Solar panel, MPPT, Battery switching

[goboatingnow]

Originally Posted by thomm225

I didn't care about the PWM action, the MPPT action, or the regulator action, I simply cared about the voltage to check my batteries charge.

When you put your meter leads on the battery terminals, you get battery voltage!

When the wires attached to the battery terminals are hooked to the controller and you read the voltage there, you get battery voltage! Same on Panel side. You get panel voltage.

And on the small 20 watt panels I had hooked directly to the batteries without a controller you also get battery voltage

I guess we techs have more real world experiences with systems as compared to some engineers......this wasn't an analysis but data gathering.

I wasn't analyzing controllers but simply checking the charge on my batteries. (since I was using $10-$15 controllers)

This from a tech that almost always used an O'scope or logic analyzer to troubleshoot with....

but that's all old school now as the speeds are so fast and most everything runs on a bus.

At present though, we are switching over to VM's (virtual machines) to replace real computers ....

Not sure what's next but whatever it is it will come fast.....

Photos from around 2017 or so.

When you put a multi meter on a PV connected to a functioning pwm you are not reading panel voltage , you are simply reading a number on a digital display and they try to explain to me something that isn’t true.

Clearly you don’t appreciate what’s actually going on at s voltage level.

A pwm controller repeatedly connects a panel to the battery and then disconnects it ,sometimes many many times a second

What you multimeter is doing is simply lying to you

The point is a pwm is very inefficient controller for panels whose operating point is far away from the battery voltage

This is my point. I don’t care what your multimeter is lying to you is saying

I designed switch mode power supplies , I know this stuff from first principles.

[thomm225]

Originally Posted by goboatingnow

When you put a multi meter on a PV connected to a functioning pwm you are not reading panel voltage , you are simply reading a number on a digital display and they try to explain to me something that isn’t true.

Clearly you don’t appreciate what’s actually going on at s voltage level.

A pwm controller repeatedly connects a panel to the battery and then disconnects it ,sometimes many many times a second

What you multimeter is doing is simply lying to you

The point is a pwm is very inefficient controller for panels whose operating point is far away from the battery voltage

This is my point. I don’t care what your multimeter is lying to you is saying

I designed switch mode power supplies , I know this stuff from first principles.

Let me put this another way to see if you can understand.

First you need to know what it is I was checking which was battery charge percentage. You seem to be stuck on the controller.

I'm getting the voltage reading that I need to determine if the batteries are charged simply by using an old volt/ohm meter.

So when my old controller / regulator got to 14.4 volts on the battery side it stayed at that voltage.

If I check the voltage on the panel side of the controller at that point it would be at 14.4 volts also.

When I checked both sides hours later, the battery side would be at 14.4 volts and the panel side would be at about 20 volts or more.

(This is about the same reading I get on my new Victron 75/15 MPPT display now when it drops to Float except the Battery Voltage is about where I set it at 13.2 volts. Yesterday panel side was 19.8 volts in Float)

This information was what I needed to tell me that current flow had decreased to such a point the panels were approaching VOC because the batteries were nearing fully charged.

It was a fun experiment that lasted a few years starting at around 2012. You got to see the exact way the batteries charged.

After the panel side reached about 20-21 volts I would disconnect this controller at the batteries.

Btw this little controller/regulator still works.

[goboatingnow]

Originally Posted by thomm225

Let me put this another way to see if you can understand.

First you need to know what it is I was checking which was battery charge percentage. You seem to be stuck on the controller.

I'm getting the voltage reading that I need to determine if the batteries are charged simply by using an old volt/ohm meter.

So when my old controller / regulator got to 14.4 volts on the battery side it stayed at that voltage.

If I check the voltage on the panel side of the controller at that point it would be at 14.4 volts also.

When I checked both sides hours later, the battery side would be at 14.4 volts and the panel side would be at about 20 volts or more.

(This is about the same reading I get on my new Victron 75/15 MPPT display now when it drops to Float except the Battery Voltage is about where I set it at 13.2 volts. Yesterday panel side was 19.8 volts in Float)

This information was what I needed to tell me that current flow had decreased to such a point the panels were approaching VOC because the batteries were nearing fully charged.

It was a fun experiment that lasted a few years starting at around 2012. You got to see the exact way the batteries charged.

After the panel side reached about 20-21 volts I would disconnect this controller at the batteries.

Btw this little controller/regulator still works.

Let’s not descend into telling an EE how to suck eggs

I fully understand what you are doing , and I fully accept the readings you mention and as you state there are enough to determine full charge etc.

What you seem to ignore is my comments that pwm controllers are inherently inefficient especially where the panel voltage is significantly higher then the battery voltage

Here’s the typical PV panel I versus V graph



As you can see when the pwm connects the panel to the battery the available power is significantly reduced if the panel nominal voltage is high.( as it is in this case ). Mppt is needed to extract power from this panel( or more correctly power conversion systems are needed , wether they track Vmp or not )

Here’s the graph with the panel power added ( normalised to cell voltage )

[hzcruiser]

Originally Posted by goboatingnow

When you put a multi meter on a PV connected to a functioning pwm you are not reading panel voltage , you are simply reading a number on a digital display and they try to explain to me something that isn’t true.

Clearly you don’t appreciate what’s actually going on at s voltage level.


[...]

I designed switch mode power supplies , I know this stuff from first principles.

It seems to me as if you both have a pretty clear understanding of what you're measuring and the limitations of each.

We understand you know how to use a scope and you keep repeating pwm is (generally) inefficient. Well, it seems we all agree. I do, anyway.

Your friend thomm uses his real world tests, data gathered over the years and covering a range of different CCs. He's aware you can't "see" the millisecond pulses on a hand-held DMM, but then again, how often do you have your scope on the boat?

So why don't you just leave it at that?

Which brings up the question: if PWM is really so inefficient, why don't they implement a buck converter instead? It's not that much more expensive from a component view, or is it?

Isn't that what MPPT does, at the end of the day?

Maybe I should just draw up some schematics of a solar CC with buck converter and full battery feedback (voltage and current) to settle this?

[goboatingnow]

Originally Posted by hzcruiser

It seems to me as if you both have a pretty clear understanding of what you're measuring and the limitations of each.



We understand you know how to use a scope and you keep repeating pwm is (generally) inefficient. Well, it seems we all agree. I do, anyway.



Your friend thomm uses his real world tests, data gathered over the years and covering a range of different CCs. He's aware you can't "see" the millisecond pulses on a hand-held DMM, but then again, how often do you have your scope on the boat?



So why don't you just leave it at that?



Which brings up the question: if PWM is really so inefficient, why don't they implement a buck converter instead? It's not that much more expensive from a component view, or is it?



Isn't that what MPPT does, at the end of the day?



Maybe I should just draw up some schematics of a solar CC with buck converter and full battery feedback (voltage and current) to settle this?

My point stemmed from the use of very cheap PWM controllers. I made the point these CAN be very inefficient.

You are right you would be better theses days simply getting a buck convertor or even buck boost and manually adjusting the constant current to ensure the PV panel operated near the Vmp , of course varying irradiance is the problem !

Ultimately mppt is a CC converter with a movable CC point based on Vmp

I carry a scope on board !

Cheap pwm is a very crude way of controlling the charge process. It works of course but it can be very inefficient

[thomm225]

Originally Posted by goboatingnow

Let’s not descend into telling an EE how to suck eggs

I fully understand what you are doing , and I fully accept the readings you mention and as you state there are enough to determine full charge etc.

What you seem to ignore is my comments that pwm controllers are inherently inefficient especially where the panel voltage is significantly higher then the battery voltage

Here’s the typical PV panel I versus V graph

Attachment 260243

As you can see when the pwm connects the panel to the battery the available power is significantly reduced if the panel nominal voltage is high.( as it is in this case ). Mppt is needed to extract power from this panel( or more correctly power conversion systems are needed , wether they track Vmp or not )

Here’s the graph with the panel power added ( normalised to cell voltage )
Attachment 260244

Thanks for that info, but it's common knowledge.

Most everyone that's messing around (doing DIY projects) with their solar know that PWM controllers are not as efficient as MPPTs.

At that time like now, I wasn't sure when I'd buy a larger boat so didn't want to put too much into this one's solar setup.

What I did then was I increased my onboard battery ah's from the $80 Walmart starter battery which had been for my old diesel by putting a so-called 12 volt deep cycle battery ($95) in parallel with it.

Then got rid of my 15 watt panel and bought a 65 watt panel for $76.

Then bought the controller/regulator for $10 and ran with this setup for 4-5 years.

And it went from there.

Skip forward to 2022 and now I still have the 65 watt panel plus a 50 watt panel hooked to my new Victron 75/15 MPPT Controller.

I won't be hooking up the 20 watt panels to the batteries directly or with a PWM controller.

My batteries now are two huge 220 ah GC2 Golf Cart true deep cycle batteries that don't seem to notice the small power requirements on my boat.

[thomm225]

Victron MPPT (Bluetooth) finally installed and attached to Laptop. (w/VE.DIRECT TO USB INTERFACE CABLE)

Notice figure taking picture of laptop display in background! (46 Watt display)

Just took my cycling helmet off after a ride to the boat. The helmet helped with my hairdo for the day.

[Paul Elliott]

Originally Posted by thomm225

When the wires attached to the battery terminals are hooked to the controller and you read the voltage there, you get battery voltage! Same on Panel side. You get panel voltage.

I think this is the claim that is causing the disagreement. When a PWM controller is rapidly cycling on and off, a meter on the panel connections will display an averaged value somewhere between the panel Voc and the battery voltage. Not exactly the same thing as "the panel voltage", but probably still informative.

Other than that, I think we all agree: PWM is inefficient when the panel MPP voltage is significantly higher than the battery voltage. This may or may not matter -- it depends on the use-case.

FWIW, I'm an EE who loves solder-smoke.

[thomm225]

Originally Posted by Paul Elliott

I think this is the claim that is causing the disagreement. When a PWM controller is rapidly cycling on and off, a meter on the panel connections will display an averaged value somewhere between the panel Voc and the battery voltage. Not exactly the same thing as "the panel voltage", but probably still informative.

Other than that, I think we all agree: PWM is inefficient when the panel MPP voltage is significantly higher than the battery voltage. This may or may not matter -- it depends on the use-case.

FWIW, I'm an EE who loves solder-smoke.

Informative being the key word here.

Again this so-called Controller I was using at the time I was doing the method I described above seems to be more of a voltage regulator than PWM Controller.

Solder smoke is another thing that isn't happening much on maintenance sites/tech shops anymore.

I think we have one partially functioning Pace Station the rest use adjustable soldering irons and wicking for the 2-4 times per year we need to solder anything

As far as PWM efficiency, this was a very telling revelation when I first hooked up my Victron Controller and checked the data

[Free2Sail]

Originally Posted by thomm225

As far as PWM efficiency, this was a very telling revelation when I first hooked up my Victron Controller and checked the data

Can someone point me to the best MPPTs for a not-yet installed 220W array (2 panels), perhaps eventually to grow to ~400W. I have cheap PWMs that panels came with, but want to do it right. A link to a recent thread is fine. Thanks.

[thomm225]

Originally Posted by Free2Sail

Can someone point me to the best MPPTs for a not-yet installed 220W array (2 panels), perhaps eventually to grow to ~400W. I have cheap PWMs that panels came with, but want to do it right. A link to a recent thread is fine. Thanks.

You could start an entire thread with that question......

[zstine]

Originally Posted by Free2Sail

Can someone point me to the best MPPTs for a not-yet installed 220W array (2 panels), perhaps eventually to grow to ~400W. I have cheap PWMs that panels came with, but want to do it right. A link to a recent thread is fine. Thanks.

Victron makes a quality product. Available on Amazon. you can buy a 100v/50amp and be able to expand, or buy one controller for the 2 panels now and just put the expansion on another controller. keep in mind that you shouldn't mix panel sizes on the same controller. 2 panels of similar output / number of cells is fine on 1 controller

[goboatingnow]

Originally Posted by zstine

Victron makes a quality product. Available on Amazon. you can buy a 100v/50amp and be able to expand, or buy one controller for the 2 panels now and just put the expansion on another controller. keep in mind that you shouldn't mix panel sizes on the same controller. 2 panels of similar output / number of cells is fine on 1 controller

Confusingly the Victron Blue solar mppt range doesn’t have integrated Bluetooth while the “ Smart Solar “ mppt range does.

[zstine]

Originally Posted by goboatingnow

Confusingly the Victron Blue solar mppt range doesn’t have integrated Bluetooth while the “ Smart Solar “ mppt range does.

yeah, to connect to the 'Blue' series they have a bluetooth dongle you have to buy separately. A little more money, Not a big deal though

[thomm225]

Originally Posted by zstine

yeah, to connect to the 'Blue' series they have a bluetooth dongle you have to buy separately. A little more money, Not a big deal though

I have had both types in the size of 75/15.

The Lighting Mode/Phase display is even different on the two.

On the Smart Solar (Bluetooth) Model Bulk Mode/Phase it's indicated by a solid Blue Light and on the BlueSolar Model Bulk Mode is indicated by a rapidly blinking yellow light.

Photo shows my laptop hooked up to my new Smart Solar Victron MPPT Controller with the w/VE.DIRECT TO USB INTERFACE CABLE that I had for the BlueSolar Model.

At least the jack was the same.

Controller is currently hooked to one 65 watt panel and one 50 watt panel.