question

alan-falck avatar image
alan-falck asked

Multiplus 3Kva Internal Battery monitor only reaches 85%

Hi Community

  • Multiplus MultiPlus-II 48/3000/35-32 Firmware version 497
  • MPPT BlueSolar Charger MPPT 150/70 rev2 Firmware version v3.11
  • Battery Hubble Lithium 5.4kVA AM2 103Ah
  • Sensor: Victron CT Clamp connected to 3.5mm sensor socket (20m)
  • Battery Cables (balanced) 35mm2
  • ESS installed

I’ve been trying to get a 3Kva Multi to get to a full SOC using the internal battery monitor.

It seems to peek at around 85% even if I increase the charge voltages. This only started when I added a 150/70 MPPT and some solar. The MPPT settings are identical to the inverter. The Charge gets ‘throttled’ at around 85%. If I choose ‘Keep Batteries Full” I can get the charge to 95%.

If I connect the battery via a CAN interface (using a RIOT cloud link adaptor supporting Victron-enhanced CAN protocol on the Hubble) I can get SOC 95%.

The battery SOC jumps (within a few seconds) from 85% (internal) to 98% on CAN.

I’ve tried every permutation of battery setting from Victron & the manufacturer. I've asked several experienced Victron installers to check my settings, but I still seem to have an issue.

Any suggestion would be most welcome.

Regard Alan

3kva-low-soc-vrm.png

hubble-victron-settings-all-modles.png


1665059589443.png

RT NewBatt Voltages 10-oct-22 c0619ab201a6_inst_ttyS4_Interfaces_Mk2_Tunnel (10).zip

Lithium BatterySOC
3kva-low-soc-vrm.png (77.7 KiB)
hubble-victron-settings-all-modles.png (29.8 KiB)
1665059589443.png (28.7 KiB)
rt-newbatt-voltages-10-oct-22-c0619ab201a6-inst-tt.zip (2.3 KiB)
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

5 Answers
nickdb avatar image
nickdb answered ·

The internal monitor has no visibility of the DC devices so will always be horribly off in a PV environment. It is thoroughly document on this site.

You require a shunt for it to work.

The hubble's also have a reputation for their SOC inaccuracies.

(personally wouldn't touch one with a 10ft barge pole and 5 sparkies strapped to the end).

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

alan-falck avatar image
alan-falck answered ·

Hi thanks for the quick response Nicdb. I agree a CAN or Shunt is ideal (this is a very low-budget system). My question is Has the MPPT had such a big impact on this? I've had this system stable and charging to 100% SOC (even if it is not very accurate) for over 3 months.

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nickdb avatar image nickdb ♦♦ commented ·

Yes, before an mppt was added, the inverter would be charging so it would be mostly accurate because it "sees" all the power flows. I saw this with pure backup systems we had. Add the mppt, now the inverter does not see the flow from solar to battery and SOC becomes inaccurate.

Add shunts and Viola, problem solved.

0 Likes 0 ·
alan-falck avatar image
alan-falck answered ·

OK understood, I have to live with the internal battery monitor and its inaccuracy, I accepted the compromise.

Q: How do I get the SOC above 85% using the internal battery monitor? If I up the Float and Absorb voltages in VE Config the MPPT still throttles the charge at around 85%. If I choose 'Keep Batteries charge" (using the remote console widget) the SOC goes up to 95%.

My understanding is the MPPT is set to 'external control' so the inverter setting takes preference and even if I manually make the charge setting in the MPPT the same as the inveter it still only reaches 85%. Am I missing something here? Are the settings in ESS also dictating the SOC level (voltages).

Q: Am I correct in saying there are 4 areas which influence SOC levels:

A: Charge settings in VE Config, B: Setting in ESS ? C: Settings on the MPPT D: The BMS on the battery?

FYI the rest of this system and ESS self-consumption mode is working perfectly.

I would really appreciate some guidance here, has anyone solved this issue with a 'workaround'.

Thanks, Alan

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nickdb avatar image nickdb ♦♦ commented ·
Hi, you can't tune this behaviour away. It is like attaching loads behind a shunt and then trying to fix the bad reading with charger settings.

Why don't you want to use the BMS as a reference? That solves all the issues and means the system is under the battery's control.

0 Likes 0 ·
alan-falck avatar image
alan-falck answered ·

OK so as this is a budget system the Hubble battery by default (unlike our preferred BSL) doesn't support the full Victron CAN protocol. So we have to purchase the Riot Cloudlik device (fairly expensive). I'm also running Venus on Rasberry Pi, it doesn't support CAN via USB, so I would need a CAN HAT and some drivers and coding to load the drivers to get basic BMS SOC values into the Multi.

I'm using a 'loan' external Cerbo so I can at least test the various options, but this is also outside of this budget.

So maybe the point I'm missing is that Hubble does have some basic CAN functionally without the RIOT Cloudlink, and this may at least solve the SOC rather than using just this internal battery monitor. ? I'll give this a try, have I understood you correctly on this one ?

PS I believe Hubble has made it onto the Victron 'recommended' battery list eventually (we have very little choice in battery supply at the moment).

Thanks

1 comment
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

nickdb avatar image nickdb ♦♦ commented ·

Hi Alan, Hubble is indeed on the list now, best check with them what you can or can't get with the basic battery. It would be ideal to have the battery manage the system.

Unfortunately with cost constraints come trade-offs.

SOC is a manufactured metric, more abstract in its calculation.

Even some managed batteries struggle to get it right, shunts can be a right pain to tune so an internal measurement which misses a big part of the picture is only going to be a problem.

The system though works on voltage, so while it reports 85% it is probably full which is why the mppts provide less. The only thing this throws off is ESS, which can be annoying but is unavoidable until you have a device capable of doing an SOC sync.

So really your choices are:

1) Try use the BMS

2) Find the lowest cost shunt option (probably cheapest and works with the Pi).

3) Live with the drift

Edit. According to the AM2 (rev2) specs it has an integrated CAN port, so as long as you have a CAN port to connect it to, and follow their setup doc for victron, it should all work. No RIOT module required.

1 Like 1 ·
alan-falck avatar image
alan-falck answered ·

Thank you !

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Related Resources

Additional resources still need to be added for this topic

Battery Compatibility

Did You Know - How to create a battery profile for non-Victron batteries?

Additional resources still need to be added for this topic