DIY MPPT SCC?

[Mike Jordan]

Has anyone DIYed an MPPT solar charge controller?

 

[LittleDaB]

Ive been thinking about using my spare PWM or MPPT for something a little different.
So, I have a 24V LiFeP04 pack and would like to run a 1000w 12V inverter off it...without tapping into the midpoint...temporarily!...of course. I was thinking about installing the SCC between the pack and the inverter, perhaps with a 12V 100AH deep cycle lead acid battery between the SCC and the inverter. So what's gonna blow up lol ?

 

[rin67630]

Has anyone DIYed an MPPT solar charge controller?

I have died a couple of SCC's, yes!
I am currently in the process of conceiving and debugging a family of WiFi controlled MPPT solar charge controllers by injecting a signal into the selection of buck converters, to cope especially with the requirements of low power needs for instrumentation projects.
Tim at Github has done it in power ranges up to 500W based on a DROK power supply.
But in the range over 20A you are probably better off buying the regular stuff.

 

[LittleDaB]

I'm a very basic meat and potatoes kinda guy. I'll leave circuit manipulation to the learned.

But seriously, what could go wrong with using an SCC powered by a 24V battery to charge a 12V battery. It won't be able to keep up to a 1000 watt draw, but it should be able to replace approx. 500 watts (26X20).

 

[rin67630]

I'm a very basic meat and potatoes kinda guy. I'll leave circuit manipulation to the learned.

But seriously, what could go wrong with using an SCC powered by a 24V battery to charge a 12V battery. It won't be able to keep up to a 1000 watt draw, but it should be able to replace approx. 500 watts (26X20).

If you are sure that your SCC is a real MPPT, go for it in confidence.
If the SCC is just pretending to be, but is really PWM electronics inside, you will get very high current pulses between your two batteries.
Enjoy the magic blue smoke...
You can separate the good from the fake with a multimeter:
If, without any connection, the minus of solar panel and battery are connected (< 1 Ohm) together, the chances are great that you have got a real MPPT circuit.
If the the minus of solar panel and battery are not connected or > 100 Ohm, I am ready to bet, that it is a PWM inside, whatever is written on the case...

 

[LittleDaB]

If you are sure that your SCC is a real MPPT, go for it in confidence.
If the SCC is just pretending to be, but is really PWM electronics inside, you will get very high current pulses between your two batteries.
Enjoy the magic blue smoke...
You can separate the good from the fake with a multimeter:
If, without any connection, the minus of solar panel and battery are connected (< 1 Ohm) together, the chances are great that you have got a real MPPT circuit.
If the the minus of solar panel and battery are not connected or > 100 Ohm, I am ready to bet, that it is a PWM inside, whatever is written on the case...

Good to know!!
It's an Outback MX60...10+ years old

 

[fat_old_sun]

If you are sure that your SCC is a real MPPT, go for it in confidence.
If the SCC is just pretending to be, but is really PWM electronics inside, you will get very high current pulses between your two batteries.
Enjoy the magic blue smoke...
You can separate the good from the fake with a multimeter:
If, without any connection, the minus of solar panel and battery are connected (< 1 Ohm) together, the chances are great that you have got a real MPPT circuit.
If the the minus of solar panel and battery are not connected or > 100 Ohm, I am ready to bet, that it is a PWM inside, whatever is written on the case...

This test assumes that the PWM controller is switching on the negative side. There are PWM controllers that switch on the high side. These will pass your test & come off as MPPT's.

 

[Mike Jordan]

I have died a couple of SCC's, yes!
I am currently in the process of conceiving and debugging a family of WiFi controlled MPPT solar charge controllers by injecting a signal into the selection of buck converters, to cope especially with the requirements of low power needs for instrumentation projects.
Tim at Github has done it in power ranges up to 500W based on a DROK power supply.
But in the range over 20A you are probably better off buying the regular stuff.

I found that github a couple of weeks ago, and I recognize your screen name from the discord over there. That build looks exactly like what I am looking for. But every time I ask a question, you guys go on 47 tangents that are all above my pay grade. I am hoping I can discern some direction out of the 47 tangents. But I am not sure I will ever make any progress.

I was hoping to find some folks as intelligent as you guys, that could also speak in non-geek. Or maybe find a translator from the geek to non-geek

What would really be awesomeness is if @Will Prowse did a build video for a DIY MPPT controller. That would fit quite perfectly into his channel. And he is a GREAT translator of geek to non-geek

 

[rin67630]

But every time I ask a question, you guys go on 47 tangents that are all above my pay grade.

We are in the middle of brewing that beer. That goes not without turmoil... Wait for a couple of weeks more and enjoy your fresh tapped draft, once the paperwork is done.

 

[LittleDaB]

I found that github a couple of weeks ago, and I recognize your screen name from the discord over there. That build looks exactly like what I am looking for. But every time I ask a question, you guys go on 47 tangents that are all above my pay grade. I am hoping I can discern some direction out of the 47 tangents. But I am not sure I will ever make any progress.

I was hoping to find some folks as intelligent as you guys, that could also speak in non-geek. Or maybe find a translator from the geek to non-geek

What would really be awesomeness is if @Will Prowse did a build video for a DIY MPPT controller. That would fit quite perfectly into his channel. And he is a GREAT translator of geek to non-geek

Way above my pay grade as well...but very intriguing. My keg is right beside me ?

I found that github a couple of weeks ago, and I recognize your screen name from the discord over there. That build looks exactly like what I am looking for. But every time I ask a question, you guys go on 47 tangents that are all above my pay grade. I am hoping I can discern some direction out of the 47 tangents. But I am not sure I will ever make any progress.

I was hoping to find some folks as intelligent as you guys, that could also speak in non-geek. Or maybe find a translator from the geek to non-geek

What would really be awesomeness is if @Will Prowse did a build video for a DIY MPPT controller. That would fit quite perfectly into his channel. And he is a GREAT translator of geek to non-geek

 

[austrisv]

What would really be awesomeness is if @Will Prowse did a build video for a DIY MPPT controller. That would fit quite perfectly into his channel. And he is a GREAT translator of geek to non-geek

As a big fun of OSPController - it was powering my off-grid shed and shallow well water pump last summer - would absolutely love to see it featured on @Will Prowse channel. Meanwhile, maybe Tim's video makes life little easier for casual brewers OSP Hardware build

 

[Mike Jordan]

As a big fun of OSPController - it was powering my off-grid shed and shallow well water pump last summer - would absolutely love to see it featured on @Will Prowse channel. Meanwhile, maybe Tim's video makes life little easier for casual brewers OSP Hardware build

I found these guys a while back. I was pretty excited. But they operate on a level above my pay grade. There was so much "mine is bigger than yours" going on, I was lost. They finally got things a little better reeled in, so I got my mind wrapped around the hardware. But the software programing has me lost. I am waiting on Tim's software video, before I try to jump in

 

[rin67630]

But they operate on a level above my pay grade.

Tim's solution isn't that expensive?
Where is your pain limit and how much power do you need to control?

 

[Mike Jordan]

Tim's solution isn't that expensive?
Where is your pain limit and how much power do you need to control?

Above my "Paygrade" is a euphemism for above my skill level. After getting Tim to reword some of the verbiage and flow of the hardware pages, to make sense to a wannabe geek, as opposed to a real geek I got that now. I am waiting on the software update vid, which I don't get. But with enough waiting, my original use project needs to move forward in another direction. But still interested in figuring it out for future use

 

[rin67630]

how much power do you need to control?

 

[Mike Jordan]

how much power do you need to control?

I was looking to start with just a DIY portable solar gen, so probably not more than 200 watts . But hardware is the easy part. Thats not my issue

 

[Horsefly]

Just stumbled upon this thread. I have two spare nodemcu ESP32s sitting around looking for something to do. Even though that component is one of the cheapest in the github project, I'm tempted to do this!

 

[austrisv]

Just stumbled upon this thread. I have two spare nodemcu ESP32s sitting around looking for something to do. Even though that component is one of the cheapest in the github project, I'm tempted to do this!

Hey, @Horsefly, come over to OSP discord and we will get you going! Depending on what setup you have, maybe all you need is to add DPS5005 to your esp32 and off you go

I was looking to start with just a DIY portable solar gen, so probably not more than 200 watts . But hardware is the easy part. Thats not my issue

@Mike Jordan, may I ask you to check the direct message on discord? Maybe I can help you to deal with the SW while Tim is working on video. (I think) I speak a little non-geek :D

 

[rin67630]

I was looking to start with just a DIY portable solar gen, so probably not more than 200 watts . But hardware is the easy part. Thats not my issue

Then Soft Power may be your friend.
Soft-Power-MPPT
It is specialized in efficient low-power operation and deliver tons of visibility without need for setting up a server/databases/MQTT stuff...

 

[Warpspeed]

Just found this thread.
A super simple cheap no frills mppt controller is fairly easy to make, but there are those pedantics that are never happy with a simple solution.

Basically what we must do is control the electrical loading on the panels, so that the voltage at the panel stays pretty much at the maximum power voltage printed on the rating plate. So at dawn, at first, the voltage is allowed to rise with no load at all.
As the sun gains some strength, we can apply some load, and increase the current output into the battery, but never overload it enough to cause the voltage to fall below the optimum maximum power voltage.

If a cloud obscures the sun, we need to throttle back the loading a bit, and increase it again when the cloud has passed. Its as simple as that.

The usual way this is done is with some fancy software that starts out knowing nothing about the system, but by trial and error (perturb and observe) gradually adjusts solar panel loading to find the power peak. There is a much simpler way to do it than that, without requiring a microcontroller or any software !

Its just a case of monitoring the solar voltage and regulating that voltage with a simple voltage regulator.
In other words classic shunt voltage regulation.
We pull enough power from the solar panel to hold the voltage at the panel constant, no more, and no less. Its as simple as that.

Likewise, with a lithium battery, we keep feeding all of the available current from our shunt voltage regulator into the battery, until it reaches full rated battery voltage. We then must throttle back the current to hold a fixed battery voltage. The battery charging current then gradually ramps down to zero, and the solar panel voltage is then free to rise above the maximum power voltage.

These two functions can both be easily performed by a single ordinary high frequency pwm control chip driving a buck converter.

Its a bit like your standard bench power supply that has a fixed regulated output voltage, and an adjustable current limit.
Two distinct modes of operation that cross over.

In this case, its a constant solar panel voltage regulator, and constant battery voltage regulator. These two parameters cross over when the battery reaches its full rated voltage. Its not difficult or complicated, and it works very well.

Now the cynics will say Ah! but what about changes in temperature or solar insolation, the peak power voltage is not a precise fixed figure.
Your cheap crappy NON mppt controller idea cannot follow that !!!

True, but the actual peak power voltage does not change all that much from what is on the rating plate. And if you build it my way, its possible to tweak the solar panel voltage potentiometer up and down under load, and watch the change in bulk charging current.

You might be surprised that the power "peak" is not a peak at all, but a very broad low hump. The solar voltage is surprisingly non critical.

What DOES change is insolation. At night there is nothing. At mid day, in a clear blue sky its at maximum. Between those two extremes we have a very wide range of solar output power to track, and a simple constant voltage circuit will do that very well, and be very fast acting too.
So my idea definitely works, and is proven to work very well.

Sure, a software mppt might beat it by a very low single digit percentages, but it would not be by much more than that.
Perturb and observe algorithms can be slow to respond, and sometimes lock onto false peaks. My system is very fast and cannot be fooled.

In the great scheme of things, it has certainly proven itself here, and its ideal for a simple do it yourself project for any tinkerers here.

If you build it yourself, its also not going to cost a fortune for multiple controllers, and be repairable, something that the commercial solar controllers very rarely are.