in the right direction.
The electrician's reasoning is based on an understanding of some onerous new regulations that are about to come down the tracks in relation to electrical installations in public spaces, a 12V system would fall outside the reach of the regulations
I can't see WMT changing overnight but it does vindicate the decision we took in regard to our requirements.
I need a better understanding of these regulations and make sure any implementation is sensible.

Hwyl

Penri

[TerryJC]

He has made a recommendation that as far as possible, particularly outside in the model town itself, that the use of mains is discontinued and that instead we use 12V DC sourced from a battery that would be charged from solar panel. To my mind that is not a complete solution but it does point in the right direction.
The electrician's reasoning is based on an understanding of some onerous new regulations that are about to come down the tracks in relation to electrical installations in public spaces, a 12V system would fall outside the reach of the regulations

I agree that we would need a better understanding of the new Regs, but the first point that I would make is that it would be better if the WMT was run from a 24 V battery instead of the 12 V that he has suggested. We could then use DC-DC converters to get power for the Pis and the voltage for the main pump motor would already be correct.

Thinking outside the small domain of the Pis (various) we could also convert all the lighting to LEDs and connect them in series / parallel to get more or less light where needed.

[Penri]

Hello Terry

I didn't want to get into a big discussion with Rev on the details of his suggestion because he has, as yet, only had a limited view of what the regs. say plus he was reporting that the industry talk is centred around 12V. You and I have a better understanding of the advantages of 24V (or perhaps higher DC voltage) hence my comment on needing a better understanding of what's actually being proposed in the regs.

By the way it was lighting the model buildings that started the conversation on how and what electrical power to route around.


Hwyl

Penri