3 hours ago, SilverNodashi said:

Sadly many sparkies won't touch solar systems. AC and DC is completely different and most of them don't know how to work with DC systems. 

The SANS regulations are clear on how the wiring should be done though, but ends on the AC side. And then it has some regulations in terms of safety for the batteries and DC connections, but it doesn't tell them how to work on DC based equipment. 

Luckily My Sparky has been involved in a couple, and also this being a small town he's realised if he can build the skills he can create business for himself as the go to guy in town.

 

G

For me I went with a few Sonoff PowR2 on certain devicess that I can monitor. I flashed these with Tasmota and then upload it through mqtt to Home Assistant. From Home Assistant into influx DB and then graph on Grafana.

For monitoring the combined  I used the pzem-004t connected to a nodeMCU flashed with tasmota that send the data to Home Assistant again. This works out to around R140 for the pzem-004t and include a display that I place next to the DB. And R70 for the nodeMCU. 

Below is a picture of the pzem-004t. It accept AC power input to measure realtime voltage and do the power usage. Bit more DIY as some products here but work for me.

Below is a extract, the geyser timer is set to start at 18:30 till 20:45 each day. We also have a municipality geyser control installed. From the graph you can see that the geyser 1st goes on around 20:20 till around 20:45 even though it is set to start around 18:30.

The 18:40 part is where we start using the stove.

 

... pretty amazing actually how many options we have to monitor everything...

it's mostly picking which we want to use and to what detail.

G

Eh. Erm, please help the n00b out here please.

So I've installed jerry rigged rednecked the ET112 into the DB board now to test. Connected a RS232 <-> RS485 converter purchased from Communica to a PL2303 USB <-> Serial converter to my RPi3, with the venus GX image loaded onto it. I can access the web UI to the Pi on the local network, but it seems that I am not getting comms to the ET112. I presume the default comms to the serial interface would look for tty0, and not ttyusb0 as it stands now. Scratched around in the web UI on the PC, but there is no place to select the com port to use to talk to the USB<->Serial converter. My 10mins worth of googling did not bring up anything that makes sense to me.

Anyone have an idea of what I can do?

Pics for posterity...

 

Just thinking, I saw in the manual of the ET112 that if you have multiple devices on a single line, one should bridge pins 3 & 5. I did not bridge them, is that my issue perhaps?

1 minute ago, Minerva10210 said:

but it seems that I am not getting comms to the ET112

Right. First off, there is a big ugly old bash script called serial-starter. This works together with udev (a kernel driven device handling layer used in most linux distros), and you really don't have to know anything about how all this works except that when a new hardware device is detected, it will try a few possible "drivers" one by one until one of them works. It is called "drivers" but of course it's not really a driver in the OS sense, it's a bit of software that can talk to that device.

Now, the options that are tried depends on what kind of thing you plugged in. You plugged in an RS232-USB converter. Venus will never try the drivers for rs485 on one of these. In other words, this is never going to work. You need a real RS485 cable.

You can however hack it. If you look in /etc/venus you will find a file called serial-starter.conf. By simply adding "cgwacs" to the default list (that's the software that reads the Carlo Gavazzi Wired AC sensors, or cgwacs for short), you can make it try the usb-serial device. This may work or may not work. Also, I cannot say if it might badly interact with other usb-serial things, eg vedirect cables.

The other less invasive way, is to first tell serial-starter to leave that port alone. Eg:

/opt/victronenergy/serial-starter/stop-tty.sh /dev/ttyUSB0

After you've done this you can manually start cgwacs:

/opt/victronenergy/dbus-cgwacs/dbus-cgwacs /dev/ttyUSB0

If your meter shows up after running this command... well... success for now.

11 minutes ago, Minerva10210 said:

bridge pins 3 & 5

That enables the built-in termination resistor. The speed used by cgwacs is 9600 baud. It's too slow to be affected by termination or the lack thereof. You only need to do this on really really long cable runs. Also, when you use a proper usb-rs485 cable (the ones Victron ship for example), the other end is already terminated.

27 minutes ago, plonkster said:

Right. First off, there is a big ugly old bash script called serial-starter. This works together with udev (a kernel driven device handling layer used in most linux distros), and you really don't have to know anything about how all this works except that when a new hardware device is detected, it will try a few possible "drivers" one by one until one of them works. It is called "drivers" but of course it's not really a driver in the OS sense, it's a bit of software that can talk to that device.

Now, the options that are tried depends on what kind of thing you plugged in. You plugged in an RS232-USB converter. Venus will never try the drivers for rs485 on one of these. In other words, this is never going to work. You need a real RS485 cable.

You can however hack it. If you look in /etc/venus you will find a file called serial-starter.conf. By simply adding "cgwacs" to the default list (that's the software that reads the Carlo Gavazzi Wired AC sensors, or cgwacs for short), you can make it try the usb-serial device. This may work or may not work. Also, I cannot say if it might badly interact with other usb-serial things, eg vedirect cables.

The other less invasive way, is to first tell serial-starter to leave that port alone. Eg:

/opt/victronenergy/serial-starter/stop-tty.sh /dev/ttyUSB0

After you've done this you can manually start cgwacs:

/opt/victronenergy/dbus-cgwacs/dbus-cgwacs /dev/ttyUSB0

If your meter shows up after running this command... well... success for now.

Seems like a "no"

I do have some more USB <-> Serial converters here, will test another just to be sure.

 

Alternatively, would it possibly be a better idea connecting directly to the GPIO pins on the Pi?

Aaaai. Fixed the issue. had the tx and rx pins reversed on the RS485 converter apparently 😐

So this is what the results are for the last few days with the ET112 running. The big 6kw peak I know was semi-on purpose. I had the kettle, tumble dryer, dishwasher and something else running at the time. The rest is for the most part "normal" usage. Obviously not really 100% into the "work efficiently" mindset yet, so working on that. Also, other than a few light globes around the house, probably using a lot of energy hungry devices still.

 

What you guys think?

 

That looks pretty normal(ish). 20 April must have been washing day with a constant load over 1kw most of the day, but the rest seems pretty similar to mine.

5 minutes ago, plonkster said:

That looks pretty normal(ish). 20 April must have been washing day with a constant load over 1kw most of the day, but the rest seems pretty similar to mine.

Yeah, mowing the lawn actually

 

Thanks for all the assistance so far guys, really liking the whole journey here. Now that ship just needs to make a move on so that I can start applying the learnings.