Mecer 5kVa settings and monitoring

[egemnaar]

Next week Monday the doctor coming to sort out my installation. I found a reliable installer in the Nelspruit . Fuses, dc breakers, even dc cables will be replaced, battery monitor (702  with temperature sensor, grounding . it will all be done. I already have the trial ICC program running on the raspberry pi. I just have a question.about the BMV 702. Do i connect the BMV to the raspberry and if so with what kind cable, or is this no need..

[___]

3 hours ago, egemnaar said:

702  with temperature sensor

This makes very little sense with ICC and a non-Victron inverter. The temperature reading can only be reported via the vedirect protocol and I doubt ICC even cares about it. You cannot adjust charge voltages based on temperature because the inverter simply doesn't support it. The only thing you can do is program the BMV to close its internal relay if the temperature goes to high and maybe do something with that. I would pretty much recommend that you rather use the auxiliary input on the BMV to do midpoint monitoring.

The temperature input on the BMV is of limited use even in Victron systems (for now), but there is a fix in the pipeline which we call Shared Temperature Sense whereby your BMV can share the temperature with your whole setup... that is going to be awesome :-)

3 hours ago, egemnaar said:

with what kind cable

You need a vedirect->USB cable. It's essentially a USB-ttl-serial converter, but with galvanic isolation between your Pi and the solar charger. They cost about 28 Euro, or around R550 once they land in the shops.

[egemnaar]

Thanks Plonkster. The one recommendation I had was to buy a Victron invertor and sell the Mecer 2nd hand. Is there a market for that?.  At the moment the installation is doing its job. i will use this setup till it gives the ghost and then i upgrade.

 

[pilotfish]

2 hours ago, egemnaar said:

Is there a market for that?

Sure, I will give you R5000 and keep it as a backup if it is MKS 5k version (mppt not PWM) and still 100% functional.

[egemnaar]

My installation has been made safe.Battery monitor is installed with midpoint monitoring. Now I need expert opinion on what has been recoeded. Mostly battery midpoint is at 0.22., but rises...

[Chris Hobson]

3 hours ago, egemnaar said:

My installation has been made safe.Battery monitor is installed with midpoint monitoring. Now I need expert opinion on what has been recoeded. Mostly battery midpoint is at 0.22., but rises...

That is typical for LA batteries. Mid point deviations occur during late absorb. If one is manually checking for out of balance batteries manually one would very often miss it as the deviation is only apparent for ½hour or so. A deviation of over 0.5V (1%) is of concern. I am not going to go into the details but you can read up on the topic in the BMV manual (page 30).

I would get a HA 02 whilst your install is new and save some pain later.

[egemnaar]

Chris, I have a HA02 installed.  I measured the batteries. They measure 12.6 - 12.7 - 13.4 - 14.0. All I can do is lowering the float and bulk charging voltage. I lowered both to 52V . the deviation is now less than 1V.

 

 

 

 

[Chris Hobson]

No at 52V you not going to charge your batteries within a solar day. What batteries are they?

How is your HA 02 wired?

[egemnaar]

The batteries are 4 x Narada 12NDT200. The HA02 is wired each output to one battery, red to positive, black to negative, in sequence.  As mentioned in another post for Narada's I

put bulk at 55.8 and float at 53.8.

[Chris Hobson]

Unless your temps  are very high should be 56.4V for bulk and 54V for float. There should not be such a big spread in battery voltage check all the connections of the battery interconnects.

If that does not remedy the situation take you batteries off to be tested. Can you confirm the HA 02 makes a high pitched whine when you have this large difference?

Check and double check your HA 02 wiring.

 

 

[Chris Louw]

Hi everyone . Does some one have experience of Axpert 5kva/5000watt with 4000 watt mppt , firmware 74 - 20.

Setting 32  Bulk charging time option (1) inverter to judge option (2)  5 to 900 min .

Setting 33 Battery equalisation. (1) Equalise.  (2) Disable

Setting 34 Equalise voltage  58.5v - 64 v

Setting 35 Equalise time  default 60 min or 5 - 900 min

Setting 36 Equalise time out default 120 min or 5 - 900 min

Setting 37 Equalisation interval default 30 days or 0 -90 days

Setting 38 Neutral to ground setting

Setting 39 Equalisation activated immediately

I Realise the settings are dependent of the battery type . Does anybody have any experience if settings are reliable .

[Chris Hobson]

Just remember you should not need to equalise  VRLA batteries. If you do you need to be aware that you will lose some electrolyte which cannot be replaced.

 

[Chris Louw]

I have a Royals batterybank on another inverter 3kva that cannot reach equalising charge . If this inverter can do a good charge it will help a lot

[Chris Hobson]

Aren't Royals sealed or semi-sealed batteries?

[Chris Louw]

Thanks Chris the batteries are marked seald . I would like to know what is best for them . Checked the balance between the 8 of them 12.60v to 12.63v after resting for 2 hours . Early morning voltage before charging 24.9 v- 25 v . These are 3 year old batteries that do only small loads , Big freezer 372 kW/ year . Just want to make the best use of them before I sell them for scrap .

[Chris Hobson]

I would be mildly concerned if the difference was 0.5V I would make sure I was trying to remedy the situation if the voltage difference was over 1V.  A difference of 30mV is not cause for concern.

[Chris Louw]

Thanks Chris I learn something everyday . My maths teacher when doing T2 always said the more you learn the more you find out how little you know . The H0 - 02 from Chris R is installed on my main system is and doing well . Thanks .

[Guest]

2 hours ago, Chris Hobson said:

I would be mildly concerned if the difference was 0.5V

After a few years one does not even know what the volts difference are - as it matters not that much anymore - as long as the water is topped up regularly (lead acids) and the SOC kept within reasonable parameters (all banks).

And now on the new lithium banks, even less to think about.

[___]

2 hours ago, The Terrible Triplett said:

And now on the new lithium banks, even less to think about.

Lithiums can be troublesome when charged all the way to the top. Their internal resistances goes up a lot (to charging, obviously not to discharging) and less than an amp is enough to push their voltage way up. And with so many inverters not having a current shunt on the DC side (this includes the Axpert).... fun fun fun :-)

[Chris Hobson]

6 hours ago, plonkster said:

Lithiums can be troublesome when charged all the way to the top. Their internal resistances goes up a lot (to charging, obviously not to discharging) and less than an amp is enough to push their voltage way up. And with so many inverters not having a current shunt on the DC side (this includes the Axpert).... fun fun fun :-)

I have never understood why everyone aims at 100% SOC with LiFePO4. It makes no sense. Lithium does not suffer chronic damage if not fully charged and has a very flat charging curve (3.2V = (roughly) 30% SOC and 3.3V = 90% SOC. Pylontech cheat as the modules as 15S  So 3.65V/cell is equal to 54.75V and  only charge to 53.2V allowing some headroom. Makes sense to me they are probably 97-98% SOC at 53.2 V

[___]

48 minutes ago, Chris Hobson said:

Pylontech cheat as the modules as 15S  So 3.65V/cell is equal to 54.75V and  only charge to 53.2V

Their BMS raises an alarm at 54V, so you don't have that much headroom. Interestingly, it transpires that just today I had the opportunity to ask someone at Pylontech about this.

You have to understand that the balancing circuits generally only kicks in in one of two places: The very top and the very bottom. At the top, energy can be transferred from high cells to low cells (active balancing) or simply bled off (passive balancing). At the bottom, you could do a similar active balancing to transfer energy to low cells. While you're in the middle (neither high nor low) no balancing can happen. So the trick is to take the voltage high enough so that balancing can happen, but to avoid 100%.

For pylontech, the balancers start to work at 90% SoC. Interestingly, many lithium makers in fact use voltage as an SoC indicator, and under charging it is actually quite reliable, so this literally translates to a charge voltage in many cases. For Pylontech, 51.8V equates to around 95% and is apparently enough to make balancing happen. 53.25V is assumed to be 100% SoC, even though the BMS allows you to go to 54V. That gives you a lot of headroom to work with.

LG Resu batteries, for example, specify a charge voltage the equates  to around 95%. At 58.1V they are 100% full, and at 57.7V (what the BMS asks for) they are 95% full. So their default configuration is to hold the battery around 95%.

[Chris Hobson]

10 hours ago, plonkster said:

Their BMS raises an alarm at 54V, so you don't have that much headroom. Interestingly, it transpires that just today I had the opportunity to ask someone at Pylontech about this.

So what other tidbits did you acquire?

10 hours ago, plonkster said:

For Pylontech, 51.8V equates to around 95% and is apparently enough to make balancing happen. 53.25V is assumed to be 100% SoC, even though the BMS allows you to go to 54V.

My batteries on reach 51.8V just before  we go to SOC 100% and hit the charging brick wall. I like that lithium spends the bulk of its time not at elevated voltages.

 

[___]

16 minutes ago, Chris Hobson said:

My batteries on reach 51.8V just before  we go to SOC 100% and hit the charging brick wall. I like that lithium spends the bulk of its time not at elevated voltages.

You see, the tidbits really goes both ways. It is feedback like this (that from 51.8V it's really quick up to 100%) that makes me very happy this morning. For that is exactly what we see: That last 5% is an extremely slippery slope and you can easily slip over if your charger's accuracy or response is lacking (and most of them are... the cost of building finely-tuned chargers is just not warranted).

It means this feeling that I am getting, that we really ought to park the sleigh at the top of the hill instead of trying to brake it lower down, is probably a healthy feeling.

It even makes perfect sense. On a cell level, what we're saying is that from 3.45V upwards it tends to go really quickly. And this seems to be the case for every battery I've seen so far. It's like driving from Cape Town to Johannesburg at highway speeds, stopping in front of your door with screeching tyres, and walking the last "5%" up the garden footpath (for in Jhb you still have gardens, presumably).

What is different, is that the different makers have a different way of translating that to an SoC percentage.

For example, Pylontech says the battery is 100% full at 53.25V, but the BMS allows you to go to 54V before it cries havoc, at which point it is technically 101% full! This translates to 3.6V per cell.

Similarly, an LG Resu asks for 57.7V at which point it is 95% full, but allows you to go to 58.1V at which point it is 100% full. What is more, the SoC they display is the real one... so customers sometimes ask why it stops at 95%, they actually expect it to go to 100% :-) Again, 58V is around 3.6V per cell.

The only important thing is to take them to a voltage high enough to activate balancing. It appears that 52V is ample for a Pylon.

Of course you are sacrificing 2.5% or so of the rated 2.5kwh of a module, or around 50Wh. In other words... it might be marketing that mandates the 53.2V level :-)

[Chris Hobson]

2 hours ago, plonkster said:

It even makes perfect sense. On a cell level, what we're saying is that from 3.45V upwards it tends to go really quickly. And this seems to be the case for every battery I've seen so far. It's like driving from Cape Town to Johannesburg at highway speeds, stopping in front of your door with screeching tyres, and walking the last "5%" up the garden footpath (for in Jhb you still have gardens, presumably).

Lovely analogy

 

2 hours ago, plonkster said:

That last 5% is an extremely slippery slope and you can easily slip over if your charger's accuracy or response is lacking (and most of them are... the cost of building finely-tuned chargers is just not warranted).

In which case the Axpert is handling it pretty well. We sometimes breach 54V (when my BMV sounds an alarm) but we do not get to 54.5V and by the time I have swiveled in my chair to look at the BMV the voltage is already dropping and we are heading to float of 50.5V ( @Coulomb and @weber's recommendation). Monster-in-law is with us while she recuperates from an op  and with winter approaching a 3rd Pylontech is on the cards. Will be interesting how the battery bank behaves with the extra module. Most literature has 100% SOC at 3.65V per cell. 

[Guest]

10 hours ago, plonkster said:

... we really ought to park the sleigh at the top of the hill instead of trying to brake it lower down, ...

If I carry that sleigh up that bloody hill, I am going down that hill on it. 

Where else would one have so much fun in the snow?