Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Solar PV inverter - Input simulation

Status
Not open for further replies.
Hi,

I want to test a SMA 4kW grid tie solar PV inverter under max load, or close to it at least. Initial thoughts are to get a 230/110V isolating transformer (about 5kA rating) & bridge rectify, then feed into the inverter PV inputs as full wave rectified DC.
I can't see why this wouldn't work, but I'm not confident about it. Does anyone have any thoughts? Would I need to smooth it?

Thanks.
 
With real solar panel input, the voltage at the input of the inverter will vary but have virtually no ripple. I would hang the biggest electrolytic capacitor I could find between the rectifier and the inverter input. Specify the nominal MPPT voltage of your panels, and the current into the inverter, and you can calculate the required capacitor C*ΔV = I*t, where t is 8.33ms and ΔV is the allowed ripple (say 10% of the DC value).


C (in Farads) = 0.00833*I/ΔV (for 60Hz line)
 
Thanks. Any idea how the MPPT would behave? I guess a rectified transformer won't drop off voltage as the current is increased like a PV panel?
 
Around MPPT, the solar array is more like a current source than a voltage source. A filtered rectifier would be more like a voltage source than a current source. You could put a light-bulb ballast in series with the transformer primary to turn it into something that looks like a current source...
 
Around MPPT, the solar array is more like a current source than a voltage source. A filtered rectifier would be more like a voltage source than a current source. You could put a light-bulb ballast in series with the transformer primary to turn it into something that looks like a current source...
Great idea thanks, but I'm not sure I'll get the power I need. It's made me wonder if a better solution would be to put current control on the DC side. Max 15A - anyone got any thoughts on best approach for that?
 
Seems to be doable to just use isolation transformer - diode bridge - capacitor. At least with EverSolar TL1500. I've been running both SMA and SamilPower inverters both from lead acid batteries and Nissan leaf battery at about 390v.

Thanks for that, and the link. I assume the inverters will always try to output as much power as they can? i.e. they run their MPPT tweaking & look for best combination of current x voltage (i.e power).
Except with the batteries they can potentially draw maximum current at nominal voltage - do they just keep on drawing more & more curernt at max voltage? Do they stop? If so where? At their max current?
I'm assuming grid connected, so the mains load is infinite
 
I can just answer how it worked for the SMA inverter tripower 8kw, and as you said it goes all the way to it's max output and seems to be stable but the SamilPower was a few years I tested and as I remember it was quite unstable and fluctuated around 50% power but I'm not totally sure, will try to test that again soon.
 
I would like to connect my inverter for experiments like done in the video as well but I don't have a proper isolation transformer. I thought that I might be able to use 2 microwave transformers "backtoback" so 230v-(a few)kv to (a few)kv-230v, do you think that would work (if i have 2 transformers with about same power capabilities and same primary-secondary ratios) and would it even be possible to parallel 2 or 3 sets like that to increase the power capability?
 
So that gives you 230V isolated? Why not stick them in series to double the voltage rather than the current? Either way though I think you need current limiting. e.g LM338 giving 5A. Possibly more than one in parallel, if that's possible
 
Seems like LM338 www.ti.com/lit/ds/snvs771c/snvs771c.pdf is for dc up to 32v. But I think don't think thats a problem.
If I connect to similar transformers back to back I guess I'll have the same voltage in as out and if I connect 2 "output" transformer in series to a single "input" transformer I guess I'll have the double output voltage but still the same amount of power cause the primary winding in the input transformer will still see 230v and same amps right?
 
The LM338 should, I believe, be floating & not actually 'know' it's voltage. I believe it'll only have 1.2V across it, and that that's the important thing.
I was supposing 4 transformers in your case, each just providing 230V floating. That can then be rectified, probably with a fat capacitor. The two identical circuits could then be put in series. But I think it needs current limiting, unless the only limiting factor is the inverters own ceiling, assuming it controls that, rather than fusing or blowing up.
 
I want to test a SMA 4kW grid tie solar PV inverter under max load, or close to it at least.


Why? What do you expect to gain or learn from it for the possible costs involved?

To me the simplest method to test it would be to set up a portable generator as the power source and run it off that.

At most use a older designed unit that uses a brushed rotor and automatic voltage regulator so that you can bypass the regulator and manually control its output power.
 
For the efforts involved building a custom DC power supply that can deliver ~4 KW and simulate a solar panel's output is not all that easy.

Whereas modifying the voltage regulator circuit of a old 4 - 5 KW portable generator to work as a solar panel simulator is not all that difficult in comparison plus unlike the line powered system it would not have the potential to make your local utility connection AC power characteristics unstable either.

To be honest, as someone who has studied and designed home built grid tie units I can say that in most cases GTI units that become unstable or that show odd/intermittent issues at their upper working range limits most often comes down to the grid connection they are connected to becoming unstable due to a number of reasons. Undersized wire and or insufficient utility connection capacity being the primary causes.

Depending in what method of power feedback they use it's entirely possible to have a unit push the local grid connection voltage up to high and shut down or to have it create an excess of line harmonic noise that cause false tripping events due to insufficient loading issues or other such issues due to the AC waveform being distorted too far in one way or another.

Then add the fact that many GTI systems are not in fact capable of running at their supposed peak ratings to begin with or at least not designed to run at them continuously. Despite the price tags some come with they are not always well designed and built. To be honest most are nothing more than a common power inverter with a few dollars worth of additional control components to make them synchronize with the grid voltage and frequency similar to how a sine wave power inverter is only a few percent more dollars in manufacturing cost to make despite carrying a 2 - 4+X price tag on the shelf.
 
Thanks for the insight. One advantage of feeding from the grid is that I'll only be using a little electricity. (assuming I can make it work). Inverter 0.96 x 0.98 x 0.98 tranformers so let's say I'll use 10% of the test load - the other 90% will be going back into the grid. Well, it'll obviously never make the grid...
So this way I can contemplate investigating intermittent errors. I do have some gear to closely monitor mains, which may be an option for intermittent diagnosis.
Interesting to hear they can rarely hit & maintain their rated capacity.
I'm sure you're right about the true viability of it, but I've got the transformers, capacitors and inverters, so I'll probably scratch this itch if I can work out a reasonable system. It bugs me having a pile of suspect inverters. I'm sure I'll learn from it regardless.
 
Thanks for the insight. One advantage of feeding from the grid is that I'll only be using a little electricity. (assuming I can make it work). Inverter 0.96 x 0.98 x 0.98 tranformers so let's say I'll use 10% of the test load - the other 90% will be going back into the grid. Well, it'll obviously never make the grid...

Unfortunately as I have found out myself from personal experimentation trying to pull a high load off of a AC line while simultaneously trying to feed that power back via a independent feedback system makes the AC line unstable due to issues with waveform distortion, harmonics, and power factor related weirdness. On a larger utility power system it's not such a problem but on a smaller one with other factors like line resistances and whatnot in play its not hard to create your own problems simply due to a sort of feedback loop effect.

Whereas using an actual independent power source to power the systems while the grid tie systems do their thing is not so much of a problem since that how those units are designed to be used.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top