Victron MPPT RS 450/200 current overshoot

[kta22]

Hi, I'm having this Victron configuration : 2x Multiplus II (5000VA) single phase ( parallel) + mppt
rs450/200 + cerbo gx + Lifepo4 560Ah + smartshunt which is working good. I've activated DVCC function in Cerbo GX, with a low 26A battery charging current, but when I'm having a transition from inverter high load to low load (e.g. from 7.2kW to ~0.2KW), the charging current overshoots up to maximum value I've set into mppt (150Ah) and not stays at 26A. Is like DVCC function is shunted for couple of seconds and current is coming down to 26A slowly ( 150A -> 26A in ~5sec) afterwards. Are you aware about this issue ?

 

[Dooug]

Sounds like at high Inverter loads the MPPT is doings it's job and provide as much power as it can to your inverter load. When the inverter load drops rapidly it takes time for the cerbo/MPPT to react and adjust to your battery charge settings.

150amosnshouldnt hurt 560AH of batteries for a short bit.

It might even just be a delay in the GX display. Would be good to see it graphed to see how the current drops

 

[kta22]

I'll try once more to connect/disconnect high load under float to observe the behaviour.

 

[timselectric]

It's functioning as expected.
A 7kw change takes time to adjust to.

 

[kta22]

It's functioning as expected.
A 7kw change takes time to adjust to.

Can be somehow adjusted/solved ?

 

[timselectric]

Can be somehow adjusted/solved ?

Only if you can limit the size of the load change.
But it doesn't need to be solved. It's operating as expected.

 

[kta22]

Only if you can limit the size of the load change.
But it doesn't need to be solved. It's operating as expected.

Thanks.

 

[sunsurfer]

Hi, I'm having this Victron configuration : 2x Multiplus II (5000VA) single phase ( parallel) + mppt
rs450/200 + cerbo gx + Lifepo4 560Ah + smartshunt which is working good. I've activated DVCC function in Cerbo GX, with a low 26A battery charging current, but when I'm having a transition from inverter high load to low load (e.g. from 7.2kW to ~0.2KW), the charging current overshoots up to maximum value I've set into mppt (150Ah) and not stays at 26A. Is like DVCC function is shunted for couple of seconds and current is coming down to 26A slowly ( 150A -> 26A in ~5sec) afterwards. Are you aware about this issue ?

I call it charge momentum. I imagine it would be hard(pricy) to design your typical SCC to be fast and efficient while also avoiding surges with quick stops to the power stream. Its like moving boxes on a conveyor belt, if suddenly, the unloading of the end of the belt slows quickly, the belt has to slow down to adjust to the new lower unloading speed. There is a slight delay in the change of the new slower speed from all the momentum. You could design a SCC that could be completely smooth but it would require a very large cap bank or something like a micro LTO bank.

 

[kta22]

I call it charge momentum. I imagine it would be hard(pricy) to design your typical SCC to be fast and efficient while also avoiding surges with quick stops to the power stream. Its like moving boxes on a conveyor belt, if suddenly, the unloading of the end of the belt slows quickly, the belt has to slow down to adjust to the new lower unloading speed. There is a slight delay in the change of the new slower speed from all the momentum. You could design a SCC that could be completely smooth but it would require a very large cap bank or something like a micro LTO bank.

It's a bit strange because I previously had 2 Growatt's(SPF 5000es) in parallel and I didn't notice any kind of voltage or current overshooting in terms of battery charging. May be due to different software control and may be due to high frequency inverters.

 

[1201]

• For CAN-bus connected batteries, check the relevant page on the Battery Compatibility manual to see if enabling DVCC has been tested with your battery type and is supported. If DVCC is not mentioned in notes relating to your battery, do not enable DVCC.

 

[GXMnow]

It's a bit strange because I previously had 2 Growatt's(SPF 5000es) in parallel and I didn't notice any kind of voltage or current overshooting in terms of battery charging. May be due to different software control and may be due to high frequency inverters.

The one box solutions do have a small advantage in that a single processor is seeing the currents and controlling things. In your Victron system as well as my Schneider/Victron/Enphase mix up, you have completely separate devices each doing a job. And a controller watching the power flow and making adjustments. My system only makes an adjustment every 5 seconds. I would bet the Victron is faster as it is all one brand and it directly knows all the data. When the batteries are near full, and the system is in DVCC or Float mode, the charge controller is running in a curtailed power mode. When a large load turns on, I bet you see the opposite spike. The batteries will need to supply the current first. Then the Cerbo will command the charge controller to crank up it's current output to cover the load and also supply your desired charge current. Then when the load turns off, the opposite happens. As you have seen, the charge controller is still putting out the commanded current to cover the load and the charging current. But without the load, ALL of the current becomes charge current. It takes some time for the Cerbo to see this current flow. Then it needs to calculate the new values, and then send the new current command to the charge controller. And then the charge controller probably also has a ramp rate and it has to adjust to match the new command. A few seconds is not a problem.

I intentionally made my system respond slower so that it would not over shoot and oscillate between charging and discharging. When my system sees a large change in load current, it will only adjust the charge current or inverter output by 80% of the error. In 5 seconds, it makes another 80% charge of the remaining error. On the 3rd 5 second loop, the current is usually in range. So it now takes my system about 15 seconds to cover a large load change. As long as the resulting current swings are well within the power ratings of the batteries and devices, don't worry about it. In my case, it results in a bit of power flowing in and out from the grid as well as the batteries, but it cancels out from the spikes going in both directions.

 

[kta22]

The one box solutions do have a small advantage in that a single processor is seeing the currents and controlling things. In your Victron system as well as my Schneider/Victron/Enphase mix up, you have completely separate devices each doing a job. And a controller watching the power flow and making adjustments. My system only makes an adjustment every 5 seconds. I would bet the Victron is faster as it is all one brand and it directly knows all the data. When the batteries are near full, and the system is in DVCC or Float mode, the charge controller is running in a curtailed power mode. When a large load turns on, I bet you see the opposite spike. The batteries will need to supply the current first. Then the Cerbo will command the charge controller to crank up it's current output to cover the load and also supply your desired charge current. Then when the load turns off, the opposite happens. As you have seen, the charge controller is still putting out the commanded current to cover the load and the charging current. But without the load, ALL of the current becomes charge current. It takes some time for the Cerbo to see this current flow. Then it needs to calculate the new values, and then send the new current command to the charge controller. And then the charge controller probably also has a ramp rate and it has to adjust to match the new command. A few seconds is not a problem.

I intentionally made my system respond slower so that it would not over shoot and oscillate between charging and discharging. When my system sees a large change in load current, it will only adjust the charge current or inverter output by 80% of the error. In 5 seconds, it makes another 80% charge of the remaining error. On the 3rd 5 second loop, the current is usually in range. So it now takes my system about 15 seconds to cover a large load change. As long as the resulting current swings are well within the power ratings of the batteries and devices, don't worry about it. In my case, it results in a bit of power flowing in and out from the grid as well as the batteries, but it cancels out from the spikes going in both directions.

Thank you for the details, I guess in the end it will not matter so much as I'll be close to 1260 Ah.

 

[sunsurfer]

It's a bit strange because I previously had 2 Growatt's(SPF 5000es) in parallel and I didn't notice any kind of voltage or current overshooting in terms of battery charging. May be due to different software control and may be due to high frequency inverters.

Or different reporting/measuring. You might not be seeing it, but it there. The meter is just giving you a 1 sec average for example.

I will say, I notice it more with lead acid then LFP using the same SCC

 

[kta22]

Or different reporting/measuring. You might not be seeing it, but it there. The meter is just giving you a 1 sec average for example.

I will say, I notice it more with lead acid then LFP using the same SCC

You're right, could have been there for Growatt, data sampling is one very important aspect.

 

[kta22]

Other issue I've been observed these days is that when I'm charging the car(7.2 kWh) and mppt's are in float mode, the smart shunt reading is displaying a low charging/discharging current (to keep 53.6V float voltage), but in fact the real current (charged discharged) from battery is arround 10-11A / pack ( I have 3 packs => ~30A). Is like mppt is trying to keep float voltage with high current and rapid change (charge/discharge 30A @ every couple of seconds) instead of going slowly to float voltage in an "analogical" way and to stay there. If I remove inverter load, things are going back to normal. The higher the load, the higher is the mppt current keeping the float voltage(e.g. : mppt =30A @ 5.5 kW load, mppt= 50A @ 7.2kW load).

 

[Q-Dog]

Other issue I've been observed these days is that when I'm charging the car(7.2 kWh) and mppt's are in float mode, the smart shunt reading is displaying a low charging/discharging current (to keep 53.6V float voltage), but in fact the real current (charged discharged) from battery is arround 10-11A / pack ( I have 3 packs => ~30A). Is like mppt is trying to keep float voltage with high current and rapid change (charge/discharge 30A @ every couple of seconds) instead of going slowly to float voltage in an "analogical" way and to stay there. If I remove inverter load, things are going back to normal. The higher the load, the higher is the mppt current keeping the float voltage(e.g. : mppt =30A @ 5.5 kW load, mppt= 50A @ 7.2kW load).

This sounds normal. Not sure what you think should be happening.