solar pwm battery charger

Thread Starter

denison

Joined Oct 13, 2018
59
Hi, have just designed and built a pwm battery charger to charge from a 20w 1amp max solar panel. before I purchase a new battery I need to know something so I don't damage the battery.
My intention is to connect the battery directly to the pwm output of my circuit. The specs I entered on ltspice for a battery with a starting voltage of 13.18v and internal resistance of 1.66 ohms. . At 20v the average of the pwm was 14.34v as set by me in the design. Current draw 700ma.
At 16v again of course 14.34v and 700ma. Great efficiency.
Will the battery be damaged by a direct connection to the pwm output which is at a frequency of 576 cps? The maximum of the pulse is 20v or lower depending on the output of the solar panel. I get the same results as above over the range of all output voltages down to 14.34 then I get approx. the solar panel voltage.
Considering that the max cycle voltage of the battery I intend to get is 15v will the battery be damaged by the high pulses over 15v?
 

MrChips

Joined Oct 2, 2009
19,933
We will have to assume that the battery type is lead acid. Exactly what is the battery type, flooded lead acid, SLAB, gel, AGM? You did not state what is the duty cycle of the PWM signal. Is there a reason for using PWM?

The charging voltage will be determined by the battery itself. You need to monitor the battery voltage. When the voltage reaches 14.8V, the charging voltage should be reduced to 13.5V for float charging.

I would highly recommend that you monitor the battery temperature and disconnect the charger if the temperature rises above 50°C.
 

TeeKay6

Joined Apr 20, 2019
569
Hi, have just designed and built a pwm battery charger to charge from a 20w 1amp max solar panel. before I purchase a new battery I need to know something so I don't damage the battery.
My intention is to connect the battery directly to the pwm output of my circuit. The specs I entered on ltspice for a battery with a starting voltage of 13.18v and internal resistance of 1.66 ohms. . At 20v the average of the pwm was 14.34v as set by me in the design. Current draw 700ma.
At 16v again of course 14.34v and 700ma. Great efficiency.
Will the battery be damaged by a direct connection to the pwm output which is at a frequency of 576 cps? The maximum of the pulse is 20v or lower depending on the output of the solar panel. I get the same results as above over the range of all output voltages down to 14.34 then I get approx. the solar panel voltage.
Considering that the max cycle voltage of the battery I intend to get is 15v will the battery be damaged by the high pulses over 15v?
@denison
You may find info at this site: https://batteryuniversity.com/learn/article/charging_the_lead_acid_battery
Search the page for "ripple voltage" for (possibly) relevant info. You may be able to reduce the PWM ripple by filtering the PWM drive a bit before it enters the battery, e.g. by inserting an inductor in series with the battery--provided your PWM driver is not upset by that. A PWM pulse with a peak amplitude of 20V is a pretty strong overcharge during the brief period of each pulse. Another approach would be to insert current limiting between the charger and the battery to limit the peak current during the pulses.
 

Thread Starter

denison

Joined Oct 13, 2018
59
We will have to assume that the battery type is lead acid. Exactly what is the battery type, flooded lead acid, SLAB, gel, AGM? You did not state what is the duty cycle of the PWM signal. Is there a reason for using PWM?

The charging voltage will be determined by the battery itself. You need to monitor the battery voltage. When the voltage reaches 14.8V, the charging voltage should be reduced to 13.5V for float charging.

I would highly recommend that you monitor the battery temperature and disconnect the charger if the temperature rises above 50°C.
The battery is a sla(sealed lead acid). The duty cycle varies from 71% at 20v to 100% at 14.34v to give an output average voltage of the pwm of 14.34 over the whole range of possible solar panel voltages.
A pwm charger is much more efficient than a linear voltage regulator. If the linear regulator is set to 14v and 20v at the solar panel 6v is lost power absorbed by the regulator whereas this 20v can be used as power to charge the battery meaning less solar panels.
 

TeeKay6

Joined Apr 20, 2019
569
If the linear regulator is set to 14v and 20v at the solar panel 6v is lost power absorbed by the regulator whereas this 20v can be used as power to charge the battery meaning less solar panels.
Your argument is true only if the SLA can efficiently convert the high charging current during a 20V pulse to chemical energy that can later be retrieved (during battery discharge). Thus far in this discussion no one has produced evidence that that is true or false. There are always losses incurred during recharging. We don't know the magnitude of those losses during very high charge rate pulses. Nor do we know whether such high charge rate pulses have any deleterious effects on the battery.
 

MrChips

Joined Oct 2, 2009
19,933
PWM or no PWM, the battery voltage will not be 20V. You will be pushing higher current through the battery which may be ok for faster charging. The 6V lost still has to be absorbed by the internal resistance of the solar panels.
 

Thread Starter

denison

Joined Oct 13, 2018
59
@denison
You may find info at this site: https://batteryuniversity.com/learn/article/charging_the_lead_acid_battery
Search the page for "ripple voltage" for (possibly) relevant info. You may be able to reduce the PWM ripple by filtering the PWM drive a bit before it enters the battery, e.g. by inserting an inductor in series with the battery--provided your PWM driver is not upset by that. A PWM pulse with a peak amplitude of 20V is a pretty strong overcharge during the brief period of each pulse. Another approach would be to insert current limiting between the charger and the battery to limit the peak current during the pulses.
I agree that 20v is a strong overcharge right down to 15v which is the recommended max cycle charge for the battery. But the overcharge is only on for a very short time. The average voltage is always 14.34.
I have thought of using a rc filter and this looks promising on the ltspice simulations. A higher resistor value means less current flowing and a smaller capacitor. A smaller resistor would mean using a larger capacitor which may lead to a ridiculously large capacitor.
Another option is to use an active filter rather than a passive using an op amp.
However none of this may be necessary if I can get a definite answer as to whether the pwm directly to the battery is going to damage it.
I have connected the pwm directly to a 12v globe drawing 0.4 amps and the pwm output was still 14.34v and the globe didn't blow. Also notably no flickering of the globe. Of course the battery is an altogether different kind of load.
What do the makers of pwm controllers do? Does anybody know?
I wouldn't have any idea of what size inductor to use or what type of current limiting device to use.
 

TeeKay6

Joined Apr 20, 2019
569
The battery is a sla(sealed lead acid). The duty cycle varies from 71% at 20v to 100% at 14.34v to give an output average voltage of the pwm of 14.34 over the whole range of possible solar panel voltages.
A pwm charger is much more efficient than a linear voltage regulator. If the linear regulator is set to 14v and 20v at the solar panel 6v is lost power absorbed by the regulator whereas this 20v can be used as power to charge the battery meaning less solar panels.
Also missing from this discussion is the fact that a solar panel is not a voltage source. If you attempt to load it heavily, it's output voltage will simply drop and its efficiency as a solar-to-electric converter will plummet.
 

Thread Starter

denison

Joined Oct 13, 2018
59
Your argument is true only if the SLA can efficiently convert the high charging current during a 20V pulse to chemical energy that can later be retrieved (during battery discharge). Thus far in this discussion no one has produced evidence that that is true or false. There are always losses incurred during recharging. We don't know the magnitude of those losses during very high charge rate pulses. Nor do we know whether such high charge rate pulses have any deleterious effects on the battery.
Yes that is the question. Will the battery be damaged? Only those that have tried this will know. I may take a risk and test it. It beats adding filters and the like.
 

TeeKay6

Joined Apr 20, 2019
569
Yes that is the question. Will the battery be damaged? Only those that have tried this will know. I may take a risk and test it. It beats adding filters and the like.
If you modeled the solar panel as a voltage source in your LTspice simulation, you will find real-world results differ considerably from the simulation.
 

nsaspook

Joined Aug 27, 2009
6,743
I have an old C40 PWM charge controller connected to a several hundred watt solar array. No filters on the PWM output and it's been charging lead-acid batteries for at least a decade. Battery charging is a diffusion process that's mechanically slow when compared the the PWM duty cycle so at reasonable source peak currents for the size of battery (it is possible to pump electrical energy into the battery faster than the chemical process can react to it, this is bad :rolleyes:) it will average the energy for the duration of the pwm cycle.




A old scope trace (bottom) of battery charge current pulses during C40 PWM voltage set-point control.
 
Last edited:

Thread Starter

denison

Joined Oct 13, 2018
59
Also missing from this discussion is the fact that a solar panel is not a voltage source. If you attempt to load it heavily, it's output voltage will simply drop and its efficiency as a solar-to-electric converter will plummet.
That is correct. But my load is only 0.175A which is holding a remote lock on my gate. When operated it draws 1 amp which is only for a moment then 0.175A to hold the gate locked continuously.
A 4.5Ah sla battery will be sufficient to hold the gate locked all night together with a 20w solar panel which can supply 1.12A at 18v max.
 

MrChips

Joined Oct 2, 2009
19,933
Your solar panel can only supply 20W.
Connect the panel in series with a Schottky diode to the battery. Monitor the battery voltage and temperature to prevent over charging.
 

Thread Starter

denison

Joined Oct 13, 2018
59
I have an old C40 PWM charge controller connected to a several hundred watt solar array. No filters on the PWM output and it's been charging lead-acid batteries for at least a decade. Battery charging is a diffusion process that's mechanically slow when compared the the PWM duty cycle so at reasonable source peak currents for the size of battery (it is possible to pump electrical energy into the battery faster than the chemical process can react to it, this is bad :rolleyes:) it will average the energy for the duration of the pwm cycle.




A old scope trace of battery charge current pulses during C40 PWM voltage set-point control.
 

TeeKay6

Joined Apr 20, 2019
569
That is correct. But my load is only 0.175A which is holding a remote lock on my gate. When operated it draws 1 amp which is only for a moment then 0.175A to hold the gate locked continuously.
A 4.5Ah sla battery will be sufficient to hold the gate locked all night together with a 20w solar panel which can supply 1.12A at 18v max.
@denison
"a 20w solar panel which can supply 1.12A at 18v max." Panels are normally rated at 1-sun insolation and with a load that causes operation at the MPP. Less light, less power. Operation not at MPP, less power. The data sheet for your panel should give more detail.
 

Thread Starter

denison

Joined Oct 13, 2018
59
At last a definite answer to my enquiry regarding direct connection to a battery of the pwm output of a pwm charge controller. Thank you very much nsaspook.
 

nsaspook

Joined Aug 27, 2009
6,743
At last a definite answer to my enquiry regarding direct connection to a battery of the pwm output of a pwm charge controller. Thank you very much nsaspook.
A really old scope display of panel voltage and battery current on the C40.
Battery charge current top, input panel voltage before charge controller bottom

Battery ripple voltage bottom trace.
 
Top