1 hour ago, Noobie said:

Unfortunately I can't upgrade the firmware on my inverter to work witj lithium batteryies because its a 24v Axpert and not the 48v unit.

Right. Your problem is that the Axpert is not compatible with Lithiums. Some of the very new 48V models finally have this now...

1 hour ago, Noobie said:

Would I have the same issue with a multiplus ii inverter?

No... except the Multiplus-II is not available in 24V (yet). You'd have to get the older Multiplus, or the MultiGrid if you intend getting NRS097 compliance.

If the problem is with solar charging (eg if you don't charge from the grid), then the solution is simply to get an MPPT that has a charging curve for Lithium batteries. Like a Victron BlueSolar.

16 hours ago, Noobie said:

Unfortunately I can't upgrade the firmware on my inverter to work with lithium batteryies because its a 24v Axpert and not the 48v unit.

No promises at all, but what version of main firmware are you running now? 

16 hours ago, Noobie said:

Would I have the same issue with a multiplus ii inverter?

No, I don't believe so (but @plonksterplease [ edit: did ] confirm). This particular blindness to the needs of LFP batteries, and the idea that absorb is over when battery current is low and the battery voltage is near the float setting, seems to be confined to Voltronic Power.

I'm keen to find out if the clone makers that seem to be evolving their firmware (Must Power seem to be the only one at this stage) have fixed these problems.

Edited April 29, 20197 yr by Coulomb

9 hours ago, plonkster said:

Right. Your problem is that the Axpert is not compatible with Lithiums. Some of the very new 48V models finally have this now...

No... except the Multiplus-II is not available in 24V (yet). You'd have to get the older Multiplus, or the MultiGrid if you intend getting NRS097 compliance.

If the problem is with solar charging (eg if you don't charge from the grid), then the solution is simply to get an MPPT that has a charging curve for Lithium batteries. Like a Victron BlueSolar.

Thanks @plonkster

I don’t have any panels connected to my Axpert, they all connect to a solis grid tie inverter. The Axpert is purely for backup power.

I see that there is a 24v multiplus-II on Victron’s website: https://www.victronenergy.com/upload/documents/MultiPlus-II-to-MultiGrid-and-MultiPlus-comparison-EN.pdf

But I don’t see it on the 2019 Q2 price list.

I will keep my eyes peeled.

7 hours ago, Coulomb said:

No promises at all, but what version of main firmware are you running now? 

No, I don't believe so (but @plonksterplease confirm). This particular blindness to the needs of LFP batteries, and the idea that absorb is over when battery current is low and the battery voltage is near the float setting, seems to be confined to Voltronic Power.

I'm keen to find out if the clone makers that seem to be evolving their firmware (Must Power seem to be the only one at this stage) have fixed these problems.

Hi @Coulomb

The Axpert I have only has a USB B port at the bottom of the inverter, not sure if this is a deal breaker?

Heres the pics of the data place and firmware, any assistance would be appreciated.

18 minutes ago, Noobie said:

But I don’t see it on the 2019 Q2 price list.

This is fairly typical. The documentation is updated but the hardware isn't in stock yet. It usually takes a few weeks for the first units to come in stock. Unfortunately can't say  how long.

Also, NRS097 certification will likely take a few months. The testing houses are very busy at the moment.

I have 2 x 24 Lithium-ion batteries in parallel

After 1 week connected to the inverter, with charging from the mains, both reached 100%

If I disconnect the mains (to simulate a power failure) and run the inverter from batteries, it draws 11 amps from both batteries

Within an hour I see that one battery is delivering 8.2 amps and the SOC had dropped to 78%, whereas the other battery shows 2.8 amps and SOC 92%

Should not the discharge from both batteries be more or less the same?   Blue Nova said a 10% difference is normal, but this is much more

Upon recharge, one battery reaches 100% SOC in 45 mins, the other takes 7 hours

3 hours ago, chrisc said:

Upon recharge, one battery reaches 100% SOC in 45 mins, the other takes 7 hours

Can you show a photo of your battery cables? It seems that they are not paralleled properly.

Is it possible that one of the cables is not crimped properly, or the connection to the inverter not tight, or the lug is tarnished?

Reading an earlier post of yours, you are using 10 mm² cable from each battery, but how are they combined? Usually you can't fit two lugs on the inverter's battery terminals.

Edit: alternatively, one is still balancing cells, since it's still quite new. But the 8.2 versus 2.8 A can't be explained away like that.

Edited April 29, 20197 yr by Coulomb

Believe me i received my Blue Nova batteries back today, The were hit by lightning on April 5.... What was damaged?  The BMS.  The BMS was replaced and upgraded. Nothing wrong with the Mecer, nothing wrong with the raspberry. All fine except to find out the ICC software no longer supports Blue Nova.

On 2019/04/29 at 10:53 AM, chrisc said:

Within an hour I see that one battery is delivering 8.2 amps and the SOC had dropped to 78%, whereas the other battery shows 2.8 amps and SOC 92%

Hey Chrisc 

Did you ever get this issue resolved? A friend of mine has a similar issue with two LFP batteries wired in parallel. The one discharges a lot faster than the other, and they are hardly ever at the same SOC. I’ve checked the cables and they look fine. He uses the “staggered takeoff” method from the terminals and this issue has been going on for three weeks now.....

When you put batteries in parallel, one battery will always (initially) do most of the work, and then the other one will join in a little bit later. That's completely normal, the capacity will never be exactly equal and the stronger battery will do most of the work. What will happen over time is that the stronger battery will wear faster, and then things should even out.

Another possibility is that the batteries have exactly the same SOC value but show different SOC values due to SOC measurement inaccuracies.

I have  BlueNova 52V 12kWh battery 😀 and find that its internal SOC measurement is hopelessly inaccurate if one has low-level loads (e.g. 200 - 300W). As a result I have installed a BMV712 to monitor the battery SOC - which works perfectly. See the comparison between the BMV and Battery BMS over the last few days.

 

According to the supplier this is a known issue and also applies to BYD batteries (see https://www.victronenergy.com/live/battery_compatibility:byd_b-box).  

6 minutes ago, NigelL said:

hopelessly inaccurate if one has low-level loads

Yup, and BlueNova is not the only battery with this issue. It's a common issue with LFP batteries.

It's another one of those things people don't understand about SOC. It's always an estimate. You get better estimates and worse estimates, but you don't really know where the bottom is until you hit it, nor do you know where the top is until the voltage rises. If you discharge a battery right to 0%, you may find that your indicated SOC hits 0% and then the inverter keeps going for a few more minutes, and conversely, you may hit 100% indicated while the battery is still accepting charge. Inverter makers know this, so if you configure things to switch off at 0% (as you might on some batteries, eg flow batteries), the actual implementation will DISABLE the soc mechanism and just run it till the voltage drops out 🙂

 

My overnight loads typically cycle between 150W and 300W so this results in wildly inaccurate SOC from the Battery BMS.  This is so inaccurate as to be effectively useless for any form of control of my system(e.g. ESS).  A while back I got a low-voltage alarm (below 50.5V) and the BMS was still showing 88% SOC.

Adding the BMV-712 solved everything.

26 minutes ago, NigelL said:

Adding the BMV-712 solved everything.

Yup. Lots of people do that, way more than you think

Well Plonkster and NigelL, it seems people must stop looking at the SOC lights and rather get a BMV-712. 

Thanks for the advice 👍

13 hours ago, NigelL said:

See the comparison between the BMV and Battery BMS over the last few days.

My word. I did not see this coming. Thank you for that picture Nigel.

Now my BMV will stay for a very long time, irrespective of the make or the materials the batteries are made of.

 

9 hours ago, The Terrible Triplett said:

Thank you for that picture Nigel

I have one from a BYD test run as well. Same issue. With a constant load of around 3A (or less), it will get stuck in the middle 90% when the real charge level is already close to 80%.

4 minutes ago, plonkster said:

I have one from a BYD test run as well. Same issue. With a constant load of around 3A (or less), it will get stuck in the middle 90% when the real charge level is already close to 80%.

So if you now have a BMV in the mix, your SOC is more accurate.

From what I gather the BMS is primarily for controlling the volts.

So the two work like a glove and hand together?

Which brings the next questions: All those banks out there without BMV's, I suppose with the BMS in play, no potential of "damage"?

7 minutes ago, The Terrible Triplett said:

So the two work like a glove and hand together?

 

7 minutes ago, The Terrible Triplett said:

no potential of "damage"?

There are multiple levels of "BMS". First, the possibility of damage is mostly voltage based. If the cell's voltage drops too low or goes too high, it might get damaged. So the first task of a BMS is to disconnect the battery if the voltage of an individual cell goes too low. This is not negotiable. The second task is to balance the cells, and in all but the cheapest of batteries this is also not negotiable.

Then comes the nice to have features. Some will have analog charge/discharge signalling, usually dry contacts/relays. That way the battery can signal when you should stop charging or discharging. Some will keep track of the SOC. Some others have full intelligent control and even tell you the charge voltage and current, and the very best ones will vary the charge voltage to accommodate for adverse conditions.

The more features you want, the more expensive the BMS.

Let's use an example, the VE.Bus BMS, or its smaller brother the miniBMS. It's almost as limited a BMS as you can get. It balances cells, and if there is an emergency condition it sends a signal to another device (either the Multi itself, or a contactor, or one of these), and that disconnects the battery. There is no SOC tracking, and no special voltage stuff. Why? Cause the Multi and/or a BMV can already do that stuff for you, so in a purely blue stack, this actually saves money.

17 minutes ago, plonkster said:

BMV can already do that stuff for you

This ties into  to our chat the other day on how important BMV's have become in today's systems.

Edited August 8, 20196 yr by Guest

1 hour ago, The Terrible Triplett said:

So the two work like a glove and hand together?

My system appears to play nicely with two "Battery Monitors" - i.e. the BMV712 and the BlueNova BMS.  You can configure the "Battery Monitor" setting on the CCGX to take its SOC readings from either device.  The additional data (temperature, cell imbalance, under-voltage/over-voltage warnings etc.) from the BMS is still available even if the SOC value is obtained from the BMV712.

EDIT: The BlueNova battery voltage readings appear to be very accurate, it's just the current measurement that is inaccurate at low values.  I think this is due to them using a dc current-clamp circuit instead of a current-shunt.

Edited August 8, 20196 yr by NigelL

31 minutes ago, NigelL said:

dc current-clamp

Hall effect sensor in other words. Yeah, those get noisy low-down, so they probably filter out the small signals as they are not accurate anyway.