Poor Grid Setpoint Tracking Multiplus IIs in Parallel

Inverters Victron
1

I upgraded my single 5kVA multiplus II to two of them in parallel.

Since the upgrade the grid setpoint tracking is A LOT worse than it was before.

With the one multi it was tracked very well. Now with the second one, there is a lot of feed in back in the grid and as the load increase, the feed back gets more. For example with a load of about 5kw there is a constant feed in back to the grid of 300W. With about 2.5kw there is a constant feed back to the grid of about 200W. And with low loads there is a constant feedback of about 50-100W. This isnt spikes when the load changes, but a constant feedback.

Luckily my prepaid meter doesnt trip - but it is charging me for the feed back into the grid…

Does anybody know how to improve the grid setpoint tracking when you use multis in parallel? Does it help to use an external energy sensor?

Thanks!

2

How old is the first Multi? If older than 18 months, its not compatible with the newer hardware models.

If they are the same hardware version, you have higher impedance on the input of one inverter and that causes the same symptoms.

Please ensure you input and output wires are exactly the same length and no loose connections. Also try making the new one the master, it might help.

4 Likes
3

You can phone me on 0713413280, and I can quickly explain what I am trying to say above.

2 Likes
4

Thanks Jaco for all the help!

5

Hi, I have exactly the same problem, can you please tell me if you managed to sort it out by adding a grid meter or if there was another solution

6

Hi - a grid meter doesn’t help. There are two possible issues - if the two multis are not the same age you might get this. Also the ac wiring needs to be different than you expect - the victron wiring manual discusses it. If I remember correctly basically you need thinner and longer than expected wires at the ac out so that the impedances can be matched.

I swapped my one new multi for and older one - to match my old one. This improved the problem to some extent. But still had a lot of power being sent to the grid. I didn’t want to change the wiring to get it perfect, so what I did is to swap the master and slave units in the setup. This causes constant current draw from the grid in stead of a push of current into the grid. I can live with it as it is now…

7

I actually think making the original slave the master did the most to make it ok. I can live with a constant current draw but not with a constant push into the grid. Its very easy with veconfig to swap around the master and slave units.

8

I’ve swopped my master and slave and I no longer constantly feed back to the grid but now it’s changed to be perfect if my load is under 2000w then it starts pulling from the grid roughly 20% of the load above that, I’ll update you if I work out anything further

9

Whenever you see a proportional pull/feed-in develop as the power level rises, it means the two parallel paths have different impedances.

A future version of Venus will have an warning for this, if all goes well.

10

Would it in future br possible for the Venus to compensate for this? I know if it was not really complex it would have been done already.

11

No, that will never happen. Not with the Multiplus or the Quattro. The physical layout of the topology makes it impossible.

Well, if I were to challenge myself: It is probably technically doable by letting the slave units run at a different PWM ratio to the master, but in the long run I think that’s an even worse idea. It masks a problem that really should be fixed higher up, and leaves capacity on the table (one unit will always do less work than the rest).

The new HF range (the Multi-RS and the future Quatro-RS) will handle it differently though.