shunt charge controllers

split off from this post in the MPPT vs PWM test thread:

tx2sturgis said:

BTW, shunt controllers apply full panel voltage and current up to the point that the shunt is applied, then they short the panel output completely. Simple: ON (full panel voltage to the battery) or SHORTED (panel voltage clamped).

Click to expand...

I have a couple questions -- trying to disentagle some ideas in my head.

Is "full panel voltage to the battery" used here to distinguish it from PWM switching?  Would also be accurate to say something like "panel voltage applied continually to the battery"?  I ask this in case an onlooker thought the battery was being charged at the panel's Voc.  UNderstanding it will also keep my head straight[er].

In a Bulk situation* is there any practical difference between a shunt's ON and PWM's 100% ON cycle?  I don't think so but maybe I am missing something.

In a Float situation* is there any meaningful distinction between shunt's ON/OFF cycles and PWM's switching when both are configged to the same setpoint?  Other than magnitudes of speed, I mean.   I think PWM would be easier to run at an exact voltage since it's not cycling between an OFF and ON setpoint.




* disregarding the fact that shunts are typically primitive and don't have staged charging in the normal sense

If you have a wind powered source the load on the generator keeps the turbine from overspeeding.  There will be a significant power dissipation heating the source panel or generator.

Otherwise I don't see any significant diffetence either

Ah! Fun stuff.

You have stated it correctly. 

But digging deeper:

We're talking about a balanced, normal, off-the-shelf, 100 watt or 200 watt system with 12v panels, and normal sized lead-acid (flooded or AGM) batteries we would find in a vehicle. 

(if we were talking about a 500 watt, 12 volt panel, trying to charge a tiny 12v smoke detector battery or the like, then this doesn't apply...nor does it apply to higher voltage 24v or 36v panels or multiple 12v panels in series)

But it's easier to wrap our mind around this concept when you remember that 12v panels can charge 12v batteries directly...

Those panels have to have a (VOC) voltage higher than 12 volts to actually charge a 12v battery, at least 14 to maybe 20 volts or so. 

But remember that solar panels themselves have no ability to regulate their output when a load is applied. Charging a partially depleted battery IS a big load and the panel voltage will be pulled down to what the battery voltage is, but of course, this value will rise as the battery accepts a charge. 

Yes, PWM is essentially a really fast, shunt type controller, (but without shunting the output of the panel) but it adds the smarts to change duty cycle when needed by the battery.

In fact, one of my PWM controllers has a setting to reduce 'tel-com' interference, which essentially slows the PWM on-off cycle to something much slower, more like a simple shunt controller. 

There are some power losses, and voltage drops of a few tenths of a volt, but the PWM or shunt controllers apply full panel voltage, or something close to it, across the battery terminals, at a duty cycle (fast for PWM, slow for shunt) that keeps the battery happy, and the controller monitors the battery voltage and then adjusts the PWM duty cycle accordingly, or in the case of a shunt, it clamps the panel output. 

MPPT controllers as we all know, are a newer technology, and do something much different.

I use a shunt controller on my 50 watt portable panel,



it has temp compensation and works well.  highdesertranger

tx2sturgis said:

Ah! Fun stuff.

Click to expand...

I figured you'd like it. 
What got me thinking about this was that PWM controllers seem to fall into two varieties:

• inexpensive without user-configurable setpoints
• expensive with user-configurable setpoints

But I have an $8 (10A) shunt and $12 (20A) shunt in my test area that have user-configurable connect/disconnect setpoints for charging and for LOAD. 

Which made me think "I'd rather have a configurable shunt than unconfigurable PWM." 

In fact, one of my PWM controllers has a setting to reduce 'tel-com' interference, which essentially slows the PWM on-off cycle to something much slower, more like a simple shunt controller.

Click to expand...

That's really clever.  

There are some power losses, and voltage drops of a few tenths of a volt, but the PWM or shunt controllers apply full panel voltage, or something close to it, across the battery terminals, at a duty cycle (fast for PWM, slow for shunt) that keeps the battery happy, and the controller monitors the battery voltage and then adjusts the PWM duty cycle accordingly, or in the case of a shunt, it clamps the panel output.

Click to expand...

That's an excellent summary.  I'm going to integrate it into the wiki for the benefit of newcomers.  With attribution, of course.

Just for grins I'm doing a 'test' with a 120 watt panel (25 VOC) on 2 different lead acid batteries. 

This is not the first time I've done this type of thing, but it's the first time I will be doing it with photos and video and put the results on the web.

I do have photos of the meter readouts with an already fully charged battery since I don't have any batteries around that aren't kept on maintainers or controllers. I am now in the process of placing a load on this battery for awhile to get the voltage and SoC down to a lower value. 

But here are preliminary readings that an already fully charged deep-cycle battery sees with a 120 w, 12v, (25 VOC) panel connected and no controller at all, and this 'test' or 'demonstration' is still on-going.

Panel VOC:




Battery at charged voltage before panel is connected:




Battery voltage after connection to panel. Red DMM is reading voltage, and the yellow DMM is reading amperage flow into the battery:



Professional driver on closed course...or some other disclaimer.

Just to state the obvious...again, this is a brief test, I am NOT recommending any person hook up any solar panel to any battery without a solar controller. 

But like, Bob, I have magic powers and doing this type of thing won't hurt me or my stuff. 

frater secessus said:

Which made me think "I'd rather have a configurable shunt than unconfigurable PWM." 

Click to expand...

I had not seen a shunt controller with configurable set points. Do you have a link or brand name and model?

As I type this I have a 3 amp load on that deep cycle battery, and will let it run for awhile, until the voltage drops to about 12.2 or so.

Of course it wont be deeply discharged, but will be low enough to do another test.

In the case of a panel connected with no controller in line, I only want to provide some test results, and not trying to say I would run a solar setup in this manner for any length of time without supervision.

Once the battery is charged, then run-away 'equalization' would begin, or in other words, overcharging. We dont want that!

Sorry for the delay, I had to interrupt my testing today, I got a call from the bank where my van was previously financed, I was waiting on notarized papers to verify title...then made a trip out of town to get some business taken care of at the bank and the DMV regional office 50 miles away. 


Yada yada yada...unrelated details....


But anyway, I got back and resumed the load on the deep cycle battery so I will have it at a much lower starting voltage in the morning when the sun comes up and I can resume the experiment.

tx2sturgis said:

I had not seen a shunt controller with configurable set points. Do you have a link or brand name and model?

Click to expand...

The same design is sold under a bunch of names, but they have three buttons and an LCD screen.  Two USB ports which do work.    Often labeled PWM but they are Shunt.  Trebor has played with similar models.
Bought a 10A one for $8 shipped on the slow boat.  Cheapest I see right now is $9.10 shipped.  

The first button cycles through the display.  Long-pressing it on a setpoint (wrench icon appears at bottom center) allows you to change it.  2nd and 3rd buttons are UP/DOWN.    Pressing the first button after adjustment locks the setting in.
BTW, there are only three setpoints not four.  Charging OFF, LOAD LVD, LOAD LVR. 

In the case of a panel connected with no controller in line, I only want to provide some test results, and not trying to say I would run a solar setup in this manner for any length of time without supervision.

Once the battery is charged, then run-away 'equalization' would begin, or in other words, overcharging. We dont want that!

Click to expand...

I love reading about other folks' experiments.  Keep us posted!

frater secessus said:

The same design is sold under a bunch of names, but they have three buttons and an LCD screen.  Two USB ports which do work.    Often labeled PWM but they are Shunt.  Trebor has played with similar models.

Click to expand...

That picture is exactly like the controller I have.  

I don't understand what you mean when you use the term PWM and the word shunt.  How do you decide that a regulator is shunt?  How do you decide that a regulator is PWM?

In my experience PWM is used to mean Pulse Width Modulation.  That is where a circuit is switched on and off varying the ratio of on time to off time.  The alternative is a continuously variable linear circuit.  Generally PWM means fixed frequency but the term is also commonly applied to variable frequency switching that varies the on / off ratio.

Shunt in voltage a regulator sense is where the regulator draws current bypassing the intended load as opposed to a series limiting device.  

Either series or shunt regulators can be switched or continuous linear.  Continuous linear circuitry will be bulky, expensive, and will produce a lot of heat.  Switched circuitry will be cheaper, smaller, and produce less heat.  

My solar charge controller uses a series pass switch to turn on and off the connection between the source, the solar panel, and the load, the battery being charged.  When the battery is full its voltage is high and the solar panel gets disconnected.  The solar panel voltage goes up to 21 volts, current drops to zero.  If it were a shunt controller it would reduce the charging current to the battery by applying a load that doesn't go to the battery being charged.   The panel voltage would be much lower than 21 volts.  

In the realm of solar charge controllers I am not aware of any that are contiuous linear devices.  The size and cost is prohibitive.

Yes I've got a box of ASCs somewhere, AF version so adjustable voltage as well as temp comp, unfortunately discontinued.

Paid maybe $15-20 each, perfect for 1:1 per panel, have to dig them up sometime.

Satisfied with some testing, taking pictures and video along the way, of the PV panel hooked up straight to the partially discharged battery for more than an hour, I moved on to testing the new Renogy PWM Adventurer that was on sale and I recently ordered.

The panel was putting out about 4 amps to the battery without the controller in the way.

VOC of this panel is about 25 volts, but of course when hooked up to a load, the voltage comes down to the battery voltage, which will rise as the battery accepts the charging current.

This is the battery voltage after an hour or so:



Here is the new controller hooked up, still about 4 amps, of course:







Interstingly, the controller shows a PV of 25 volts when first attached to the panel, then that reading dropped to 13 when the charging began. This helps verify that the controller is also pulling the panel voltage down, by applying a good load to the panel. 



I have not tested my Victron MPPT controller...I'm curious what it will show when hooked up to my set of 2 100 watt panels in series. 

I've used it several times, but have not actually done any testing or comparisons from one type of controller to the other.

Again, disclaimers are needed:

The testing I have done here was only for information...dont do this at home! Do NOT hook up a panel directly to a battery and leave it unattended or overnight.

Always use a solar controller of the proper size and type between your solar panels and your batteries.

(just to cover the boss's behind)

Next....

Trebor English said:

In my experience PWM is used to mean Pulse Width Modulation.  That is where a circuit is switched on and off varying the ratio of on time to off time.

Click to expand...

That is what I mean.

My solar charge controller uses a series pass switch to turn on and off the connection between the source, the solar panel, and the load, the battery being charged.  When the battery is full its voltage is high and the solar panel gets disconnected.  The solar panel voltage goes up to 21 volts, current drops to zero. 

Click to expand...

That's what I was calling a shunt.  I think the term gets used in a solar context because the end effect to the bank is like a shunt;  charging on-off at a particular setpoint as if shunting had occured.  I admit the term is not accurate.

So what do we call a non-PWM on-off charge controller?

If it were a shunt controller it would reduce the charging current to the battery by applying a load that doesn't go to the battery being charged.   The panel voltage would be much lower than 21 volts. 

Click to expand...

As with hydro or wind, I think.  I always thought of that as a "dump load" controller but shunt appears to be the correct term there.

Yeah a true shunt controller actually shorts the input from the solar panel, once the battery has reached the set point. It's low tech, but it works. 

Of course PWM just rapidly turns the input from the panel on and off. Very rapidly. Dozens or hundreds of times per second.

This WAS the high-tech way of controlling a solar charging system until MPPT came along.

At one of the solar seminars I attended in Quartzsite, the guy from Blue Sky Energy explained that MPPT controllers tend to dump more and more of the input power as heat, as the series array of panels gets farther and farther above the battery voltage. 

In other words, an MPPT controller with 4 panels in series (about 80-100 volts) will dump more power as heat than the same controller with 4 panels in series/parallel, (about 40-50 volts) assuming the same 12 volt battery bank...

Learned something new that day!

I have posted before that the fact that my three 250w panels in series running at 105 volts cost me 5.5a over running them in parallel.

Yeah it makes sense, given that many of the higher wattage and voltage MPPT controllers have rather large heat sinks...

frater secessus said:

So what do we call a non-PWM on-off charge controller?

Click to expand...

On another thread I posted a link to a 1980s diy Mother Earth News solar panel controller that uses a relay to turn on and off the solar panel to battery connection.  Typically it turns on in the morning and off in the afternoon.  An early vendor of PWM products gave it the name "on off controller" as it previously had no name.  You could call it "Original Recipe" but Colonel Sanders already had that trademarked.  

The one cycle per day on off controller will shut off when the battery gets to the upper limit.  A PWM controller will do the same tens or hundreds of times per second to have the effect of maintaining a continuous constant charging voltage at the battery for an extended period of time.  There is a class of controllers that are between the two.  The frequency of the pulses is much higher than daily or per cloud passing but not nearly high enough to give a constant voltage.  I would describe them by comparing a light dimmer to flipping the light switch on and off at an aggravating speed.  The on off action will cause the battery to continue to charge over an extended period of time.

That may actually make this topic very applicable to charging LFP banks!

The ideal scenario is no stages, not even two.

The CC-CV "transition" with lead charging (Bulk to Absorb) is actually LFP's end point.

In other words, start charging, trying to feed say 13.75V (a bit lower since low current, as in below 100's of amps)

When the combined controller/bank voltage reaches that setpoint, just stop charging, no Absorb, no Float.

Ideally maybe a restart if bank/load voltage drops below say 12.8V.

Are there non-MPPT units out there "primitive enough" to do this?

The tricky part I guess is user-adjustable charging V setpoint.

The ASC "AF" variant is no longer made, may otherwise be a candidate.

John61CT said:

That may actually make this topic very applicable to charging LFP banks!
The ideal scenario is no stages, not even two.

Click to expand...

The controller I have, like the one in the photo in post number 9, does nothing fancy.  I have lead acid so I set it to 14.6 volts in winter, 14.3 in summer.  When the voltage at the controller gets to that level it becomes an open circuit.  When the voltage at the controller drops below that level it turns on.  If you set it to 13.6 your LiFePo battery would get to about that voltage then it's off.  As loads turn on and the voltage drops it stays on more.  

It has no stages.  The solar panel limits the current.  If the battery is low enough it just stays on.  When the voltage at the controller gets high enough it shuts off.  The on off frequency depends on the solar panel, the battery, and the wire resistance.  The controller voltage will be higher than the battery voltage depending on the charge current and wire resistance.  E=I×R.  In my lead acid battery application the "absorb stage", a function of the lead acid battery,  terminates not at a low tapering charge rate but when the sun goes down.  The instructions that came with the controller call it "float" rather than constant voltage, absorb, or whatever other people might call it.  Functionally, it is the peak voltage. 

Since it has a low voltage disconnect load output with settable on and off points you could get a resistor like a car headlight bulb and have it turn on the headlight at 13.7 volts and turn if off at 13.5 volts.  That would really be a shunt regulator, draining off the excess current, good for one 100 watt solar panel.  You could use that instead to run a motor to tilt the solar panels to a non-productive angle.  That way a $10 controller could control a multi-kiliwowatt solar farm.  You could use it to turn on / off an automotive style relay coil, 1/4 amp for the coil, 40 amps contact capacity.  Using the normally closed contact for charging would mean no wasted coil current until the battery is full.

For $10 you could get one and have a fine toy to play with.  You could buy a bunch of 2n3055 transistors and make a contiuous linear, no pulses, no emi, shunt or series pass regulator but that will cost way more than $10.

> I have lead acid so I set it to 14.6 volts in winter, 14.3 in summer.

For temp compensation or for some usage that varies with the seasons?


> For $10 you could get one and have a fine toy to play with.

For sure.