MPPT solar controller and LiFePO4 battery for backpacking

[karrak]

How can the GV-5's OCV problem be easily circumvented? Boz at Genasun suggested that I look into a TVS, but I'm not dealing with transient voltage spikes. Is there some type of fast acting, weather-proof-able circuit that can keep the max load terminal voltage at or below 15V without fail, without adding significant size or weight and without drawing significant current from the battery?

Edit: Would something like this work? If, as described, it adds noise, and a capacitor wouldn't completely eliminate the noise, this might not be good for a ham radio application. On the other hand, the benefit of preventing damaging voltage would outweigh the inconvenience of noise as long as the noise only happens when the circuit is actively limiting high voltage.

You are doing very well at learning about all this, before you know it you will be an electronic engineer.

Did Boz confirm the GV-5 OCV problem? His comment regarding TVS makes me think that the overvoltage will only be a transitory event.

A better device than a zener is a Transient voltage suppression diode.

Edit 2: According to Bioenno, the battery balances at any voltage as it's being charged or discharged. So I should be able to save money and go with a default 14.2V GV-5.

When you say default you do mean the 14.2 volt Li version not the Pb version?

Simon

[Dave C]

You are right that if you can get a GV-5 programmed for 14.6 volts that all the cells should be balanced on each charge if the balance voltage is 3.65V. As 3.65 volts is the maximum voltage that you should charge a LFP cell to I would think this would be the cell balance voltage.


The OCV "feature" is a nuisance but is not a show stopper and can easily be circumvented. A battery without a low charging threshold will still have the same problem because the charge current will taper off to virtually zero as an LFP battery has virtually no "leakage" current. To the GV5 a disconnected battery looks exactly the same as a battery that will not accept any more current.

Simon

Okay, I think I finally understand at least a sketchy, high level picture of how LFP charging works, and how the GV-5 interacts with LFP batteries.

How can the GV-5's OCV problem be easily circumvented? Boz at Genasun suggested that I look into a TVS, but I'm not dealing with transient voltage spikes. Is there some type of fast acting, weather-proof-able circuit that can keep the max load terminal voltage at or below 15V without fail, without adding significant size or weight and without drawing significant current from the battery?

Edit: Would something like this work? If, as described, it adds noise, and a capacitor wouldn't completely eliminate the noise, this might not be good for a ham radio application. On the other hand, the benefit of preventing damaging voltage would outweigh the inconvenience of noise as long as the noise only happens when the circuit is actively limiting high voltage.

Edit 2: According to Bioenno, the battery balances at any voltage as it's being charged or discharged. So I should be able to save money and go with a default 14.2V GV-5.

[karrak]

If no cell reaches the balancing voltage at 14.2V, in either of those batteries, then the GV-5 has been eliminated. No, wait, it just occurs to me that a GV-5 custom programmed at 14.6V would probably work. That's Bioenno's recommended charging voltage, which I think means that balancing should occur.

You are right that if you can get a GV-5 programmed for 14.6 volts that all the cells should be balanced on each charge if the balance voltage is 3.65V. As 3.65 volts is the maximum voltage that you should charge a LFP cell to I would think this would be the cell balance voltage.

I think the GV-5's OCV "feature" is a major drawback. However, it seems to me that it might work seamlessly (i.e., never go to OCV) with a battery that has no low charging threshold, and that has a HVD that protects against only catastrophic over voltage.

The OCV "feature" is a nuisance but is not a show stopper and can easily be circumvented. A battery without a low charging threshold will still have the same problem because the charge current will taper off to virtually zero as an LFP battery has virtually no "leakage" current. To the GV5 a disconnected battery looks exactly the same as a battery that will not accept any more current.

Simon

[Dave C]

The CV voltage is set by the solar controller not the battery. The GV5 sets the CV voltage at 14.2 volts. The lower the CV, the slower that the current will taper off. Within reason, it doesn't matter how fast the solar controller has to reduce the charge current. It is hard for the solar controller to go from supplying maximum current to zero current instantaneously without the voltage momentarily going higher than is should. With this fairly basic setup and with the low power involved I don't think it would be too difficult to design the charge controller to make sure that the voltage didn't go too high.

Can you give more details about the sequence of events, what was the max current, was it sunny all the time that the battery was being charged, what was the battery voltage when you started charging the battery, how long did it take before the voltage climbed to 14.2 volts and exactly how long did it stay at 14.2 volts and exactly how fast did the current taper off when the voltage got to 14.2 volts.

You as a user don't need to know this information unless you are interested in how the charging process works. To try and work out what happened I need this information.
Simon

I am very interested to know how the charging process works. Unfortunately, back when I had a GV-5 I didn't record any of the information you requested, and I won't purchase another until I'm reasonably comfortable with the answers to my pre-sale questions.

It turns out that it's not a 14.6V HVD at which the BMS ends charging; rather, it disconnects at 0.02C. Again, not sure why it took so long to arrive at this information.

I asked Bioenno and K2 about the voltage at which balancing occurs, but have received no answers yet.

If no cell reaches the balancing voltage at 14.2V, in either of those batteries, then the GV-5 has been eliminated. No, wait, it just occurs to me that a GV-5 custom programmed at 14.6V would probably work. That's Bioenno's recommended charging voltage, which I think means that balancing should occur.

I think the GV-5's OCV "feature" is a major drawback. However, it seems to me that it might work seamlessly (i.e., never go to OCV) with a battery that has no low charging threshold, and that has a HVD that protects against only catastrophic over voltage. Not sure how I would find one, given that specs at this level of detail seem to be available only through persistent digging. I suppose I could email a number of manufacturers/distributors with my requirements and ask them if they have a battery that matches.

[karrak]

I meant to say, "the battery is not yet at it's specified CV voltage of 14.6V." In other words, since the battery is not yet fully charged when all is at 14.2V, does the controller have to work harder or faster to bring the current down due to the battery still looking for full charging current? Maybe my question is meaningless. I'm just grasping for a reason why the controller would back the current down from max current to < 90mA so quickly.

The CV voltage is set by the solar controller not the battery. The GV5 sets the CV voltage at 14.2 volts. The lower the CV, the slower that the current will taper off. Within reason, it doesn't matter how fast the solar controller has to reduce the charge current. It is hard for the solar controller to go from supplying maximum current to zero current instantaneously without the voltage momentarily going higher than is should. With this fairly basic setup and with the low power involved I don't think it would be too difficult to design the charge controller to make sure that the voltage didn't go too high.

Can you give more details about the sequence of events, what was the max current, was it sunny all the time that the battery was being charged, what was the battery voltage when you started charging the battery, how long did it take before the voltage climbed to 14.2 volts and exactly how long did it stay at 14.2 volts and exactly how fast did the current taper off when the voltage got to 14.2 volts.

You as a user don't need to know this information unless you are interested in how the charging process works. To try and work out what happened I need this information.

Simon

[Mike90250]

I meant to say, "the battery is not yet at it's specified CV voltage of 14.6V." In other words, since the battery is not yet fully charged when all is at 14.2V, does the controller have to work harder or faster to bring the current down due to the battery still looking for full charging current? Maybe my question is meaningless. I'm just grasping for a reason why the controller would back the current down from max current to < 90mA so quickly.

The battery has high internal resistance

[Dave C]

All not being at CV means is that the controller is supplying as much current to the battery as the solar panel will supply. When at CV means that the controller is limiting the amount of current to the battery to keep the battery voltage constant at say 14.2V. The MPPT controller is in CV mode when it is cutting the current down.

I meant to say, "the battery is not yet at it's specified CV voltage of 14.6V." In other words, since the battery is not yet fully charged when all is at 14.2V, does the controller have to work harder or faster to bring the current down due to the battery still looking for full charging current? Maybe my question is meaningless. I'm just grasping for a reason why the controller would back the current down from max current to < 90mA so quickly.

[karrak]

Since my last post, I learned that the battery has a low current threshold of 90mA. Based on that, my understanding was that at its CV voltage, the controller was reducing current past that threshold, causing the BMS to disconnect. Then the overvoltage occurred, and the BMS wouldn't close until the panel was removed. Both manufacturers seemed to think that this would explain what happened. But your point about timing makes that conclusion questionable, and your point about balancing raise a new question: At what point does the battery's internal balancing occur? I know that for at least some batteries, balancing occurs at full charge. If that's the case with my battery, then I wonder if balancing is happening at all when the controller ends charging at 14.2V. Another question for Kevin at Bioenno...

It is all becoming clearer now, so I was right thinking that the Bioenno disconnects the battery when it thinks the battery is full. The only way to really check if it is the 90mA threshold or the battery balance would be to log the battery current as it is charging and have a look at the plot. From your point of view it doesn't matter which it is. The issue is what the MPPT controller does when the BMS disconnects the battery from the outside world.

You are right that the point at which the internal balancing occurs is of interest. The balancing occurs when aany cell reaches a preset voltage. If the cell voltage does not reach this voltage no balancing will occur. If your overall charge voltage does not cause any of the cells to go above the balancing voltage no balancing will occur. There is nothing wrong with this as long as all the cells are fully charged at the charging voltage you are using.

Ideally the MPPT controller should just keep the voltage at 14.2V and that should be the end of it. If the MPPT controller does let the voltage rise to the solar panel voltage it also doesn't matter if the BMS can handle the higher voltage. That is why I wanted you to find out the maximum voltage that the Bioenno battery could handle. If it can't handle the voltage it is very easy to keep the output from the GV5 at 14.2 volts by hooking up a small load to the output of the GV5. Putting a load onto the GV5 to solve the problem of the voltage going too high is about the only thing that Sunking and I agree on.

I'm beginning to think that it doesn't make sense to try to mix and match components from different manufacturers that have not been tested and approved for use with one another. If this is correct then I have two problems: Genasun doesn't market any small batteries, and Bioenno's MPPT controller is too big and heavy.

It would be a good start if the manufacturers gave you a list of equipment that their equipment should work with. Also more technical information like Absolute maximum voltages and minimum load currents and other specification that an engineer could look at and work out whether or not the different equipment would work together would also be nice.

This is certainly not territory for a sane non-engineer.

I hope that this has not been too truamatic for you

Edit: regarding timing of the current drop off: would the fact that the battery is not yet at CV cause the controller to have to "work harder" or faster to get the current down?

All not being at CV means is that the controller is supplying as much current to the battery as the solar panel will supply. When at CV means that the controller is limiting the amount of current to the battery to keep the battery voltage constant at say 14.2V. The MPPT controller is in CV mode when it is cutting the current down.

Simon

[Dave C]

Since my last post, I learned that the battery has a low current threshold of 90mA. Based on that, my understanding was that at its CV voltage, the controller was reducing current past that threshold, causing the BMS to disconnect. Then the overvoltage occurred, and the BMS wouldn't close until the panel was removed. Both manufacturers seemed to think that this would explain what happened. But your point about timing makes that conclusion questionable, and your point about balancing raise a new question: At what point does the battery's internal balancing occur? I know that for at least some batteries, balancing occurs at full charge. If that's the case with my battery, then I wonder if balancing is happening at all when the controller ends charging at 14.2V. Another question for Kevin at Bioenno...

I'm beginning to think that it doesn't make sense to try to mix and match components from different manufacturers that have not been tested and approved for use with one another. If this is correct then I have two problems: Genasun doesn't market any small batteries, and Bioenno's MPPT controller is too big and heavy.

Actually, at one point, Genasun recommended a K2 battery, but not the one that's nearer my weight requirement (although that one adds a full pound to my pack). I'll reach out to Mark at K2 with some additional quesitons.

From a weight perspective, it would be a toss-up between the Bioenno MPPT controller and the K2 battery. The K2's dimensions are a better fit. The saga continues. This is certainly not territory for a sane non-engineer.

Edit: regarding timing of the current drop off: would the fact that the battery is not yet at CV cause the controller to have to "work harder" or faster to get the current down?

[karrak]

Simon you are full of crap. You can look the info up and see it for yourself. A varible Power Supply can go to 0% if designed to go mili-volts, But for a solar controller on a battery has no need to go to zero and will always bbe in a state of On to some degree. You can read it right out of Fairchild and TI application notes.

Wrong yet again, if the output voltage from the solar controller is greater than it should be the solar controller should disconnect the solar panels from the output (0% modulation) until the voltage drops down to what it should be. If the solar controller control electronics is being powered from the output they will supply the necessary load to reduce the voltage. No external load or battery is required. When the voltage drops to the correct voltage the controller should start switching through just enough current from the solar panels to supply the control electronics and the battery/load to maintain the correct voltage. It there is no battery/load, the current will be the amount that the control electronics needs.

Geez you dunno chit about electronics or batteries. Why the hell do you think controllers cannot operate without a battery.

You obviously don't take any notice of what other people say do you?

On MPPT controllers, the housekeeping supply (the power supply that provides power to run the oscillator, gate drive, feedback amp, processor etc) is powered from the SECONDARY or load side. This is because the power from the PRIMARY or line side goes away sometimes (i.e. at night) whereas the power on the secondary side (the battery) never goes away while the system is commissioned.

That is why MPPT controllers can't work without a battery.

If they could [operate without a battery] you could bypass the battery and operate directly from the controller if it worked the way you you imagined. .

If you run the control electronics from the solar panel as the GV-5 does, you could do just that. Only problem is that solar power is an unreliable supply. Lets say you have a 200W panel connected via a solar controller to a 100W 12V light globe. In full sunlight the panel supplies 200W and the solar controller only lets 100W through to the light to maintain the voltage at 12 volts. If a big cloud comes between the sun and the solar panel the power delivered from the panel to the solar controller could easily drop to say 50W. The controller now cannot supply the 100W required by the light and you get a brownout. It is not the controller that needs the battery to work, the battery is there to make sure that the load has a reliable power supply.

Simon

[karrak]

Then what did the Genasun rep mean when he said "Yes the Genasun will hold the battery down at 14.2V. All regulators will have this function to prevent over charge of the battery"? To me, that sounds like the controller is designed to prevent the voltage from rising above 14.2V. Are you saying that he was mistaken, or that I'm misinterpreting him? The PCM doesn't come into play until the battery's voltage has risen to the cutoff, right?

I don't think you are misinterpreting him. I think either he is wrong or the unit you had is faulty or maybe it is a Pb unit for use with Lead Acid batteries which may operate differently to the Li unit.

From the information I have I would think the PCM will come into operation if the voltage on any of the four cells that make up the battery go outside the range 2.5V to 3.65V. This would occur under the following conditions.

• The battery charger supplying more than 14.6V
• The output voltage being under 10.0V under load i.e. flat battery
• The battery being out of balance and one of the individual cells going outside the range 2.5V to 3.65V before the others.

Simon

[karrak]

When you say "slowly," what kind of time frame would be typical? I observed the GV-5 first indicating full charge for at most a couple of seconds, and then showing high battery voltage.

If the battery was charging normally I would expect the reduction in current to occur over minutes or tens of minutes. If the drop in current from 200mA to 0mA occurred in a few seconds I would think it was caused by the battery being internally out of balance and the BMS disconnecting the battery from the outside world because one of the cells voltage went to high.

I forget, do you speak from experience with the GV-5?

No, I have never had anything to do with a GV-5. My comments were are my educated guess as to why the GV-5 would work with Lead Acid batteries if it will not maintain a regulated 14.2V with little or no load current.

Do you think any of the 5 scenarios in my last post could be valid?

From the information I have I think:-
1. Is correct, but given the time frame of the drop in current is not the reason for the battery BMS disconnecting the battery
2. Is not correct, the controllers voltage is 14.2V but it can't maintain regulation with little or no load current
3.Is not correct, the battery's BMS HVD is around 14.6V
4. Is not correct, if the controller is regulating its output, the output will be 14.2V, The battery's HVD is around 14.6V
5. This is unlikely.
Simon

[Sunking]

Yes, I understand this. There was a disconnect involved. I was responding to Simon's statement that at CV the current goes slowly from 200mA to 0A, and I was asking him if he considers a couple of seconds slow in the realm of electronics.

An eternity, but it depends on the size of filter caps. Your observations is telling you Simmon does not have a clue.

[Dave C]

That is when the battery disconnected, and the controller voltage pumped up to 18 volts with no place for the charge to go.

It is impossible for your controller to go from 14 volts to 18 volts in a hew seconds if a battery is connected. To do that the current would have to be a few hundred amps, and the battery would explode. Only way for that to happen is the BATTERY DISCONNECTS. You can prove it with a volt meter.

Yes, I understand this. There was a disconnect involved. I was responding to Simon's statement that at CV the current goes slowly from 200mA to 0A, and I was asking him if he considers a couple of seconds slow in the realm of electronics.

Edit: Actually, I meant to say that I was asking him if he considers a couple of seconds to be slow in the realm of an LFP battery at CV going from 200mA to 0A. I'll have to wait until he gets back.

[Sunking]

I observed the GV-5 first indicating full charge for at most a couple of seconds, and then showing high battery voltage.

That is when the battery disconnected, and the controller voltage pumped up to 18 volts with no place for the charge to go.

It is impossible for your controller to go from 14 volts to 18 volts in a hew seconds if a battery is connected. To do that the current would have to be a few hundred amps, and the battery would explode. Only way for that to happen is the BATTERY DISCONNECTS. You can prove it with a volt meter.