Still struggling to fully understand top balancing with bench top power supply

[Solarfun4jim]

@Will Prowse @FilterGuy @snoobler
Looking at the operation of these bench top power supplys, you can either use as a regulator source or a steady flow source. In the regulator mode, you adjust the current knob clockwise to maximum, adjust voltage to the value required then connect the load. In this case the CV led is illuminated.
In steady flow mode, you adjust the voltage knob to the required voltage, adjust the current to minimum, connect the load, then adjust to the current value you wish. In this case the CC led is lit.
Now, in Will's video(at around 3:20), he appears to do the first, by cranking up the amps to full, before connecting the load/battery bank.
When i done this in my top balance, the current/wattage continually fell all the way through the charge, whilst the voltage climbed up to the set point.
I had been expecting the current to remain static up until the set point was reached and then see the expected fall off in current, but that didn't happen. However, i felt i was following the instructions given....so this had to be correct.
Looking at charging charts, the first phase is the CC phase where the current remains constant.....so, this begs the question, did i set up my bench top supply wrongly? Should it be set up for steady flow, so that the current remains the same till the voltage hits the 3.65v, then expect it to drop? Unfortunately, having zero experience in all this, i made the assumption, that the falling current/wattage input all the way through was a function of the bench top power supply providing less than the needed 0.05C end point charging current, not that i had screwed up following the instructions provided in wills video.
Hoping this makes sense and that someone can illuminate. When top balancing, should the CV led or the CC led be lit up until the voltage reaches 3.65v?

 

[snoobler]

Very common issue.

You're seeing the result of high resistance in the system. The cells may only be at 3.3, but the PS reads 3.65 because of resistance from the connections, wires, plugs and internal to the unit. At that point you're in CV mode, and the current will decrease to hold 3.65. Thus, your power decreases.

There's only so much you can do - short, thick wires, ring terminals and as solid a connection as you can get where you plug into the power supply, but the reality is these cheap power supplies have a lot of internal resistance when trying to push 10A @ such a low voltage.

Best practice is the series the cells with the BMS and charge at 10A until the BMS trips off. THEN you wire them in parallel and go through the process.

WORST practice, that I've recommended several times, is to set the voltage to something around 3.9V, BUT you have to watch the actual cell voltage like a hawk. The second you get to 3.65V, you cut the current.

LFP can be charged to 4.2V. There is essentially no benefit as you might only gain 1-2% of capacity between 3.65 and 4.20, and above 4.20, damage can occur; however, the occasional run up to that point is likely not going to harm anything, and keeping it at something lower, like 3.9V reduces the risk further.

Worst practice is ONLY to be considered when you can nearly continuously monitor the direct cell voltage (what a cell measures, not what the power supply reports), AND you're keeping a running tally of how many Ah you've input in all charging.

 

[Solarfun4jim]

Very common issue.

You're seeing the result of high resistance in the system. The cells may only be at 3.3, but the PS reads 3.65 because of resistance from the connections, wires, plugs and internal to the unit. At that point you're in CV mode, and the current will decrease to hold 3.65. Thus, your power decreases.

There's only so much you can do - short, thick wires, ring terminals and as solid a connection as you can get where you plug into the power supply, but the reality is these cheap power supplies have a lot of internal resistance when trying to push 10A @ such a low voltage.

Best practice is the series the cells with the BMS and charge at 10A until the BMS trips off. THEN you wire them in parallel and go through the process.

WORST practice, that I've recommended several times, is to set the voltage to something around 3.9V, BUT you have to watch the actual cell voltage like a hawk. The second you get to 3.65V, you cut the current.

LFP can be charged to 4.2V. There is essentially no benefit as you might only gain 1-2% of capacity between 3.65 and 4.20, and above 4.20, damage can occur; however, the occasional run up to that point is likely not going to harm anything, and keeping it at something lower, like 3.9V reduces the risk further.

Worst practice is ONLY to be considered when you can nearly continuously monitor the direct cell voltage (what a cell measures, not what the power supply reports), AND you're keeping a running tally of how many Ah you've input in all charging.

Thanks snoobler...so that is setting up using the first method then with the CV led lit. What happens if you charge using the second method with the CC led lit? At the constant current of 10A, would the voltage fail to come up to 3.65/cell?

 

[Gazoo]

Thanks snoobler...so that is setting up using the first method then with the CV led lit. What happens if you charge using the second method with the CC led lit? At the constant current of 10A, would the voltage fail to come up to 3.65/cell?

Having just went through this using a 12 amp Riden power supply in charge mode this is what I know for sure. I did not receive any connecting cables with my power supply. Usually the cables that come with power supplies are cheap. I ordered 12AWG cable and soldered on ring connectors to connect to the cells, and fork connectors to connect to the power supply. I am sure the result was a much better connection than using banana plugs and alligator clips. I noticed a .2 to.3 voltage drop between what the power supply was displaying and the voltage measured directly at the cell terminals. There will always be some voltage drop due to the high resistance of the cells.

Make sure you set your power supply to 3.65 volts and 10 amps before connecting to the cells. When you do connect to the cells the power supply should remain in CC until the tail end of the charge. During the tail end of the charge the power supply should go into CV mode and the current will taper off as the voltage goes up. Eventually the current should be zero, or close to it when the voltage displayed is 3.65 volts and it will take patience to get there. It took me about 5 days to parallel top balance my 8 cells. It is also a good idea to monitor the tail end of the charge.

I don't know which power supply you have but some of them don't like to run at the full current for a long time. So you might want to ease up on the amps a bit.

 

[snoobler]

Thanks snoobler...so that is setting up using the first method then with the CV led lit. What happens if you charge using the second method with the CC led lit? At the constant current of 10A, would the voltage fail to come up to 3.65/cell?

The unit is both CC and CV. It will be CC until the unit registers 3.65V (set max), then it will enter CV mode and taper current to hold.

 

[Solarfun4jim]

Having just went through this using a 12 amp Riden power supply in charge mode this is what I know for sure. I did not receive any connecting cables with my power supply. Usually the cables that come with power supplies are cheap. I ordered 12AWG cable and soldered on ring connectors to connect to the cells, and fork connectors to connect to the power supply. I am sure the result was a much better connection than using banana plugs and alligator clips. I noticed a .2 to.3 voltage drop between what the power supply was displaying and the voltage measured directly at the cell terminals. There will always be some voltage drop due to the high resistance of the cells.

Make sure you set your power supply to 3.65 volts and 10 amps before connecting to the cells. When you do connect to the cells the power supply should remain in CC until the tail end of the charge. During the tail end of the charge the power supply should go into CV mode and the current will taper off as the voltage goes up. Eventually the current should be zero, or close to it when the voltage displayed is 3.65 volts and it will take patience to get there. It took me about 5 days to parallel top balance my 8 cells. It is also a good idea to monitor the tail end of the charge.

I don't know which power supply you have but some of them don't like to run at the full current for a long time. So you might want to ease up on the amps a bit.

Thanks Gazoo,
When i was parallel charging the first time, i turns the voltage to 3.65v and had the current dialed up to maximum. Then i connected the load/battery pack. The unit stayed in CV mode and the current did not stay level until it reached 3.65v. In fact by the time i got it to 3.64v the current had also dropped to 0.15A
Now however, on using it to charge up my series connected pack at 29.2v, the current is staying level at 10A and the CC light lit. What i dont understand, is that i connected it in the exact same manner both times.
Anyhow, i will wait for the BMS to kick me off or else for the current to drop. I have the balancing function on, however the cells are still closer than 10mv , so i dont expect any balancing till i reach right at the top end, if at all.

 

[Solarfun4jim]

The unit is both CC and CV. It will be CC until the unit registers 3.65V (set max), then it will enter CV mode and taper current to hold.

Thanks snoobler, but this was not my experience. On my top balance, the CV light was lit right at the start and the current did not stay level, but continually dropped.

 

[ArthurEld]

Thanks snoobler, but this was not my experience. On my top balance, the CV light was lit right at the start and the current did not stay level, but continually dropped.

I'm sure snoobler will answer but Riden CV mode starts reducing the current before it gets to the target.
Like you say, if the target is 3.65V when the Riden is nearing 3.65 it will switch back and forth from 3.64 to 3.65 and reduce the amps. So, I guess that isn't really constant voltage.
I think my Mean Well power supply does use constant voltage and reduces the amps but when the amps get real low it doesn't stop charging. And it charges the cells over 3.65. It hasn't burned me but I have to be paying attention.
I don't think the Riden will over shoot the target voltage which is probably why it is popular.

 

[Solarfun4jim]

I'm sure snoobler will answer but Riden CV mode starts reducing the current before it gets to the target.
Like you say, if the target is 3.65V when the Riden is nearing 3.65 it will switch back and forth from 3.64 to 3.65 and reduce the amps. So, I guess that isn't really constant voltage.
I think my Mean Well power supply does use constant voltage and reduces the amps but when the amps get real low it doesn't stop charging. And it charges the cells over 3.65. It hasn't burned me but I have to be paying attention.
I don't think the Riden will over shoot the target voltage which is probably why it is popular.

Thanks Arthur, mine is a Long wei, bought solely for this purpose and probably wont see the light of day again once i have completed the balancing and capacity testing.
Never having used a bench top supply before, it is all a bit daunting, when you can easily cook your battery pack....lol.

 

[Solarfun4jim]

Guess my first top balance was crap. Now, on series charging the pack, i got to 98.5% SOC and the cells went crazy, with the cell diff mV rising sharply and the balancer failing to cope. Got kicked off very soon thereafter when one cell hit 3.65, but by this time cell diff was around 150mV :-(

Voltage difference is still 80mV as it approaches resting voltage. How close do the cells need to be before i can safely parallel them again and give it another go?

 

[Deleted member 9967]

Guess my first top balance was crap. Now, on series charging the pack, i got to 98.5% SOC and the cells went crazy, with the cell diff mV rising sharply and the balancer failing to cope. Got kicked off very soon thereafter when one cell hit 3.65, but by this time cell diff was around 150mV :-(

Voltage difference is still 80mV as it approaches resting voltage. How close do the cells need to be before i can safely parallel them again and give it another go?

As for paralleling. I have been told it is always safe to parallel the cells.
It is in series where the fun begins and the special smoke can happen.
hence the reason i have no desire to charge mine beyond 80 to 90% while in series.
Only in parallel can I and will I bother to try to go to the full 3.65.
And then only the one time to top balance them

 

[snoobler]

Thanks snoobler, but this was not my experience. On my top balance, the CV light was lit right at the start and the current did not stay level, but continually dropped.

That's exactly what I was describing in my initial post.

Thanks Gazoo,
When i was parallel charging the first time, i turns the voltage to 3.65v and had the current dialed up to maximum. Then i connected the load/battery pack. The unit stayed in CV mode and the current did not stay level until it reached 3.65v. In fact by the time i got it to 3.64v the current had also dropped to 0.15A
Now however, on using it to charge up my series connected pack at 29.2v, the current is staying level at 10A and the CC light lit. What i dont understand, is that i connected it in the exact same manner both times.
Anyhow, i will wait for the BMS to kick me off or else for the current to drop. I have the balancing function on, however the cells are still closer than 10mv , so i dont expect any balancing till i reach right at the top end, if at all.

Because you're charging at a much higher voltage. Resistance is worst at high current, low voltage. You have nearly 10X LESS resistance with series charging.

Guess my first top balance was crap. Now, on series charging the pack, i got to 98.5% SOC and the cells went crazy, with the cell diff mV rising sharply and the balancer failing to cope. Got kicked off very soon thereafter when one cell hit 3.65, but by this time cell diff was around 150mV :-(

Voltage difference is still 80mV as it approaches resting voltage. How close do the cells need to be before i can safely parallel them again and give it another go?

If you're saying all your cells were 3.50-3.65V, that's great.

I'm totally okay with a voltage deviation of less than 0.2V for UNKNOWN states of charge. You have a high confidence that all cells are at high state of charge. Parallel immediately.

 

[Solarfun4jim]

As for paralleling. I have been told it is always safe to parallel the cells.
It is in series where the fun begins and the special smoke can happen.
hence the reason i have no desire to charge mine beyond 80 to 90% while in series.
Only in parallel can I and will I bother to try to go to the full 3.65.
And then only the one time to top balance them

I thought the cells had to be fairly close in voltage before you connect them in parallel or else there is too much inrush current? Just how close do they need to be? I can bleed off some wattage with a 30ohm resistor to get them closer before i connect them in parallel, just dont know how close they have to be to start with?

 

[Solarfun4jim]

That's exactly what I was describing in my initial post.



Because you're charging at a much higher voltage. Resistance is worst at high current, low voltage. You have nearly 10X LESS resistance with series charging.



If you're saying all your cells were 3.50-3.65V, that's great.

I'm totally okay with a voltage deviation of less than 0.2V for UNKNOWN states of charge. You have a high confidence that all cells are at high state of charge. Parallel immediately.

Brilliant snoobler.....starting to grasp this a bit better....steep learning curve for one so thick as me. Thanks for bearing with me.

 

[Gazoo]

I thought the cells had to be fairly close in voltage before you connect them in parallel or else there is too much inrush current? Just how close do they need to be? I can bleed off some wattage with a 30ohm resistor to get them closer before i connect them in parallel, just dont know how close they have to be to start with?

Others have connected their cells with much higher voltage differences. I don't know how much of a voltage difference would cause a problem. I do know a 150mV is very safe to parallel connect.

Snoobler pointed out your charge in series went very well and I agree. The second time time I charged my 8 cells in series the highest cell voltage was 3.650 and the lowest 3.567. So you did good...

 

[Deleted member 9967]

Brilliant snoobler.....starting to grasp this a bit better....steep learning curve for one so thick as me. Thanks for bearing with me.

When I did mine they were pretty out of whack.
What I did was put the lowest then the next lowest, on up to the highest voltage and then connected the power supply directly to the lowest one in the bunch.
It worked for me anyhow.
But ONLY do this for parallel. For series I am still scared and learning LOL.
EDIT: Only add your BMS when in series. I broke mine trying to use it in parallel sadly.
Just unplug it. No need to remove all the wires.

 

[Gazoo]

Thanks Gazoo,
When i was parallel charging the first time, i turns the voltage to 3.65v and had the current dialed up to maximum. Then i connected the load/battery pack. The unit stayed in CV mode and the current did not stay level until it reached 3.65v. In fact by the time i got it to 3.64v the current had also dropped to 0.15A
Now however, on using it to charge up my series connected pack at 29.2v, the current is staying level at 10A and the CC light lit. What i dont understand, is that i connected it in the exact same manner both times.
Anyhow, i will wait for the BMS to kick me off or else for the current to drop. I have the balancing function on, however the cells are still closer than 10mv , so i dont expect any balancing till i reach right at the top end, if at all.

I see what you are saying now. In addition to what snoobler said that could have also been caused by the cables being too small that came with the power supply. From the description: Cable Conductor Wire Size: 0.15mm Copper Wire, 18AWG, super flexible. That's really too small especially when dealing with low voltages. In fact I used a wire gauge calculator and came up with 6awg for a small voltage drop. 12AWG is working ok for me.

 

[Solarfun4jim]

Others have connected their cells with much higher voltage differences. I don't know how much of a voltage difference would cause a problem. I do know a 150mV is very safe to parallel connect.

Snoobler pointed out your charge in series went very well and I agree. The second time time I charged my 8 cells in series the highest cell voltage was 3.650 and the lowest 3.567. So you did good...

That's reassuring you folks.....i thought those results of mine were way off. Somehow, because i had in my mind that all cells were calibrated to the one maximum charge level, that the cells would remain in balance at the top end, but get mightily out of whack at the bottom. Because my cells remain within about 10mV of each other all the way up till 98.5% soc or up to around 3.41v before they diverge, i just assumed after calibrating/balancing, they would remain very much closer all the way to 3.65

Appreciate all the input from folks.

 

[ArthurEld]

That's reassuring you folks.....i thought those results of mine were way off. Somehow, because i had in my mind that all cells were calibrated to the one maximum charge level, that the cells would remain in balance at the top end, big get mightily out of whack at the bottom. Because my cells remain within about 10mV of each other all the way up till 98.5% soc or up to around 3.41v before they diverge, i just assumed after calibrating/balancing, they would remain very much closer all the way to 3.65

That is exactly what happened to me and I thought something was wrong. But like you say, that is 98.5% state of charge.
It took me a while to realize that.
That is out of the usable area anyway. It is also why I think top balancing should be done closer to the high of the usable area.

 

[Solarfun4jim]

So, have just completed second parallel balance.....set it at 4v , gave me higher amps which initially lit the CC led, then shortly after, changed to the CV led. Watched it rise constantly on the meter and terminated immediately on hitting 3.65v
Going to call this done now. Charge voltage when pack was in series got to 3.42v before it started going whacky, so i will just charge in future to 3.4v
From the bottom end coming up, 3.2375v corresponded to around 25% SOC, so i think i will set the inverter lvd to 3.24 and the BMS to 3.23
That should roughly give me a range of 25% to 97.5% or 72.5% useable capacity, without putting the cells anywhere near the crazy gyrations. Hopefully this, along with low discharge rates, will keep the pack balanced for a very long time.