Dynamic Power Limiter Limitations
You are probably expecting a lot from the Dynamic Power Limiter (DPL).
And sometimes you may feel disapointed, realising that the DPL does not balance the "zero feed throttle" (Nulleinspeisung) as well as you would expect.
Although it is understandable you may consider the cause of error to be in the DPL programming, there are some other areas you need to look for! We have listed some cases below.
In Summary: There are Limitations! Not due to the programming of OpenDTU-OnBattery Dynamic Power Limiter (DPL) but, due to inverters' design and behaviour. In most cases, if you leave enough margin, DPL will keep the "zero feed throttle" (Nulleinspeisung) despite the oscillations and you will hardly notice this behaviour.
If you are a perfectionist, you will notice the oscillations and start trying to understand whether there is an error in your setup. This Wiki article is aimed at you, in order to spare you hours of frustration... (and do not ask us how we know about it).
Community member peff74 made some interesting observations:
On close inspection of the performance of the Hoymiles HM-700, I noticed that an inverter with many Inputs connected to the battery, has a brief power dip of 50W about once a minute.
You can only see this on the electricity meter itself, as it spits out 2 values per second.
If the information one of these power dips reaches the OpenDTU-OnBattery, the DPL will proceed to the necessary adejustmnets and later back again. Result --> oscillation.
I have noticed this particularly strongly at power levels around 100-200W and when many Inputs are connected:
With only one Input connected to the Battery:
With many Inputs connected to the Battery:
whereas as of 250W, the oscillations seems to disappear.
The problem only exists in the 15-35% relative limit range. Whereby it occurs most strongly around 25%.
You can see it clearly in the live view. The limit has not changed, however inverter power jumps almost 20W per channel from one moment to the next:
We can not confirm whether this problem as a general issue of all inverters, or something related to a specific model or specific Firmware Version.
On a HM600 for example, the fluctuation happens also in the area of 15%-35% but, at much smaller scale ~6-7W per Input.
All Hoymiles inputs should be connected to the battery
What happens if:
- "All DC+ cables" do not have the same length? or
- "All DC- cables" do not have the same length? or
- "All DC+ & DC- cables" do not have the same length?
SW-Niko and spcqike were kind enough to write some explanations:
First of all, different length of cables = different losses.
It makes sense therefore, that all DC- cables have the same length and all DC+ cables have the same length.
However, there is a much more important reason:
If the DCminus inputs of the Hoymiles are connected internally, then the return current to the battery flows preferentially via the connection with the lower-impedance .
In other words, more current flows in one return line than in the other return line.
This could lead to an overload of the internal conductors or the external cables.
In the event of a fault (lower line high resistance or interrupted), the entire return current could flow via one line.
The external line can cope with this, but whether the internal connection can do the same, is questionable.
This is not the first time that Ground (DC-) is so important: This is the reason why galvanic isolation for the earth of the VE.Direkt cable is so important level shifter - optocoupler
A quick check on a HMS-2000 showed that all DC- lines are connected internally. No problem with galvanically isolated power sources (= solar panels), but if all inputs are connected to one source (= battery).... Yes!!!!
In Summary: Have same length DC- cables 👍
A more detailed explanation is here: https://github.com/helgeerbe/OpenDTU-OnBattery/wiki/HOYMILES
Most probably, differences in length of the DC+ cables is less of concern as the electronic limiter of the Hoymiles should offer enough protection
See here: https://github.com/helgeerbe/OpenDTU-OnBattery/discussions/984
One member describes:
The string inputs to the inverter were completely assymettric (I was too busy one Sunday afternoon to put things together again that I mixed a set of cables):
Another member describes: Took me a while to understand what your issue was, but yeah, I can see it now. The inverter is set to a certain limit, but it does not produce as much.
I had the exact same learning curve. My setup was working fine, then some day I saw the inverter was not producing as much as the limit allowed. The power produced by channel was wildly different and fluctuating a lot. I also thought that I had tightened all screws properly. And I had, but on some only one out of two screws reached the cable. See this image:
The screw below is down a lot more than the screw above, and that's because the cable does not reach the second screw. I removed the blue collar from the ferrule, such that the cable went down further and the second screw could reach it. There were other connections and screws that loosened a little over time. After going over all of them, the issue was gone.
I thought this was the perfect opportunity to share this experience with a little more detail. I hope I can find and link to this discussion the next time I need it 😉
Regarding contact grease: I also thought about it, and learned that it is actually not conductive. One uses contact grease to lubricate fasteners and prevent corrosion, but you actually don't put it between contact surfaces that shall conduct electricity. At least that's what I researched a couple of months ago. I don't use any and rewired the whole setup into a new cabinet recently. After learning that connections must be very good, I took care to double check every single one and I had no issue firing up the new wiring.