Author Topic: Home automation rebuild  (Read 3646 times)

0 Members and 1 Guest are viewing this topic.

Offline Wolvenar

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 1466
  • Karma: +40/-0
  • Mr. Murphys pawn
Home automation rebuild
« on: June 14, 2014, 04:44:58 am »
I am attempting  to simplify and centralize my automation systems.
I originally patched stuff together, most times with a ton of extra power hungry devices spread out over the house/property.
Other than just inefficient, this became a nightmare to troubleshoot and maintain.


This is my attempt rebuild all of this.
My goal is to make the entire system more resilient, efficient, and much easier to work on if/when something goes wrong.
Trying to make power from alternative energy any which way I can.
Just to abuse what I make. (and run this site)

Offline Wolvenar

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 1466
  • Karma: +40/-0
  • Mr. Murphys pawn
Re: Home automation rebuild
« Reply #1 on: June 14, 2014, 04:57:23 am »
So far, I have been rewiring the house and attached garage with many more runs of romex.
My initial planning of the house when I rebuild it included easy to open channels for high voltage electrical, and a seperate low voltage, so this was part was not terribly hard, just time consuming.

My "power wall" in  the shop however was a much tougher job.

Initially I had two breaker panels per floor one each for grid and off grid.
Yeah I know, a mess. This came about for various reasons, mainly because of room and lack of the attached garage.
Now with the attached shop, I have the room to centralize everything.
Most of my time so far has been moving romex runs, and removing these subpanels.

But moving a bunch of circuits, while running web services, and working around family schedules gets complicated (and MESSY).
I sure hope the electrical inspector doesn't see this image, it will certainly be much more tidy when finished.
I have a lot of temp wiring until everything is ready to go in properly. Definitely not proud of this mess.
Trying to make power from alternative energy any which way I can.
Just to abuse what I make. (and run this site)

Offline Wolvenar

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 1466
  • Karma: +40/-0
  • Mr. Murphys pawn
Re: Home automation rebuild
« Reply #2 on: June 14, 2014, 05:21:23 am »
At the core of the automation will be two controllers from oeezee.com.

One (pictured nearly installed) has 16 gpio that normally control relays on the board below it.


These boards optically isolate the relays from the main controller board which is a nice perk.
However in this case I am trying to remove physical contact relays from the equation for (hopefully) better reliability.
I can therefore connect this directly to these (S)olid (State) (R)elays.
Each of the SSR are mounted on a heatsink, and air channels are already made up for directing air, on the top is a fan like found in a PC.
I am not sure yet that this is enough to handle the heat, but given most of the circuits are not generally loaded much when on, I think it should do.


This oeezee board as configured has a multitude of both analog and digital inputs, perfect for monitoring things.

The second board controls 32 relays, it does not have the input that the 16 does, as its repurposed the gpio to run the extra relays.
This controller will control mostly low voltage things, or relays/SSR at the device for a few things, which is an unfortunate but necessary trade off for both the space, and how some of the devices are built. It would have been nicer to have *all* high voltage switching right here.

(to be continued)
Trying to make power from alternative energy any which way I can.
Just to abuse what I make. (and run this site)

Offline Wolvenar

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 1466
  • Karma: +40/-0
  • Mr. Murphys pawn
Re: Home automation rebuild
« Reply #3 on: June 25, 2014, 08:20:25 am »
Yesterday I started hooking circuits into the SSR bank.
I figured I would start with the circuit easiest to test things with.
One of the shop ones of course.  ;D
At first I plugged a small fan in, all worked just as expected.
Then I plugged in the air compressor, Turned on the SSR, and it fires up just as it should.
As it was getting close to the shut off pressure I figured now was the time to see if the SSR turns off.
Well it didn't  >:(

These SSR fail in the "ON" state. and it had failed.
These are 40 amp 28-380 volt SSR, on a 20 amp circuit. They are supposed to have a surge capacity of something like 10x the rated.
I'm not sure if this is the expected result, or was this a dud.
I had previously in testing  a couple of these (from another supplier) to see if they would do the job for me.
I loaded the test one with not only the same air compressor but a wire feed welder doing small test welds, and an air conditioner with the intent to massively load this thing with some varied loads, until it trip the breaker or roasted the SSR.
It did well for the ~5 min it took for the *30* amp breaker to trip.
Just got the heatsink quite warm.

As far as other loads on other circuits I moved to other SSR. 
So far lights and pretty much any electronics like TV and other such sort the other SSR handle well.
I would rather not wreck a bunch of these things in the learning process, or by someone plugging in a vacuum or similar if I leave them in place.
But I would still like to use them if possible.

Is it these SSR just cannot handle high current motors and my first test was a fluke?
My understanding of larger SR is somewhat lacking I suppose.
Does anyone have any advice on these?

The datasheet on these SSR
http://www.fotek.com.hk/solid/SSR-1.htm


Trying to make power from alternative energy any which way I can.
Just to abuse what I make. (and run this site)

Offline oztules

  • Forum Advisors
  • Hero Member
  • ****
  • Posts: 1177
  • Karma: +104/-8
  • Village idiot
Re: Home automation rebuild
« Reply #4 on: June 26, 2014, 03:31:23 am »
Motors and large transformers can cause havoc with zero crossing.

At rest a loaded motor will appear as a short circuited transformer, and at rest a big transformer will appear as a big transformer... no prizes for that one.

The interesting thing is the inrush currents for transformers. If you were to start the transformer at the peak of the sine wave, then things will be ok, but if you start at the zero crossing point, the inrush currents will be huge, particularly if the saturation point of the motor or transformer is close to it's operation flux.

If you start at the peak of the wave form, the applied voltage needs to be countered by a back emf of similar value. So we get this from a changing flux, which means a rapidly changing current to make the rapidly changing flux to generate the back EMF.... This will be about the same as the normal current waveform... all goes swimmingly.... the current/flux is at zero when the voltage is at peak,,, exactly the same as steady state where the current lags the voltage by 90 degrees.. no inrush at all

But at zero crossing this does not happen, instead the applied voltage is zero, but the flux is also zero.... as the wave form leaves zero, there was no time where the flux was coming down from fully negative to provide the back emf to counter the rising voltage.... so the current will rise with the voltage, but still needs to get the flux required to get the back emf generated to counter the terminal voltage.... yes we need twice the flux in the same time to even things out.... or get normal..

In short, this means we need twice the flux in the same half cycle... ideally this would mean just twice the steady state current...... but the thing will go straight into saturation ( generally we run close to saturation to keep size down) so twice the current will not get us twice the flux.... as we are saturated, we are now trying to generate the required  flux with an air core.... and suddenly the resistance of the wire is the only real limiting factor.... do you see where this leaves us....... miles and miles over current, and it will take a few cycles to get to steady state...... the motor was already looking like a shorted transformer at rest... so it can only get worse from there.

This could explain the SCR having convulsions..... try with a bigger one for the motor circuits or big transformers... particularly torroids....because the inrush currents can be horrendous if it has to start at 0 volts from the get go... zero crossing is the worst case scenario.... peak voltage start is the best case.

I think that was what happened at zero crossing anyway.... Ross?

edit: your peak currents will be nearly the peak voltage / the resistance in the windings. big motors may get as low as an ohm or so for the run and less when the start is incorporated.... our 240vac is actually 320v peak, so peak currents will be about 300 amps for a few cycles for those motors. ( assuming no leakage)

My torroids I have wound for my pure sine inverter is only 114 turns of 7mmsq wire...for the 240v winding.... very low resistance torroid with 80 lbs weight.... blows 30 amp mains overloaders like tissue paper on a hard start with no load if you mistime the test waveform... hit and miss testing...... :-[ very low leakage.


................oztules
Flinders Island...... Australia

Offline rossw

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 866
  • Karma: +32/-0
  • Grumpy-old-Unix-Admin
Re: Home automation rebuild
« Reply #5 on: June 26, 2014, 06:04:07 pm »
I think that was what happened at zero crossing anyway.... Ross?

I can't say I've spent an aweful lot of time thinking about it Oz, but your logic seems sound, and certainly reflects my own observations over years.

The other thing - may not be so common now as it once was - not all ZCD are made equal! Some SSRs would not turn off with highly inductive (or capacitive) loads as it screwed with the zero-crossing.

Offline Wolvenar

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 1466
  • Karma: +40/-0
  • Mr. Murphys pawn
Re: Home automation rebuild
« Reply #6 on: June 26, 2014, 06:52:27 pm »
Ok I follow how that works.
So peak amperage may be way out of the scale of these SSR with a large motor since they claim to take 410 amp, only for *one* cycle however.
So they may present the high amperage for several more than just that one they can handle

But as Ross pointed out. There may be a difference in the SSR I ordered for testing, and the one I roasted.
Today tried the very same one ordered for the testing before I ordered the bulk of these in the same manner as the one I roasted..
Switched it over and over, randomly timed ( no telling where the phase was) with varying, and larger loads than what took out the other SSR and it worked just fine each time.
In fact I took out a 30 amp relay, both in a quick manner, and over a few minute time period, this applied to a second SSD from the first test batch.
I roasted yet another of the second batch with a simple large fan motor.

My guess is these were cheap knock offs of already China made devices.
Either that or they were made even cheaper on this batch. They are slightly off in color and all around build quality comparatively.
This batch was ordered from a different seller who bundled heat sinks with the SSD.

Maybe the next step is to contact the seller of each device, and see if there is either a difference in suppliers, and if the seller of the less capable batch and see what they have to say about any warranty for them since they are not up to the same quality as my test order.


Trying to make power from alternative energy any which way I can.
Just to abuse what I make. (and run this site)