Can you please help troubleshoot the issue with this installation (see enclosed diagram)?
I am not an electrician or an engineer and my diagram is very simplified to give an idea of the installation (negative wires and fuse omitted). There is a fuse between the lithium batteries and the installation, and another one between the AGM batteries and the installation.
The goals of the installation are:
When the alternator is off, charge the AGM batteries from the lithium batteries through a Orion-Tr Smart charger.
When the alternator is on, charge the Lithium batteries through 2 Orion-Tr Smart chargers.
When the alternator is on, charge the AGM batteries directly from the alternator (without chargers in-between).
When the alternator is off, it is working correctly (AGM batteries are charged from the Lithium batteries through the Orion).
When the alternator is on, I observed short voltage spikes (of up to 40V for a few seconds) on the input of 2 of the Orion chargers and the output of the other Orion charger. I did not have time to measure it elsewhere, but it suggests that the line from the alternator to the AGM batteries was having those spikes.
Upon seeing these spikes, I turned off all 3 Orion chargers from Victron Connect to avoid damaging the batteries. When the chargers are off, the alternator output a stable 28V (27.95V measured) to the AGM batteries without issues.
What can cause those voltage spikes and how can I fix it?
I think your problem is the input to the 2 Orion chargers feeding the lithium batteries is isolated from the AGM battery by the Argofet. The Orion chargers need the AGM battery to buffer (smooth) the voltage response. Look at all the Victron diagrams for Orions, the alternator connects to the lead acid battery and direct to the input of the Orions.
Try the system without the Argofet and the lithium to AGM Orion turned off. I am sure the alternator will then charge both batteries.
The on thing you need to be careful about is setting up the lithium to AGM Orion so it only charges when the alternator is off, you can do this by setting the output voltage of the lithium to AGM Orion to a lower voltage than the alternator output. If the battery is higher voltage than the Orion setting then it will not charge.
It makes sense. However, if the end goal is to charge the AGM even when the alternator is off, then I cannot remove the Agrofet permanently. If I do, the output of the Lithium->AGM charger will be connected to the input of the Alternator->Lithium chargers (via the AGM battery).
It will create a circular charging pattern where the Lithium->AGM charger raises the voltage of the AGM battery, which is connected to the alternator and the input of the Alternator->Lithium chargers. This raise voltage will activate the Alternator->Lithium chargers.
What I could do is remove the Agrofet, and connect the Alternator->Lithium charger remote sense H-pin to the D+ output of the alternator. In that case, these 2 chargers would only be on when the alternator is also on, and completely be turned off when the alternator is off, even if the input has a high voltage due to the AGM charger. Right?
There are two main differences between DIODE charge splitters, and FET Charge splitters (like the argofet).
The most commonly known difference is the voltage drop associated with diode splitters is eliminated in FET splitters.
What is not realise is that with FET slpitters, the current can flow EITHER way through a FET when it is turned on - depending on the voltage difference.
Thus it would be possible under the right circumstance, for power to flow from the lithium battery, through the dc/dc that charges the AGM, backwards through the ArgoFET, then through the dc/dc converter that charges the lithiums, thus completing the loop.
To stabilise your alternator, the AGM should be connected directly to the Alternator output. The Orions can then charge the lithium, being switched by the D+ output from the alternator. If you then decide that you need a float charge for the AGM battery, then that can be added, but this should only be sourced from the house bank if you have another source of charge for the house bank i.e. wind or solar.
I believe I have narrowed down the issue that causes the voltage spikes. I have not tested this reasoning, but it is the only explanation that makes sense to me. Feel free to disagree or point out errors in my reasoning; I would appreciate it.
This setup is likely to cause voltage spikes at the alternator and the batteries. An Agrofet should not be used with an Orion charger when the damping battery is behind an Agrofet output. Here’s why:
Remember that the Agrofet will close (allow current to flow) when the voltage at the input is slightly higher than the voltage at an output. The sequence that causes spikes is as follows (example in a 24V system):
The alternator is turned on and generates a steady 28V. This is higher than the output voltage at both the battery and the Orion. The Agrofet allows current to flow to both outputs.
The battery voltage rises to 28V as expected.
After a while—depending on your Orion configuration—the Orion starts charging. This causes an immediate drop in voltage that can be measured at the alternator.
The battery is slower to lower its voltage (it’s a battery!), so for a brief moment the output side of the Agrofet (to the battery) has a higher voltage than the input of the Agrofet. The Agrofet opens this circuit and disconnects the battery completely.
Now the damping battery is no longer connected to the running alternator. This causes the voltage to fluctuate. The Orion, if it is cycling on and off, compounds the problem (for example, due to voltage lock).
When the alternator voltage rises above the battery voltage (say it spikes to 30V), the Agrofet allows current to flow to the battery. This raises the battery voltage, stabilizes the alternator at 28V, and then immediately disconnects the battery again (since the battery is now sitting around 30V, higher than the alternator’s stable voltage).
This cycle—lasting only a few milliseconds—repeats over and over. In my case, I observed brief voltages as high as 40V.
Now I will check if the batteries have been damaged (hopefully not). I will change the setup so that the damping battery is not behind the output of the Agrofet, or remove the Agrofet completely. TBD.
