I think there are still quite some bugs in the implementation of the IP43 Smart Charger:
DC current reading is wrong in PSU mode and bulk mode. Here at 29V (also tried bulk at 28.8V): reports 0.5 or 0.6A instead of 1.2A. Going through a big 24 ohm 1kW resistor:
Setting charge state to PSU mode in Node Red.. nothing happens. Only way to do it is in the GUI
Curious to hear your opinions or solutions!?
DC current going through the resistor has been measured with a Velleman DVM030 multimeter. I also checked that 29V is measured over the resistor.
I also tried with a 1kW 16 ohm resistor, which should invoke a current of 1.8A. Yet 1.2A was displayed in the GUI and Node Red. Seems like an underestimate every time.
There is only a Cerbo GX connected to the Smart Charger, with firmare 3.66.
When running 29V through a 48Ω resistor, it should give 0.6A. So 0.5/0.6A sounds very good to me.
Is your multimeter working correctly?
Running 29V through 32Ω should give 0.9A, so 1.2A seems a bit off (let’s first check your measurements)
The charger is not intended as a measuring device. The measurement circuits and calibration are designed to provide the accuracy needed for the purpose of charging.
This means that the AC current measurement is more of an indicator. There is a reason it is displayed with a single digit (0.2A means it is somewhere between 0.15A and 0.25A).
I will pass your issue of not being able to switch between PSU and charger mode from node red, to our Node-Red team. As well as the missing documentation of certain error numbers.
PS: to double check, do you mean Charger firmware v3.65?
Thanks for thinking along and reporting the node-red bugs already!
I made a typo with the resistor values in my original post (now corrected). I am using up to 3 1kW 48ohm resistors in parallel to create the different loads. I now redid the experiment with the parallel resistors and also with a 2,5ohm resistor. These are the results:
Victron IP43 Charger at 29V output:
1 resistor of 48ohm (measured with multimeter): 0A in Victron
2 resistors of 48ohm in parallel: 24ohm (calculated) =1.1198A measured with the multimeter but 0.6A shown in Victron Connect app
3 resistors of 48ohm in parallel: 16,5ohm (measured with multimeter)=
1.8A measured with the multimeter but 1.2A shown in Victron Connect app
Victron IP43 Charger at 24V output (multimeter can measure max 10A):
1 resistor of 2,5ohm: 9A measured with the multimeter but 8.4A shown in the Victron Connect app
(Probably 0,15ohm to be added for connections and wires? 25V/2,65ohm=9A)
So 0,6A difference each time. Is there maybe an offset applied in the firmware to compensate for the internal power draw of the Charger?
Firmware of the Cerbo is version 3.66 (the Cerbo and Charger are bought only 1 month ago):
There is also no setting in the Victron IP43 Smart Charger Control node to set the Max DC Current. However that setting is there in the Victron Connect app:
Setting the Max Charge current so that it limits the current actively: normally my 2,65ohm load would consume 9A at 24V. When the Max Charge Current is set to 6.2 at 24V, the voltage drops to 16V and the current to 6.2 (Victron always offsets by 0,6A for some reason, see above posts). Setting the Max Charge Current higher makes the voltage revert to 24V, but this works only once! After that the charger hangs forever at 16V..
This also happens with no other Victron apparatus connected to the Charger. Connecting the Cerbo makes no difference in this behaviour.
When you limit the charger’s output current to 6.2A, the voltage drops to 16V, (internally) causing an undervoltage condition. For safety reasons, the charger will limit the output current to the lowest possible value (6.2A).
Even though you set the current limit back to 25A, the undervoltage still remains, so the charger will not increase the output current until the undervoltage is resolved.
This behaviour is as intended.
To work around this issue, you can either switch the load off and on (if possible), or switch the charger’s output off and on.
I agree with it doing undervoltage but I don’t agree with the inconsistent behaviour it shows. If the limit is set back to 25A there is no more reason to decrease the voltage and it should return to normal (>=24V).
This good behaviour is even what happens the first time the limit is increased again (like I described in my post). The second time it does not work anymore.. which is inconsistent.
In my test setup, the charger consistently remains at 16V, 9A after setting the current limit back to 25A (also the first time). I suspect you are seeing different behaviour the first time, due to timing (before editing you said “this works only a few times“). For how long did you leave the output limited to 6.2A?
Btw, the charger is not lowering the (target) voltage, it is only lowering the current limit.
Thanks for testing along, I appreciate it. Let’s sort this out.
We investigated further and indeed it is a timing issue! Going through the menus fast allows to keep on switching low and high current for many times!
So we retried again with the same set-up and also managed to reproduce that it didn’t came back to 24V/25A (see video below). I waited another 7 minutes after the video and the current and voltage did not go back up again. You can see I go slowly through the settings.
Are you at the same firmware? And are all IP43 Smart Chargers the same in terms of hardware or have changes been made between batches?
In either way, we are a company and did not buy Victron to loose so much time to this. We hope the issue can be resolved the coming days.
In the firmware changelog I found that there have been very similar issues before, with the 24/16 IP43 Smart Charger:
The undervoltage condition has a 10 second delay; When the voltage drops below the undervoltage threshold (22V your case) for 10 seconds, the undervoltage condition gets activated. When the voltage rises above the threshold, the undervoltage time counts backwards (so intermittent undervoltage can still lead to activation of the undervoltage condition ).
The undervoltage condition serves to protect the internal components of the charger from overheating. So the firmware is doing exactly what it is supposed to do.
For what it is worth, I have checked undervoltage condition in v3.65, v3.61 and v3.50 and also the regression in v3.31 has nothing to do with this feature.
Thank you for describing the timing behaviour. It would be good to put this in the manual.
However, I agree that it is allowed to go under 24V but after that it never goes back to 24V.. that’s the hole problem. The backwards counting that you are describing clearly does not work in my unit as I demonstrated in the video. There the undervoltage stayed 17 seconds, after that I waited for 7 minutes (!). So whatever duration the timer is counting, .. it is not working to allow for the voltage to rise again.
That your charger does not return to 24V is as expected.
First you lower the current limit. Then you cause undervoltage by connecting a load that is too large.
After 10 seconds this triggers the undervoltage protection. The undervoltage protection limits the current to 6.2A.
Next when you increase the current limit back to 25A, the undervoltage protection’s current limit (of 6.2A) remains active (to protect the charger).
Meaning that the output voltage cannot reach the 22V threshold. So the undervoltage condition remains (and so does the undervoltage protection).
To get out of this, you can either disconnect the load, or briefly turn off the output of the charger.
I don’t agree that a load of 2.6ohm is ‘too large‘, as empty batteries go easily below 1 ohm. This seems to imply the Charge Current limit will not work at all? Or how will the Charger react on that? (I don’t have batteries at hand)
In numbers, this means the load may not be bigger than 24V/6.2A = 3.87ohm or you’re stuck with permanent undervoltage. (Disconnecting the load is not an option for the use case I intend.) At 24V/25A = 0.96ohm it is looking better but as said, empty batteries go even below this.
When the battery voltage is below 22V, then it is over-discharged. In that case the battery will be charged slowly, until the battery voltage has recovered sufficiently. Exactly how it should be.
The load should be no bigger than 22V/6.2A=3.55Ω IF you set the output current limit to 6.2A.
Not all possible errors are included in the Node Red , there are at this moment over 100 possible errors defined. Non-zero means something is wrong.
That being said, I do agree that the errors 32 and 33 should be documented, while for example error 67 does not apply for the IP43.
The Victron Node Red nodes, are intended for monitoring and control, not for configuration. This means changing the charger’s function (charger or power supply) is not supported from Node Red.
Configuration must be done using Victron Connect (via bluetooth, VE.Direct or VRM).
Also be aware that the ChargeCurrentLimit setting is a setting, not a control parameter. If you try to use it as a control parameter, then you will eventually wear out the internal flash storage in the charger.
Ok, that clears some things up. The battery brings voltage of its own, but if the internal voltage of the Charger is lower than the battery’s, no current will flow? For PSU mode I think it would be good to see the info about different loads and the internal protection and its timer make it to the official documentation.
How can a user differentiate between a setting and a control parameter? Because indeed we were planning to write to it every few seconds. All we have here is which setting is writable/readable/both at https://github.com/victronenergy/node-red-contrib-victron/wiki/Available-nodes. So control parameters live in RAM?
The charger will try to generate a voltage, but the voltage can be limited by the current.
This is why the battery voltage slowly climbs during the bulk charging phase. The charger is set to generate the absorption voltage, but the current is limited to whatever current-limit is set.
Basically any parameter that persists after a power cycle (of the IP43) is stored in flash. Overwriting those parameters every few seconds will eventually wear out the flash memory.
I am getting the feeling that you may be complicating things more than they could be. Maybe it is good idea to ask for some advice on how to approach your use case here on the community, or to talk to your Victron distributor.
