Hey everyone, I’ve recevently (7.12.) updated my setup from a 3x Multiplus II 5000 to a 6x to be able to charge more energy during the night, when the price is low.
My main objectiv (more charge power) works fine, but as sideeffect I see that the ESS-controll loop seems a lot slower now.
Below you can see the grid meter readings (EM24). I thought that this could be caused by an overload of my rpi2 (load values often closer to ~0.8) so i switched everything to an Ekrano (20.12.) but that didn’t have a big effect.
The Gridsetpoint is set to -50W. In total we’re talking about ~2kW/day, not the end of the world but also slightly disappointing. Can anyone help me with this? would it be possible to switch off 3 of the 6 (as I only need all 6 to charge at night) ?
You have a system with parallel inverters now, this could also be an artifact of unbalanced wiring. They are significantly more fussy to set up.
You could remove them from the parallel set up (in software set up and otherwise and just have them as chargers - no ess- from the grid with some kind of progamming (node red or as a generator) for charge control. Since that is all you need out of them.
What you can’t do is switch them off while they are part of the 3phase parallel set up. The whole system will switch off.
A different approach since you seem to have another option for gxs then use the other gx for control and still use ess with scheduled charge.
Hi and thanks for your input. I agree that I could run them as two independent setups as last resort. I have to admit that this setup would be quite disappointing.
Can you explain the unbalanced cables a bit more? All AC cables are the same (same diameter, same length) but for sure will have variations in the resistance. I don’t understand as of now how that would influence the ess speed. In my head the ess would do the exact same job as it did with 3 MP-2 and just divided all power values in two halves and send each half of it to master and slave
Looking at the indivitual inverters it appears like all Master inverter behave somewhat similar which is what I would expect (ESS is set to regulate total power, not each phase individually).
L2 Slave is pretty flat but at least close to the masters, L3 a bit distant and L1 Slave is really out there .. doing his thing .. and causing the problem
Obviously not all cables are the exact same length, which will always produce a variation, same for every connection in the system .. not very satisfying.
Can somebody help me to find documentation how the Multiplus is doing its Power generation and what parameters are send to the slave?
Is anyone of victron here and is willing to comment on this one?
I’m curious because it looks to me like an offset voltage is send to the slave, which results in the required power (as in P=power(dV,2) / R) …
Or does anybody else have an idea why the AC cable resistance makes such a big difference?
I guess Victron won’t change the firmware “all of the sudden” to use send Power values directly to the multiplus .. but at least to me this appears to be a lot less error prone. Or can I somehow “calibrate” the system? Are multi multiplus per phase setups so uncommon / untested?
Its all about ohms law. The devices function as one inverter and power is managed like that. There is a reason devices, cabling etc must be symmetrical. The split will never be perfect. If you’re below 7% you’re doing well. It will continue to change as uneven wear affects performance. I have various topics on this subject. I would never again opt for multiple units. The fewer the better.
yup, agreed .. every change effects the entire setup ..
last resort will be to switch two independant triple setups as suggested by @lxonline .. and use the second triple only as charger ..
but that would imply to send battery values from A to B, two Venus OS setups, two MQTT setups, DVCC will not work easily etc .. highly annoying
do you have any idea why victron is not sending power values to each multiplus?
OGPS
(Ed @ Off-Grid Power Systems - offgridps.com)
8
If I lived in Europe, I would agree with you. But here in the USA, Victron offers relatively few inverter/charger options. For larger systems, only the Multi-II 48/5000 operates efficiently. We don’t have any MultiPlus models (old or new M-II) greater in capacity than 5kVA. If you need two inputs, the Quattro 48/5000 has more than 2x the idle consumption. The next largest model is the Quattro 48/10000 and it has an idle standby draw more than 2x of the Quattro 48/5000. So in Canada, USA, and Mexico, you either put a system together using 2x, 4x, 6x, or more Multi-II 48/5000 or suffer large idle draws that are wasteful.
So, the best solution here - even for off-grid - is to either not use Victron products or go with multiple inverter/chargers. I envy those of you in Europe, Australia, Africa, etc. that have access to Victron’s full suite of products. For whole-home power with the grid, you need four Multi-II 48/5000 just to pass through the typical 200A service in N.A. since we don’t have ESS or purchase an external transfer switch (more $$$$) just to work around the limitations.
Victron is still the best solution on the market (at least in North America) for small, mobile systems. But beyond small, single phase systems Victron has chosen to lose relevance here. And my bitching about it hasn’t changed anything so I’m trying to bitch less and focus elsewhere. I obviously failed today
Agreed, on top of the need to have multiple devices in some situations the problem seems to be quite solvable.
The workaround that I’m testing at the moment adjusts the setpoint once every 100sec to match a low pass value of the grid with my desired value. A control loop on top of a control loop .. Bad design, but worked for the last 2h. Just too bad that victron isn’t implementing a better multi device setup.
I’ll keep testing. Maybe others have more good ideas
