Axpert standby/off mode.

[Chris Hobson]

After 5 years (yes one of those back burner projects) I am communicating with my Axpert using my own scripts. A Czech Axpert user, Josef Kriegelstein  has a pair of commands MNCHGC1497 and MNCHGC1498 to cause the Axpert to go into standby mode and revert to normal. They look more like charging current commands.

I have used them to no success, getting a NAKss response from the inverter. Anyone had any success getting a inverter to switch off. I have a pair of inverters and after 6pm there is not much sense having both running.

[Gnome]

@Coulombis the most likely guy to be able to provide an answer here because he has seen a lot of the commands from the firmware itself.

I've looked at WatchPower source code and there is nothing there to turn it off.

When flashing a new firmware the unit is powered down, not sure if that is because it is placed into flashing mode or there is an actual power down command.

Also the Axpert King I suspect may be using commands to power on/off the inverter.

[Coulomb]

4 hours ago, Chris Hobson said:

A Czech Axpert user, Josef Kriegelstein  has a pair of commands MNCHGC1497 and MNCHGC1498 to cause the Axpert to go into standby mode and revert to normal. They look more like charging current commands.

Indeed, these are standard commands with non-standard parameter values. These will be understood only by fully patched firmware, such as 73.00e for PF0.8 models, 72.20e for PF1.0 64 V models, and 74.40e for PF1.0 58.4 V models. See the AEVA PIP-5048MS topic for details and downloads. Factory firmware will NAK these parameters, as you have seen.

There is no fully patched Axpert King firmware as yet. Hopefully this year.

[Gnome]

6 minutes ago, Coulomb said:

these are standard commands with non-standard parameter values

So that begs the question, are there non-patched firmware commands for standby and running mode?

Edited September 15, 2020 by Gnome

[Coulomb]

Just now, Gnome said:

So that begs the question, are there non-patched firmware commands for standby and running mode?

No. That's why that patch exists, to fill a need / request.

[Gnome]

2 minutes ago, Coulomb said:

No. That's why that patch exists, to fill a need / request.

Ah ok
Too bad

[Energy-Jason]

8 hours ago, Chris Hobson said:

After 5 years (yes one of those back burner projects) I am communicating with my Axpert using my own scripts. A Czech Axpert user, Josef Kriegelstein  has a pair of commands MNCHGC1497 and MNCHGC1498 to cause the Axpert to go into standby mode and revert to normal. They look more like charging current commands.

I have used them to no success, getting a NAKss response from the inverter. Anyone had any success getting a inverter to switch off. I have a pair of inverters and after 6pm there is not much sense having both running.

@Chris Hobson You back ;)) !?

[wolfandy]

Really interesting, gents 

Do you know how a parallel setup with 2x Axperts would behave if the slave was turned off through this command? And then later turned on again? Would we end up up with a bunch of error messages and an unstable system? Or could this actually work to turn off 1 inverter for the night to save on battery?

[Chris Hobson]

16 hours ago, Coulomb said:

Indeed, these are standard commands with non-standard parameter values. These will be understood only by fully patched firmware, such as 73.00e for PF0.8 models, 72.20e for PF1.0 64 V models, and 74.40e for PF1.0 58.4 V models. See the AEVA PIP-5048MS topic for details and downloads. Factory firmware will NAK these parameters, as you have seen.

There is no fully patched Axpert King firmware as yet. Hopefully this year.

These are a pair of PF0.8 models running LF1  72.70 C patched software. So reflashing with your latest firmware (73.00e) will enable these commands?

Edited September 16, 2020 by Chris Hobson

[Chris Hobson]

Hmm  @Coulomb I have been looking at your documentation and am not sure whether dynamic load control is implemented by 72.70c or not. I have found my  USB to serial cable and can reflash to 73.00e but am a firm believer in "If it ain't  broke do fix it" principle.

How is the integer used  in  the MNCHGC0497/MNCHGC0498  pairing derived?

[Coulomb]

5 hours ago, Chris Hobson said:

So reflashing with your latest firmware (73.00e) will enable these commands?

Yes.

[Coulomb]

1 hour ago, Chris Hobson said:

Hmm  @Coulomb I have been looking at your documentation and am not sure whether dynamic load control is implemented by 72.70c or not.

Dynamic Load Control came out with 73.00d. So I'd say 72.70c just missed out.

Quote

I have found my  USB to serial cable and can reflash to 73.00e but am a firm believer in "If it ain't  broke do fix it" principle.

I have been running 73.00e on my (now pair of) PF0.8 machines from 2015 for years with no problems. Think of 73.00 as "72.100".

Quote

How is the integer used  in  the MNCHGC0497/MNCHGC0498  pairing derived?

The last 3 digits are abritrary. We want to stay away from legitimate parameters for these commands, which are up to 140. We are already using numbers from 500 up to represent current limits, and we are already using 499 for something else. It's that simple.

The leading "0" is of course a machine number, with 0 representing the master. If you have one slave, it will always be machine 1. I always turn off the slave when I can be bothered doing it; turning off the master may not work. I'm sure it will be in the documentation.

[Chris Hobson]

On 2020/09/15 at 6:22 PM, Energy said:

@Chris Hobson You back ;)) !?

Lol yes in between  all the madness I find I have a little development time for myself.

[Energy-Jason]

12 minutes ago, Chris Hobson said:

Lol yes in between  all the madness I find I have a little development time for myself.

Nice to have you back!

[Chris Hobson]

Ok finally I have it working 

wake.py

One could include a loop to ensure you have an ACK  but I'll leave that for the weekend.

 

Thank you so much @Coulomb

 

Edited September 16, 2020 by Chris Hobson

[wolfandy]

3 hours ago, Chris Hobson said:

Ok finally I have it working 

Would you mind providing us with feedback here on how this is working in your parallel setup please? Do you get any error messages when putting one of your inverters (I assume the slave) into standby? And does everything work as normal when you wake it up again?

On 2020/09/15 at 9:40 AM, Chris Hobson said:

I have a pair of inverters and after 6pm there is not much sense having both running

I'm in a similar situation - and being able to avoid self-consumption of 1 inverter for 10h or so during the night will save almost 0.5kWh of battery

[Coulomb]

On 2020/09/17 at 7:37 AM, wolfandy said:

being able to avoid self-consumption of 1 inverter for 10h or so during the night will save almost 0.5kWh of battery

It doesn't save all the self-consumption, most of the electronics is still running. If you let it turn off completely, it can't be woken with a command (only with switch coming on, or PV available). What it saves is the overhead of switching the DC-AC converter (the inverter in the inverter-charger). That can still be significant.

I wonder if one day we can get a machine to turn off completely until morning, get woken at sunrise, and still be able to use the DC-AC converter after sunrise.

[wolfandy]

6 hours ago, Coulomb said:

It doesn't save all the self-consumption, most of the electronics is still running.

Thanks for the detailed insights, Coulomb 

[Chris Hobson]

Hi @wolfandy it works perfectly an switches  immediately.  Since I have not integrated the script into Node red yet I am using SSH at the moment at the appropriate times. I will after a while be able to see how much I am saving.

When I flashed 73.00e I had a "jelly bowel" inducing moment where the flash failed and the inverter appeared to have been bricked. I re flashed again and the inverter came to life with myself being the only thing suffering from the ill effects of the failed flash.

@Coulomb is there a version of the firmware that has the standby features of 73.00e and the fan algorithm of 72.70c? I find the fan noise now excessive. 

[Coulomb]

4 hours ago, Chris Hobson said:

is there a version of the firmware that has the standby features of 73.00e and the fan algorithm of 72.70c? I find the fan noise now excessive. 

Oh. I didn't pay much attention to the fan noise; it doesn't bother me where my inverters are. No, no such firmware exists. I suspect that the 73.00e fan level is the minimum that you'd want (from a cooling perspective).

[Chris Hobson]

26 minutes ago, Coulomb said:

Oh. I didn't pay much attention to the fan noise; it doesn't bother me where my inverters are. No, no such firmware exists. I suspect that the 73.00e fan level is the minimum that you'd want (from a cooling perspective).

My inverters are in our family room/TV room/office next to my desk. The inverters have never bothered me before - when someone used the microwave the fans would be noticeable but other that that the PC made more noise than the inverters. Now with the 2kW baseload I have during the day the inverters are 8 degrees cooler but irritatingly noisy.

[Coulomb]

3 minutes ago, Chris Hobson said:

2kW baseload I have during the day

Wow.

Though with two inverters in parallel, that would be 1 kW each. The fans should not be running higher than the minimum 30% until 30% x 4000 = 1200 W each.

Ah, the problem will be the SCCs. Your solar charge controllers are likely 3 kW nominal, so anything above 30% x 3000 = 900 W per inverter will increase the fan speed and volume. During the day, you will likely be running a lot more solar than that.

[wolfandy]

8 hours ago, Chris Hobson said:

it works perfectly an switches  immediately

Thanks a lot for the feedback, Chris. Appreciate it

8 hours ago, Chris Hobson said:

When I flashed 73.00e I had a "jelly bowel" inducing moment where the flash failed and the inverter appeared to have been bricked. I re flashed again and the inverter came to life with myself being the only thing suffering from the ill effects of the failed flash.

I had a similar experience when I flashed my last inverter. Flashing failed and display remained dead afterwards. But thanks to Coulomb and Webers explanations on AEVA, I also managed to flash it on the 2nd attempt

[flamegrilled]

12 minutes ago, wolfandy said:

I had a similar experience when I flashed my last inverter. Flashing failed and display remained dead afterwards. But thanks to Coulomb and Webers explanations on AEVA, I also managed to flash it on the 2nd attempt

Me 3.same using moving to 73.00e.Just bought Noctua 8cm R8 from Aws.They are supposed to be quiet. I am yet to install as the wiring for the 4 pin is different. Justin Case YT : 

and wiring diagram. 

[Coulomb]

I don't know if the situation has changed, but colleague Weber had to replace stock fans in a customer's inverter due to intermittent "fan locked" errors with new fans (which  were Noctuas as well). It's a pest. There might be a simple modification to the fan checking hardware to fix this, but that means a lot more work than just swapping the fans.