Voltage spikes

[MikeintheShed]

Hi,
I have a few questions for the group.

I have a single panel 120w setup for a summerhouse just to power my chargers for a tablet, phone and a rechargeable lamp.
I know it is a "cheap" setup. Eco-worthy, with a 30a Lifepo4. The controller is the standard PWM 30a.
It ran great on a SLA battery but as I added a Lifepo4 and set the controller to Life I get voltage spikes over 16v. This shuts off my speaker so I have put the SLA battery back on to save killing anything.
I see that is seems to be from the BMS shutting off etc.

So to the questions:
Will a new battery solve it or are the BMS all the same?
Advice seems to be lower the voltage, if I do this will it still charge the battery fully? ( I tried it and at 13.5 it still spikes but to 14.6v)
Will the spikes do any damage to the battery or controller and could I let it spike and smooth out the 12v, a capacitor?
Would a MPPT controller control the spikes?

Thanks,

Mike

 

[sunshine_eggo]

Will a new battery solve it or are the BMS all the same?

Yes or no. Imbalance in a new battery is extremely common across all brands. Just as likely a replacement will behave similarly.

Advice seems to be lower the voltage, if I do this will it still charge the battery fully? ( I tried it and at 13.5 it still spikes but to 14.6v)

Ensure the charger is not set to equalize. Recommend you leave it in the 13.5V config for several days or until you notice it stops spiking. Once it stops spiking, raise to 13.8V and assess.

Will the spikes do any damage to the battery

Probably not.

or controller

Probably not.

and could I let it spike and smooth out the 12v, a capacitor?

Doubt it, but it might be fun to try.

Would a MPPT controller control the spikes?

No. They have the same problem, but the magnitude might be less. Regardless, most are not willing to trust cheap PWM with LFP.

 

[Texas-Mark]

Regardless, most are not willing to trust PWM with LFP.

I agree with all of what you said except this. I've been running dual Xantrex C60 controllers on one of my lifepo4 battery banks for a few years now with zero issues. I have mulled over swapping them out due to some of the "talk" here about using them. However, I have not found/seen a compelling reason yet. So IMO "most are not willing" = "people read something on an internet board, but never actually did it". Sort of like the those that say mixing of lifepo4 and FLA/AGM batteries won't work (which I also have done for years and still have one bank set up like that and also works fine). YMMV of course, but I am not one of the "most".

 

[sunshine_eggo]

I agree with all of what you said except this. I've been running dual Xantrex C60 controllers on one of my lifepo4 battery banks for a few years now with zero issues. I have mulled over swapping them out due to some of the "talk" here about using them. However, I have not found/seen a compelling reason yet. So IMO "most are not willing" = "people read something on an internet board, but never actually did it". Sort of like the those that say mixing of lifepo4 and FLA/AGM batteries won't work (which I also have done for years and still have one bank set up like that and also works fine). YMMV of course, but I am not one of the "most".

MOST people do not invest in quality PWM controllers. Xantrax "C" PWM controllers are certainly an exception to the statement.

I have revised my post to add "cheap" before "PWM"

 

[Hedges]

PWM will connect PV panel to battery and deliver full current.
I would expect that to drive a LiFePO4 cell to over voltage. Might not happen if cells are well balanced and controller disconnects again fast enough.

MPPT will deliver a regulated voltage, not exceed that. Only way a cell would go over-voltage is due to imbalance.
Imbalance can happen, but if absorption voltage and time allows BMS to balance sufficiently, then shouldn't be a problem.

For instance, $62 buys a Victron 75V input, 12V/24V output, 10A MPPT

Amazon.com

 

[Texas-Mark]

PWM will connect PV panel to battery and deliver full current.
I would expect that to drive a LiFePO4 cell to over voltage.

Have not had it happen yet. I should add that mine do have the ability to set the voltage exactly where you want it, so I never come close to an over-voltage situation. Also, having more than one battery/cell in parallel, will mitigate any issue with one battery acting up.

MPPT will deliver a regulated voltage, not exceed that.

Not entirely true. There are plenty of posts here where a BMS opens and the MPPT voltage spikes before the controller can react. In either case (controller type), if a BMS does not react fast enough on cell OV, you can have an issue. But once it opens, there is no longer an issue.

 

[sunshine_eggo]

PWM will connect PV panel to battery and deliver full current.
I would expect that to drive a LiFePO4 cell to over voltage. Might not happen if cells are well balanced and controller disconnects again fast enough.

MPPT will deliver a regulated voltage, not exceed that. Only way a cell would go over-voltage is due to imbalance.

There are MANY accounts of MPPT controllers, including Victron, failing to clamp down on current and spike the system voltage when the BMS cuts the battery out of the charge circuit due to charge protection from cell imbalance.

 

[MikeintheShed]

Great..
The controller will not equaliser in life mode so that is off.
I will swap back to the Life and give it some time and see if it stops.
Thank you,

Mike

 

[sunshine_eggo]

If this is due to battery cell imbalance (and it almost certainly is), the BMS simply needs time to balance the cells. To give you an idea of severity, a typical passive BMS can balance ABOUT 1Ah per 24 hours when it's held at elevated voltage or in charge protection mode.

Whenever possible, this is best done on the bench on an AC charger or power supply that can hold the battery at 13.8V or higher.

 

[DaveF]

There are MANY accounts of MPPT controllers, including Victron, failing to clamp down on current and spike the system voltage when the BMS cuts the battery out of the charge circuit due to charge protection from cell imbalance.

Yes, I've had spikes while using Morningstar Tristar MPPTs.

 

[Hedges]

There are MANY accounts of MPPT controllers, including Victron, failing to clamp down on current and spike the system voltage when the BMS cuts the battery out of the charge circuit due to charge protection from cell imbalance.

Yes, but does PWM full current pulses cause a runner cell to go over-voltage (where MPPT would not have), resulting in BMS disconnecting?

I'm trying to address whether MPPT would result in a dependable system while PWM would result in shutdown, possibly failure.
MPPT will deliver steady voltage, and battery current draw tapers off. PWM force-feeds approximately PV Imp into battery.

 

[sunshine_eggo]

Yes, but does PWM full current pulses cause a runner cell to go over-voltage (where MPPT would not have), resulting in BMS disconnecting?

I'm trying to address whether MPPT would result in a dependable system while PWM would result in shutdown, possibly failure.
MPPT will deliver steady voltage, and battery current draw tapers off. PWM force-feeds approximately PV Imp into battery.

Not sure. Both types will spike their output voltage when the BMS cuts the battery out of the circuit thus triggering an inverter over-voltage shutdown. Frankly, I've heard of it happening far more on MPPT than PWM, but that's probably because more folks use MPPT with LFP than PWM.

 

[Texas-Mark]

Yes, but does PWM full current pulses cause a runner cell to go over-voltage (where MPPT would not have), resulting in BMS disconnecting?

Don't see how it matters how a cell receives the current. If the cell goes OV, it goes OV and the BMS will open. If you are charging that close to a cell's limit, time to fix the pack balance issue.

 

[sunshine_eggo]

Don't see how it matters how a cell receives the current. If the cell goes OV, it goes OV and the BMS will open. If you are charging that close to a cell's limit, time to fix the pack balance issue.

Which is commonly required on a freshly purchased "sealed" LFP battery. Like one should fully charge an automotive battery before installation, one should fully charge and evaluate the balance of a new LFP.

 

[Hedges]

Don't see how it matters how a cell receives the current. If the cell goes OV, it goes OV and the BMS will open. If you are charging that close to a cell's limit, time to fix the pack balance issue.

Difference being PWM applies say 10A @ 14V, (21 Voc) to a battery as square wave.
MPPT applies 14V (CV) and battery only received 0.5A as SoC rises toward 14V.

MPPT is CV, ~CC.
PWM applies PV panel's IV curve to battery, dropping to zero current whenever voltage reaches setpoint.

 

[Texas-Mark]

Difference being PWM applies say 10A @ 14V, (21 Voc) to a battery as square wave.

Still don't see how that matters. You have other cells in the pack that will load the controller. On my PWM controllers, current still tapers down to near zero when the battery is nearly fully charged. A non issue IMO.

PWM applies PV panel's IV curve to battery, dropping to zero current whenever voltage reaches setpoint.

Again, whats the problem?
Like I said, I have been using PWM for years without any issues. <shrug>

That said, MPPT controllers are certainly more flexible with panel configurations, but IMO there is nothing inherently wrong with using them with lifepo4.

 

[Hedges]

On my PWM controllers, current still tapers down to near zero when the battery is nearly fully charged. A non issue IMO.

How could that be, with PWM? Unless Voc is barely above battery voltage?

Average current may taper down. But every time FET turns on, it is connecting PV array directly to battery, and you get approximately Imp.

 

[Texas-Mark]

How could that be, with PWM? Unless Voc is barely above battery voltage?

Not exactly "barely" above. They are "12V" panels (22.5Voc) going to a 12V bank. Matching input to output of PWM controllers is critical, and why I said MPPT is more flexible on the panel configurations.

While I don't know the specifics of how my PWM controllers work internally (since I don't have schematics for them), all I know is as the batteries get closer to full charge, the charge current goes down just like it does on my MPPT controllers. Never have a problem with any battery/cell going OV even if it is full sun and the batteries are still full.

As was noted earlier, some of the cheaper PWM controllers may not work as well, but if they don't slow the charge down (via PWM) then how would it be any different than just connecting a (large) panel to a battery without any controller at all (which actually would result in OV issues)? My point is even if PWM is giving high voltage/current pulses for a brief period, the battery/cell is not going to suddenly go OV unless it was already too close in the first place. I assume (based on observing mine) that as the battery voltage reaches "full", the PWM controller just stops charging. Mine even "float" properly.

 

[efficientPV]

I'd put a large capacitor on the battery > 10,000F to absorb some of the spike. I would never put a PWM on a lithium. I've done it but only after a buck converter set to 14.5V.

 

[Hedges]

You're right, capacitor and suitable switching speed to fix it. PWM pulse between current source and capacitor becomes voltage sawtooth.

PWM without capacitor into lithium battery would force ~ Imp. Into disconnected BMS would apply Voc.
(MPPT is PWM into inductor, makes variable current sawtooth.)

Inverter capacitors probably do the job. So long as not separated from PWM SCC connection point by fuse, breaker, switch.
System topology is key.

Makes sense Texas' system worked fine. Including Quality not Cheaper PWM.

 

[MikeintheShed]

Hi,
In the last 10 days I have had the Life connected and the charge rate at 13.6v. It's been stable here but after a week of this, if I raise it to 13.7v it spikes again. If 13.6v is the max stable charge I get, will this be good to continue with without any problems later?

I got hold of customer support but they did not have any suggestions why it it like this but said they are sending me some replacement parts. I will just have to wait and see what they send and go from there.

Mike.

 

[sunshine_eggo]

I can only guess that 13.6V is too low to encourage balancing, or it's happening to slowly. Some BMS will only balance when actively charging or when in protection mode (which results in a spike).

Has this charger controller ever successfully charged ANY battery to 14.4V?

As far as 13.6V being a workable voltage, possibly. The risk is that the battery isn't actually getting to full charge, and it won't deliver full capacity. If the cells are balanced, yes, but they likely aren't in this case. 4 cells at 3.40V are essentially fully charged, but an imbalanced battery with one cell at 3.65V and 3 at 3.32V is not fully charged. A capacity test would help answer this question.

 

[MikeintheShed]

I can only guess that 13.6V is too low to encourage balancing, or it's happening to slowly. Some BMS will only balance when actively charging or when in protection mode (which results in a spike).

Has this charger controller ever successfully charged ANY battery to 14.4V?

As far as 13.6V being a workable voltage, possibly. The risk is that the battery isn't actually getting to full charge, and it won't deliver full capacity. If the cells are balanced, yes, but they likely aren't in this case. 4 cells at 3.40V are essentially fully charged, but an imbalanced battery with one cell at 3.65V and 3 at 3.32V is not fully charged. A capacity test would help answer this question.

Thanks.
The charger has only been connected to this battery.
Even setting the charge rate to 14.8v the charge stops at 13.7 max... I guess it is the bms in the battery causing the issues.
Let's hope they send me a new battery.

 

[MikeintheShed]

I got another controller and it is the same as the first.
Tomorrow I will be asking for a replacement battery.

 

[HarryN]

I will look more carefully at what the bogart PWM does on some BBs when fully charged. I remember being surprised that my volt meter didn't see any significant spikes during charging ( as in literally nothing ) and have never seen it force the BMS closed. May take a few days.