diy solar

diy solar

Series or Parallel

From your description, it sounds like you conditioning unit has a soft-starter, which is good news.

In my experience though your numbers are not going to be enough to support that load for that long. First, let's look at your sun. The panels might be close to full power at ~1:00pm, but no way at 5:00pm, even in the tropics. At 1:00pm, I'd go with 75% output, or ~ 2475W. At 5:00pm, I'd say it's more likely to less than 25%, or about 825W. Any shortfall in power is going to almost immediately drain your battery. It's just too small for the application.

If you go by the yardstick of not draining an Li battery at more than 1/4C, then to support a 2400W load, you need 200Ah battery. To support that same 2400W load at 5:00pm, without strikingly draining the battery, you need 9600W of panels.

If you can't stomach those numbers, I'd say increase battery capacity to at least 200Ah, and at least double the solar you have now, and be ready to add more at some later date.
 
From your description, it sounds like you conditioning unit has a soft-starter, which is good news.

In my experience though your numbers are not going to be enough to support that load for that long. First, let's look at your sun. The panels might be close to full power at ~1:00pm, but no way at 5:00pm, even in the tropics. At 1:00pm, I'd go with 75% output, or ~ 2475W. At 5:00pm, I'd say it's more likely to less than 25%, or about 825W. Any shortfall in power is going to almost immediately drain your battery. It's just too small for the application.

If you go by the yardstick of not draining an Li battery at more than 1/4C, then to support a 2400W load, you need 200Ah battery. To support that same 2400W load at 5:00pm, without strikingly draining the battery, you need 9600W of panels.

If you can't stomach those numbers, I'd say increase battery capacity to at least 200Ah, and at least double the solar you have now, and be ready to add more at some later date.
Thank you Michael. If I understand you correctly, 9600W, I would need 17 panels of 550W. My inverter is 5000W so I would need another one. So 17 panels and 2 inverters to run one AC for 5 hours. Or add more batteries. Seems a lot for one AC. Most houses here have an average of 20 panels for the whole house. They must have spent a lot of $$$$$ on batteries. Did I understand you right? Thank you for all your help.
 
You can parallel East facing and West facing strings for more total panel watts than what the inverter can process through its MPPT. That gives more hours, lower peak current.
 
Thank you Michael. If I understand you correctly, 9600W, I would need 17 panels of 550W. My inverter is 5000W so I would need another one. So 17 panels and 2 inverters to run one AC for 5 hours. Or add more batteries. Seems a lot for one AC. Most houses here have an average of 20 panels for the whole house. They must have spent a lot of $$$$$ on batteries. Did I understand you right? Thank you for all your help.
No. You do NOT need two inverters. What you need are panels pointed towards the 5pm sun. You have to understand that there's a difference between what panels can make, and what panels do make.

At noon, when the sun is close to directly perpendicular to the panels, you will be at full output. Your panels might be putting out 90% of what they are rated at. The power level is high enough to start and run the air-conditioner with no problems. By 5pm though, the sun is no longer directly over the panels, but at an acute angle. That means they'll be producing far fewer watts then they were at 1pm.

As an alternative to having all your panels facing one way, you can have multiple arrays facing in different directions. Maybe one array of panels on the roof to catch the midday sun, but another array facing west at an optimal angle for 5pm sun. Look at this pic. This is one of my West-facing arrays positioned for collecting late afternoon sun. Notice the steep angle? That's because at 5pm, the sun is lower in the sky, and almost due West. At 5pm, it's in full sun, whereas my South-facing arrays are starting to become fully shaded.

If you have different arrays facing in different directions, you will never have full output from both arrays at the same time. The one facing the sun will be close to full power, but the one at the wrong angle for that time will be outputting very little. That keeps you within your amp limit of your charge controller. That's why you don't need a second inverter. As the sun rotates across the ski, the West array reaches full power as the South array is dwindling. Voltage of both arrays stays about the same, but the amperage varies with the intensity of the sun it's getting.

Another advantage of having multiple directions (called virtual tracking) is that on cloudy days, when the light is very scattered, both the arrays will be putting out at least something, maybe 10% of sunny output. But, two arrays putting out only 10% is better than just one array putting out 10%.
 

Attachments

  • panels on array frame.JPG
    panels on array frame.JPG
    1.1 MB · Views: 3
Thank you Michael. If I understand you correctly, 9600W, I would need 17 panels of 550W. My inverter is 5000W so I would need another one. So 17 panels and 2 inverters to run one AC for 5 hours. Or add more batteries. Seems a lot for one AC. Most houses here have an average of 20 panels for the whole house. They must have spent a lot of $$$$$ on batteries. Did I understand you right? Thank you for all your help.
Don't forget, many homes with solar are grid tied, and they don't have any batteries. They sell solar to the grid, and buy it back...
The solar rarely covers their own loads.
 
No. You do NOT need two inverters. What you need are panels pointed towards the 5pm sun. You have to understand that there's a difference between what panels can make, and what panels do make.

At noon, when the sun is close to directly perpendicular to the panels, you will be at full output. Your panels might be putting out 90% of what they are rated at. The power level is high enough to start and run the air-conditioner with no problems. By 5pm though, the sun is no longer directly over the panels, but at an acute angle. That means they'll be producing far fewer watts then they were at 1pm.

As an alternative to having all your panels facing one way, you can have multiple arrays facing in different directions. Maybe one array of panels on the roof to catch the midday sun, but another array facing west at an optimal angle for 5pm sun. Look at this pic. This is one of my West-facing arrays positioned for collecting late afternoon sun. Notice the steep angle? That's because at 5pm, the sun is lower in the sky, and almost due West. At 5pm, it's in full sun, whereas my South-facing arrays are starting to become fully shaded.

If you have different arrays facing in different directions, you will never have full output from both arrays at the same time. The one facing the sun will be close to full power, but the one at the wrong angle for that time will be outputting very little. That keeps you within your amp limit of your charge controller. That's why you don't need a second inverter. As the sun rotates across the ski, the West array reaches full power as the South array is dwindling. Voltage of both arrays stays about the same, but the amperage varies with the intensity of the sun it's getting.

Another advantage of having multiple directions (called virtual tracking) is that on cloudy days, when the light is very scattered, both the arrays will be putting out at least something, maybe 10% of sunny output. But, two arrays putting out only 10% is better than just one array putting out 10%.
 
Thank you Michael for helping me (a newbee). I learned a lot on this forum. It makes a lot of sense. If I would have a large backyard I would build a tracking system. But for now I will do something like this photo.
 

Attachments

  • different angle.png
    different angle.png
    2 MB · Views: 3
Guys let me know if I should open a different thread for this. Subject: Ground. Many YT videos show how to ground your panels. But then I saw a video from Will Prowse saying do not ground your solar panel it will cause a ground loop very dangerous. Then I think he means ground t but separately. Needless to say I am confused. So Please look at my design and let me know if it is a good wiring plan or what it should be. Keep in mind that I will not be sending electricity to the electricity company. Only using the electricity from the city to supply the Hybrid inverter. If that makes a different.
 

Attachments

  • Ground scenario A.jpg
    Ground scenario A.jpg
    274.6 KB · Views: 2
Any solar array that is in full sun at 11am can NOT be in full sun at 5pm
So not true anywhere towards the middle of the earth, looking at my future solar roof, it is in the sun from about an hour after sunrise to an hour before sunset.
The amps on the York 24,000 BTU ramp ramp up gradually
No surge then, easy to start.
You should in theory be able to run without a battery but any system will be much more stable with one.
In your case, even 4 lead acid car batteries would do the trick, no need for Li.
 
You want to ground the frames of the panels to your house ground if possible, their own ground rod is second best but still OK.
This is mainly to bleed off static and induced voltage from lightning...neither of which is likely in your beautiful area of the world.
The secondary reason is to protect you IF the panel leaks voltage to the frames if the insulation breaks down.
 
You want to ground the frames of the panels to your house ground if possible, their own ground rod is second best but still OK.
This is mainly to bleed off static and induced voltage from lightning...neither of which is likely in your beautiful area of the world.
The secondary reason is to protect you IF the panel leaks voltage to the frames if the insulation breaks down.
No.
Their OWN ground is very bad, unless it is ALSO connected to your house panel, or there is no house panel.
 
No.
Their OWN ground is very bad, unless it is ALSO connected to your house panel, or there is no house panel.
Not true, 99.99% of the time, the frames are connected to nothing at all, so it does not matter.
Static or induced spikes, just dump them to ground.

You split phase people are so confused by grounding lol.
 
Not true, 99.99% of the time, the frames are connected to nothing at all, so it does not matter.
Static or induced spikes, just dump them to ground.

You split phase people are so confused by grounding lol.
At least per NEC requirements the panel metal frames, and metal mounts, must be bonded to a ground system.
 
At least per NEC requirements the panel metal frames, and metal mounts, must be bonded to a ground system.
I meant the frames are not connected to electrical circuitry, not that they are not or should not be grounded.
It was not clear what I meant, sorry.
 
So not true anywhere towards the middle of the earth, looking at my future solar roof, it is in the sun from about an hour after sunrise to an hour before sunset.
Maybe I should make my statements a bit more clear. I would define full sun as the sun hitting the solar panels at an exact 90 degree angle, so they are at maximal power. Just because there is still sun on your roof an hour before sunset does NOT mean they are at an optimal angle for solar production.

Once you actually have panels on your roof, and you are actually producing power, will you understand the reality of this.
 
Don't forget, many homes with solar are grid tied, and they don't have any batteries. They sell solar to the grid, and buy it back...
The solar rarely covers their own loads.
Good point. In the future yes. But here it is a very long process to get permit and approval to tie it to the grid so I’ll start off with off grid for now. Thank you.
Maybe I should make my statements a bit more clear. I would define full sun as the sun hitting the solar panels at an exact 90 degree angle, so they are at maximal power. Just because there is still sun on your roof an hour before sunset does NOT mean they are at an optimal angle for solar production.

Once you actually have panels on your roof, and you are actually producing power, will you understand the reality of this.
Very clear thank you.
 
Ok
At least per NEC requirements the panel metal frames, and metal mounts, must be bonded to a ground system.
so I should change my drawing picture to only show grounding the panels and rails to the ground. Correct? Sorry stil confused. Or if someone post a diagram on how to ground a similar system that would help me a lot to get my head around that. Thank you!
 
Panel frames and rails should have a wire connecting them back to chassis of inverter or SCC.
If panel frames are grounded to earth, then inverter chassis can be driven to an AC voltage by AC that couples to PV wires inside inverter.
Will discovered that belatedly, after several forum members got shocked by panel frames.
One forum member had ground mount panels and inverter was on an RV (insulate from earth by rubber tires). The RV gave his dog a shock. Running the wire fixed it.
 
Back
Top