Electronics > Power/Renewable Energy/EV's

Testing Jig(s) for home solar inverters


A couple of years ago, I queried about using a dc power supply in place of a solar array in troubleshooting home solar inverters, most of which, if not all , are grid-tied inverters.

The discussion diverted into the more theoretical aspects, like the non-linearity of the V-I curve and so on, without any useful results or conclusion as to how to construct or purchase such a dc power supply.

Looking on the Internet recently, there is no much information on this subject. Most of the information revolves around, again, on the theoretical side of emulating the solar panel array with a dc power supply and goes deeply into mathematical equations .. in other words, nothing practical from a technician's point of view.

So to keep things simple.  I am interested in troubleshooting inverters from a hardware point of view ONLY, that is, if  components or cards or modules are in fully working order.

Not a bit interested  in the software side of the inverter

Effectively, I am looking for suggestions, in setting up a jig consisting of a dc power supply (suitable for all inverters out there to simulate the solar panel array,   the inverter under test and  connection(s) to  appropriate loads.  The objective being, again, to perform live testing on the hardware, by taking voltage readings, etc to determine the cause of a problem.

The above method in the long run is more productive than some of the methods I have used and other may be using to date.

The most popular method, at least by me, is to search for defective components, initially, visually ( it takes some experience even to do this and I am not talking about blown up components ), subsequently, checking for other bad components around the defective component, then if everything looks ok, followed by cold testing of  suspect components, looking up the history of possible faults associated with a particular inverter, naturally, taking into consideration the error displayed on the lcd screen (but sometimes this may not be known ) and eventually board substitution (if you got the right board, that is)

However, having gone through all this process, a defective resistor/diode or even a cap  may hold up repairs considerably.

What input ranges are you looking at for your inverters? Something like a HP 6010 or 6030 can do 0-200v upto 1kw (Max of 17 amps or 1kw, whichever is less). They are auto ranging so at 100v you can get 10 amps, but at 200v you can get 5 amps.

Another one is the 6035, they are a similar idea but 0-500vdc to a max of 5 amps.

This should give you a bit of beef to test the inverters usability without the whole IV curve solar array tester bullcrap.

I do a lot of ELV testing of solar regulators, chargers, batteries etc... Most things are no higher than 48v. I have a 6012 0-60v 1kw/50a and a 6011 20v 1kw/120a unit (this needs a bit of repair from the previous owner) and these work great for what I need.

If I was to get a 60v 1kw power supply, most of them will only give 17a or thereabouts at the whole range, therefore most of the time when I use it at 12-20v it is not that big a power supply.

I hope my ramblings are on the right track for what you are wanting.

So am I right in assuming you really just need to check that it operates at a particular power rating and that any MPPT function actually bothers MPPT'ing?

For most purposes only really require something that gives some negative slope to the IV curve. A resistor can do this quite nicely. You can up the game a little and use something akin to an amplified diode kind of arrangement, this makes the emulator system a bit more efficient at the maximum power point (MPP) but still incredibly lossy at open open circuit - so its up to you whether you want to dump the power in a resistor or transistor.

Now, i know what i just said isn't perfect, but it will give you a (very simple and dependable) platform which you can use to test some of the more critical parameters.

There may be a few arguments to sharpen the IV curve and put a sharper kink in it like a real PV module, thinking whether there may be some control instabilities which will only be triggered by that sharper response... controllers can go unstable, but the culprit would probably more likely be input filtering, which you could scope out quite easily by looking at the input waveforms and to high frequency switching the PV module can filter current and voltage pretty well, to the non-linearity will rarely be seen by the SMPS, only at a large signal low frequency perspective from the MPPT algorithm which will probably be digital and will unlikely go untuned with time.

You can go a little more complex and make a slightly more custom controlled supply, there is an IET Electronics Letter called "low cost solar emulator for evaluation of maximum power point tracking methods", the paper is a little sketchy on its exact design but it can mostly be calculated from inspection rather than analysis - I published that design during my PhD but to be honest i did most testing using a resistor and power supply.

My primary concern, as I explained, is the hardware side of the inverter.

In other words, does the inverter has a short diode or a cap with a high esr or a resistor has gone high, perhaps a faulty IC. Why the inverter is not switching on, etc..  hardware issues only.

Some inverters may incorporate MPPT (maximum power point tracking), the purpose of which is to sample the power output (I-V curve) from the solar arrays and subsequently apply the appropriate load to maximize power output. However, strictly speaking, if the hardware is ok, this becomes a software issue. In which case, most likely would require replacement of the cpu board.

The likelihood of convincing the manufacturer ( if they are still in business) to sell the cpu only, astronomical.  However, apart from ICs holding programs or data, on the cpu board, a particular IC may have gone faulty.

To find out if that is the case, I need, power from the dc power supply, at least, so I can take a measurement to determine which of these ICs is not working.

So my main concern is to be able to dial up a voltage ( this voltage must be in the region which will turn the inverter on, if working properly) and a current setting to make the inverter "think" as if it was connected to the actual solar array.

Can you provide schematic diagram of your power supply that can be used to simulate this condition.

Sorry to zombie an old thread, but I too am interested in building a test jig for solar inverters. I came up with the variac/35 amp bridge and some old TV power supply caps. All from the junk draw, and the variac used for soft starting amps after repairs completed back in the day. I found some cheapo meters on ebay that measure current/volts/kwh for both the DC and AC sides, and intend to connect the AC out back into the mains in. The only input power from the mains would be a couple of hundred watts, the inverter feeding power back into itself via the rectifier.

I have the meters still coming on slow boat from china, but concerns I have for my jig, are limiting current if the MPPT side of the inverter(a 1.5kw unit with 50 volt turn on point) tries to drag max current out at lower input voltages. This would take down my supply (bridge on heatsink with fan is only rated to 35 amps).

I too came to using a bar heater or jug element in water as a ballast resistor, or maybe even a ballast transformer for a metal halide lamps(600 or 1000w)to limit the overall power going into the power supply on the AC side.

I dont have an isolation transformer, and the Inverter may see some error with the mains(and its own output) not being isolated from its  input. Did you get a decent test rig up and going in the end??


[0] Message Index

There was an error while thanking
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod