Charging with Sol-Ark 15k question

[RCinFLA]

When in doubt, check cell voltages with no cell current for at least five minutes to give cells a chance to reach equilibrium on no-load voltage.

If your no-load rested cell voltages are greater than 3.45vdc on each cell the battery is fully charged, and likely Columb counter is off.

3.45v x 16 cells is 55.2vdc, but total battery voltage does tell you if cells are out balance, so some cells are still below 3.45v while higher SoC cell voltages are making up the total battery voltage being over 55.2vdc. Battery monitors and inverters can only make judgement of 100% state of charge by total battery voltage, unless they are getting other info from BMS communications that tells it a different story.

BMS's can have various ways to declare full charge (like common method of forcing a cell overvoltage charging shutdown). It may only be based on total battery voltage or individual cell voltage achieving a minimum voltage. The later criterion is rare as it can run into conflicts when there is out of balance conditions that cause a cell overvoltage charge shutdown before lowest state of charge battery reaches minimum required voltage for 100% charge declaration.

Even allowing a BMS single cell overvoltage charge shutdown does not mean the lowest cell rested voltage is above 3.45v assuring the pack is fully charged.

If you do not let the Columb counters in BMS, battery monitor, or inverter's Columb counter to periodically reset to full state by fully charging, then 100% full is just based on Columb counter record keeping of positive and negative current flow over time. Without a full reference reset every few months the Columb counter error just continues to grow over time.

The best sure way for a Columb counter reset by a full state of charge is with BMS checking every cell voltage and reset Columb counter to 100% when charging current tapers off to very low level and all cells are above 3.50v. Allowing charging current to taper down to low levels minimizes cell overpotential voltage bump up due to charging current and the 3.50v minimum cell voltage allows for a little bit of charging overpotential bump up voltage margin at low charging current. If cells are out of balance you will unlikely be able to meet this criterion before the highest SoC cell trips a BMS overvoltage charge shutdown at which point the Columb counter will not be reset to 100%.

 

[SunDave]

When in doubt, check cell voltages with no cell current for at least five minutes to give cells a chance to reach equilibrium on no-load voltage.

If your no-load rested cell voltages are greater than 3.45vdc on each cell the battery is fully charged, and likely Columb counter is off.

3.45v x 16 cells is 55.2vdc, but total battery voltage does tell you if cells are out balance, so some cells are still below 3.45v while higher SoC cell voltages are making up the total battery voltage being over 55.2vdc. Battery monitors and inverters can only make judgement of 100% state of charge by total battery voltage, unless they are getting other info from BMS communications that tells it a different story.

BMS's can have various ways to declare full charge (like common method of forcing a cell overvoltage charging shutdown). It may only be based on total battery voltage or individual cell voltage achieving a minimum voltage. The later criterion is rare as it can run into conflicts when there is out of balance conditions that cause a cell overvoltage charge shutdown before lowest state of charge battery reaches minimum required voltage for 100% charge declaration.

Even allowing a BMS single cell overvoltage charge shutdown does not mean the lowest cell rested voltage is above 3.45v assuring the pack is fully charged.

If you do not let the Columb counters in BMS, battery monitor, or inverter's Columb counter to periodically reset to full state by fully charging, then 100% full is just based on Columb counter record keeping of positive and negative current flow over time. Without a full reference reset every few months the Columb counter error just continues to grow over time.

The best sure way for a Columb counter reset by a full state of charge is with BMS checking every cell voltage and reset Columb counter to 100% when charging current tapers off to very low level and all cells are above 3.50v. Allowing charging current to taper down to low levels minimizes cell overpotential voltage bump up due to charging current and the 3.50v minimum cell voltage allows for a little bit of charging overpotential bump up voltage margin at low charging current. If cells are out of balance you will unlikely be able to meet this criterion before the highest SoC cell trips a BMS overvoltage charge shutdown at which point the Columb counter will not be reset to 100%.

Indeed a full charge seemed to help a lot. The battery still stays at 90% a little longer than the other percentages, but then shoots up faster after it passes 90% so its just a very minor thing that I was worried about at first.

The Columb counter idea, and charging to full once in a while seems to be exactly the case. Thanks to all the smart people on this forum for pointing that out.

 

[SunDave]

The fog burnt off and we are getting a bit of ? Here's a few screen shots of what my system is doing today

That's awesome. Imagine summer now! It's nice that we set up our systems in November, that way we get to see the worst of it right away, and be happy with the good production, instead of bummed out by the bad production after a good start.

 

[Danke]

When in doubt, check cell voltages with no cell current for at least five minutes to give cells a chance to reach equilibrium on no-load voltage.

If your no-load rested cell voltages are greater than 3.45vdc on each cell the battery is fully charged, and likely Columb counter is off.

3.45v x 16 cells is 55.2vdc, but total battery voltage does tell you if cells are out balance, so some cells are still below 3.45v while higher SoC cell voltages are making up the total battery voltage being over 55.2vdc. Battery monitors and inverters can only make judgement of 100% state of charge by total battery voltage, unless they are getting other info from BMS communications that tells it a different story.

BMS's can have various ways to declare full charge (like common method of forcing a cell overvoltage charging shutdown). It may only be based on total battery voltage or individual cell voltage achieving a minimum voltage. The later criterion is rare as it can run into conflicts when there is out of balance conditions that cause a cell overvoltage charge shutdown before lowest state of charge battery reaches minimum required voltage for 100% charge declaration.

Even allowing a BMS single cell overvoltage charge shutdown does not mean the lowest cell rested voltage is above 3.45v assuring the pack is fully charged.

If you do not let the Columb counters in BMS, battery monitor, or inverter's Columb counter to periodically reset to full state by fully charging, then 100% full is just based on Columb counter record keeping of positive and negative current flow over time. Without a full reference reset every few months the Columb counter error just continues to grow over time.

The best sure way for a Columb counter reset by a full state of charge is with BMS checking every cell voltage and reset Columb counter to 100% when charging current tapers off to very low level and all cells are above 3.50v. Allowing charging current to taper down to low levels minimizes cell overpotential voltage bump up due to charging current and the 3.50v minimum cell voltage allows for a little bit of charging overpotential bump up voltage margin at low charging current. If cells are out of balance you will unlikely be able to meet this criterion before the highest SoC cell trips a BMS overvoltage charge shutdown at which point the Columb counter will not be reset to 100%.

With passive balancing in the BMS, won’t the higher voltage cells equalize with the lower voltage cells?

 

[Partimewages]

Passive takes from the high cell only.

 

[Danke]

Passive takes from the high cell only.

Correct.

 

[Partimewages]

It will not top up the low cells.

 

[Danke]

It will not top up the low cells.

Correct.

 

[RCinFLA]

With passive balancing in the BMS, won’t the higher voltage cells equalize with the lower voltage cells?

As long as the higher voltage cells don't trip a cell overvoltage that terminates charging.

Even if overvoltage cell trips and shuts down charging, the overvoltage cell should continue to bleed down by BMS balance dump and eventually reset BMS back to charging. The overvoltage shutdown may occur multiple times, extending the required absorb charging time, until the cells get balanced well enough to prevent a single cell overvoltage trip.

Also have to watch charger operation. Many chargers see a BMS charging shutdown with its charge current drop off as battery being fully charged and charger may drop from absorb phase to float phase voltage setting which will not likely do further balancing once BMS resets.

When you have BMS to inverter communications it usually prevents this premature charge termination by BMS telling charger to back down its charge current level to keep highest voltage cell from tripping overvoltage limit. It is not a sure thing, depending how out of balance the cells are the BMS may still trigger a cell overvoltage shutdown. With a BMS resistor dump balancing of only 50-100 mA, the charger may not be able to reduce its charge current low enough to prevent further rise in highest voltage cell.

 

[Rodney L]

That's awesome. Imagine summer now! It's nice that we set up our systems in November, that way we get to see the worst of it right away, and be happy with the good production, instead of bummed out by the bad production after a good start.

Circling back now that I've driven my system round the block a time or two...
If someone knows what BMS Pytes 48100r Has in them, I'd like to know.

As to your original question. I noticed my batteries were seemingly hanging at that Sol Ark displayed battery value of 90% for a unusually long time a couple of times during what seemed like several weeks of overcast weather .

So I fired up the console connection to my laptop and ran the latest version of PytesBat and what was causing what appeared to be a hold up was in fact one battery that was at like 78% and climbing while the other 5 were from 90 to 99% (PytesBat displayed values)

So it appears that Solark is/was displaying the average of my 6 batteries and seeming to take for ever to get off the solark displayed battery SOC value of 90%

My assumption is heavy evening load draw, followed by overcast days that limited charging to less than full for a fews days had a battery drift lower than the others.

Even with that said, all the cells in each battery pack had Delta of .006v or less, so the cells seem to be healthy and BMS doing great at what it does.

 

[Lt.Dan]

Even with that said, all the cells in each battery pack had Delta of .006v or less, so the cells seem to be healthy and BMS doing great at what it does.

If the battery is at 78% then it's normal to have the delta voltage that low. When battery % gets above 99% (3.4v/cell) is when they will diverge.

I would not take a .006v delta at 78% as a sign the cells or the BMS are good

 

[Rodney L]

If the battery is at 78% then it's normal to have the delta voltage that low. When battery % gets above 99% (3.4v/cell) is when they will diverge.

I would not take a .006v delta at 78% as a sign the cells or the BMS are good

Good info to know. I'll verify tomorrow but iirc it was close to that at 100%

 

[Oldphile]

I have 4 Pytes V5 batteries and I've also seen SOC stall at 90%. Once it moves to 91% charge current begins to ratchet down.

 

[SunDave]

I have 4 Pytes V5 batteries and I've also seen SOC stall at 90%. Once it moves to 91% charge current begins to ratchet down.

Well, at least we are all seeing similar things with the Sol-Ark Pytes combo. Indeed it seems like if I go a long time without a full charge, then it can stay on 90% longer. The good news is that the new is wearing off a bit, so it doesn't bother me. It seems to be working very well overall.

It also shoots up quickly after the 90% mark. If I were to draw a straight line from the 90% to 100% it is not terribly far off the rest of the curve.

 

[Rodney L]

Well, at least we are all seeing similar things with the Sol-Ark Pytes combo. Indeed it seems like if I go a long time without a full charge, then it can stay on 90% longer. The good news is that the new is wearing off a bit, so it doesn't bother me. It seems to be working very well overall.

It also shoots up quickly after the 90% mark. If I were to draw a straight line from the 90% to 100% it is not terribly far off the rest of the curve.

Today...

 

[Oldphile]

Here's a more accurate description of what I'm seeing. This is the corner where it changed from 90%.

 

[SunDave]

Very similar to what we see. Again, if you drew an imaginary line from 90 to 100 it would be a more normal looking curve.

My best guess is that Pytes doesn't attempt to report the range of 90-100 super accurately. At least in this firmware version. With the voltage curves of this battery type, it's probably within fractions of a volt between 90, 91, 92% for example. Just a guess.

Could also be waiting for one battery in a stack to "catch up". Another wild guess since I have nothing to back the theory.