Hi all, I’m new to Victron, but experienced in electronics.
Problem is that I’m seeing voltage > 14.7 to 15.0 volts on the battery side when clouds pass. Everything is normal during full sun.
My batteries, inverter and other 12VDC devices required < 14.5V
Setup is:
SmartSolar Charger MPPT 100/30 firmware v1.74, bluetooth v2.53
SmartShunt 500 firmware v4.19
400W solar (2 x 120W, 2x80W) solar panels I’ve tried wiring as 4 parallel (~17-20V) or as 2S2P (for >34V)
4 x 105Ah 12V AGM batteries, that want 13.6V float, and 14.3V absorption.
6AWG from MPPT
Bluetooth networking between the SS and MPPT.
Temperature compensation is OFF right now.
I have checked settings, and can’t figure out why it’s spiking like this, wondering if there is a problem with the MPPT.
I tried disabling the VE.networking to take the current, temperature, and voltage sensing of the SS out of the picture.
When things are sunny, normally I see about 14.3V and 13.6V. Adding load, no spikes.
If clouds come along, sometimes it starts with these spikes, and the inverter and protection for 12v devices complain and shut down.
If I don’t snag a screen shot, the spike history is not present in SolarHistory.csv as it’s too summarized.
The previous solar controller worked inefficiently, and I successfully used the SS to monitor battery.
It doesn’t explain why the charger output voltage shoots beyond the set maximum voltage for the batteries. I’m expecting it would be clamping it to never exceed the setup voltages.
Shadows on the modul are genarating additionally power peaks. So the working point will be near the lower voltage peak. If the batterie is full the working point moves a little to higer voltage to reduce the output power. When the cloud disappears the power can not reduced by moving to highter voltage because there is the most power peak. So the output voltage will be a short time too high
So how do I protect all of my equipment from these voltage spikes?
OGPS
(Ed @ Off-Grid Power Systems - offgridps.com)
6
I haven’t seen this behavior before on the output side of the MPPT. Or, it’s been so quick I’ve never noticed in the last five plus years of installing these. I agree with you. While the voltage on the input side can vary widely and quickly, it should be stable on the output side if the batteries remain connected. I have seen this behavior when no batteries are connected to buffer the DC buss.
Are you measuring this output voltage with or without the batteries connected? How old are the batteries and are they in good condition?
normaly there is an inverter connectet. Inside of the inverter are capacitors. If there is no inverter or the capacitors are old the overvoltages can appear if the battery is full charged. You can connect an electrolytic capacitor of abeout 30000microfarad to test it. If you have an inverter you shoult test the internal capacitors: fist connect the inverter to power but switchde off. Disconnect the inverter an conent a multimeter to it. It will show the battery voltage. Than connenct a Resistor an observe the time of the falling voltage. Take a resistor so that te voltage will reach 0 volts after a time of about 10 sekonds. Now you can calkulate the internal capacitor
Output voltage measured with batteries connected - SmartShunt 500 is the data source. These higher voltages aren’t necessarily in the viewable data since it gets averaged. I wish there was a way to download the raw data from the SmartShunt - there doesn’t seem to be an export capability.
Batteries are not new, and I swapped out a weak one when I was suspicious. My plan is to switch to LiFePO4 soon, but these AGMs have been excellent.
Temperatures here have been swinging wildly, and that makes it hard to get a good test of capacity. Plan is to tackle that this weekend.
So a theory, they probably have a high internal resistance.
So a short charge raises the voltage quickly but as soon as the charge amps stop the voltage falls. Its falling enough to trigger the rebulk so it tries again, then the cycle continues.
Each battery tests about 3.0-3.3 milliOhm, so it could be a bit better, but that isn’t very high internal resistance.
Battery age is likely the bigger problem; I’m going to load test them tomorrow if I get some time.
You have selected 14,3V absorbtion. But the battery voltage rises to about 14,8V. That matches good if the battery temparature is about -5°C. Without temperature sensor the MPPT uses the temperature from early mornig when the sun comes out. If the Battery temperature is rising during charge, the voltage will to be too high. With a temperature sensor you can additional reduce the absorbtion voltage to 14,2V. Here is a graph of a full charges battery stored outside in a wooden housing:
In this case the voltage in the late afternoon will be abaout 81mV too high. But the thermal capacity of a MPPT is lower a the thermal capacity of the battery. So the detected mornig temperature of the MPPT may be significant lower. If the Battery in your case would be similar to my graph but the MPP temperature in the morning would be -5°C the battery voltage would be 243mV too high
it has only a little effect to the PV voltage but no effect to battery voltage. I think overvoltge can occur when the working point ist located left of a Powerpoint. If the working point is located between two powerpoints it will be always left of a power point! Normaly it will be positioned right of a power Point and the output power is regulated at the flage of the falling edge. Moves the working point to left, the power increases. moves it right, output power decreases. Is the working point left of a power point, the direction is reveresed. So the power will be increased instead to decrease it if the most power point has recovered meanwhile.
I do have a temperature sensor attached to the battery, and it is connected to the SmartShunt.
Temperature compensation (TC) was turned OFF so I could simplify the problem.
With TC Off, the voltage should be 14.3V, regardless of battery and MPPT ambient temperature. Right?
My plan is to turn TC On, after I figure out why it is exceeding voltages.
Voltage in my data showed 400mV to 700mV too high. (14.3 + 700 = 15.0, max is 14.5 in my system).
Here’s another graph from around the same time, showing spikes >14.85 Volts, in mid afternoon, ambient temperature +18C, TC OFF. (Sunrise T=+7C)
OGPS
(Ed @ Off-Grid Power Systems - offgridps.com)
14
I’m back from a long weekend trip, but I don’t have anything else I can contribute. I don’t have a lot of experience with lead-acid chemistries having started in this world just six years ago and working with lithium batteries. In my shop I have a 3kW programmable DC load and we could put this on the output of your MPPT and see if these issues persist, in which case we’d replace the MPPT if. Have you talked with your dealer about this? You did purchase from a local Victron dealer and not online, right?
Did you make a “Ve.Smart Networking” Network with Smartsolar and Smartshunt? Than the temperature should be displayed in Victron Connenct in the Smartsolar.
Try to aktivate Temperature compensation and lower the absorbtion voltage. I think 14V could be enough. (A batteriy will be fully charged with minimal 13,8V but it needs more time)
Where ist the Mpp wired to? To the Battery or to the inverter? Do you have a schematic including fuses wire legth and wire cross section? Or a picture?
Yes and Yes - he is happy to keep selling me more stuff and offered some good suggestions - which have eliminated several possibilities. I’ll try a full system load after I reassemble everything again.
such switches are often in a poor qualiy. Better is a 30A Midi fuse. In a system with a smartshunt kommunicating to the mpp voltage losses of the charging wire are no problem because they are compensated. You can use the charging wire as a resistor of an low pass filter. The corresponding capacitor is located in the inverter. Take a wire as thin as possible. For 30A is 10AWG enough. If the legth is 5m it is like a resistor of 0,02Ω. A current spike of 30A produces a voltage loss of 0,6V. So the spike will be surpressed by the capacitor.
and offered some good suggestions - which have eliminated several possibilities. I’ll try a full system load after I reassemble everything again.