Author Topic: Diagnosing problems with solar mains supply  (Read 860 times)

0 Members and 1 Guest are viewing this topic.

Offline herodotusTopic starter

  • Newbie
  • Posts: 2
  • Country: es
Diagnosing problems with solar mains supply
« on: November 25, 2018, 02:24:06 pm »
Hi All,

First time poster with fairly limited knowledge, so be gentle! I live off grid and our power is supplied from a 3kw inverter connected to a solar-charged battery bank. The inverter is reasonably high-end (Outback) and has generally been pretty solid for us, but periodically we get problems with all the LED lights in the house flickering annoyingly. I have tried some rudimentary debugging with a multimeter, and also by unplugging all devices/disconnecting wiring spurs and trying again, but so far I've failed to identify any clear cause.

At this point I really want to be able to actually look at what's going on with the supply, but I'm really not sure how to proceed. The obvious tool here would be an oscilloscope but;
  • I'm really uncertain how to do it safely given that pretty much by definition I can't isolate the DUT (it's the mains supply!)
  • I don't have hundreds of euros to spend on this
My inclination here is to buy a very basic USB scope for ~100 euros (entry level Picoscope e.g.) and test using a battery-powered laptop. I'd be interested to hear if people think this is unsafe for some reason I don't understand or if there's a better way to approach this.

Thanks!
 

Offline Ian.M

  • Super Contributor
  • ***
  • Posts: 13216
Re: Diagnosing problems with solar mains supply
« Reply #1 on: November 25, 2018, 04:45:51 pm »
The usual problem with scoping the mains supply safely is the risk of a significant potential difference between Neutral and Ground, possibly caused by a fault outside your control and maybe even not on your property, and the risk of dangerous high voltage transients on the incoming supply.

In a single dwelling, entirely off-grid situation neither of those apply, (unless your panels are struck by lightning, which will probably blow up the whole system and all connected loads etc.) and provided you test Line with respect to Ground, not Neutral, do so near to the system's Neutral to Ground bond, and use an appropriately rated x100 HV probe (CAT III or better), it should even be safe to power the scope from the supply you are testing, so you can do some long-term data logging.

However I would *NOT* like to attempt such a measurement with a budget USB or other software controlled scope, as its input coupling capacitor is almost certainly not rated for mains across it, and to safely attenuate a voltage exceeeding the max. input rating, the probe *REQUIRES* that the scope input *ALWAYS* presents a 1Meg DC resistance to ground, which many do not do when AC coupled.   You are therefore one click or software glitch away from blowing the scope input.  Due to the x100 probe's very high series resistance, you'll *probably* get away with it on a purely  AC source, and if you do blow the scope, its unlikely to damage anything else.     If your scope is rated for 100V or higher input voltage, connecting the probe via a BNC T, and wiring a NE2 neon bulb directly across the other arm of the T will provide extra protection for its input when using a x10 or x100 probe, as a NE2 neon typically strikes at under 100V, without affecting low frequency measurements.   If like many cheap USB scopes, its only rated for +/- 30V or 50V you are out of luck - one mistake and it will let out the holy smoke and join the choir invisible.

However brand-name* CAT III rated x100 probes don't come particularly cheap, and you'd probably be better off  splashing the cash on a low end (limited bandwidth) isolated differential HV probe, which will guarantee there are no hazardous voltages on the scope side of it.

* When a fake or faulty probe could be lethal, buying an off-brand probe or one from EBAY etc. wolld be rather unwise.
« Last Edit: November 25, 2018, 08:00:08 pm by Ian.M »
 
The following users thanked this post: herodotus

Offline malagas_on_fire

  • Frequent Contributor
  • **
  • Posts: 591
  • Country: pt
  • Kernel Panic
    • Malagas Lair
Re: Diagnosing problems with solar mains supply
« Reply #2 on: November 25, 2018, 04:52:03 pm »
Does the inverter produces modified or pure sine wave? If you can't find the probes try at least one isolation transformer example 230 to 12V, if 230V is you're mains live supply voltage. be sure to measure isolation voltage in respect to ground before pluggin into scope and recheck with multimeter.

If one can make knowledge flow than it will go from negative to positve , for real
 

Offline herodotusTopic starter

  • Newbie
  • Posts: 2
  • Country: es
Re: Diagnosing problems with solar mains supply
« Reply #3 on: November 25, 2018, 06:02:49 pm »
The inverter is pure sine wave. I could use an isolation transformer - that's probably going to be the cheapest approach to this that isn't dangerous. My main concern with this was whether doing so would potentially fail to pass through some of the distortions that I'm trying to track down.

I can't really justify buying a proper differential HV probe - it's just too much money.  I'm not quite sure why I'd need cat 3 passive probes, I would have assumed that cat 2 would be fine for testing the output of a 3kw inverter? I don't think it's really capable of delivering anything like grid currents even in a fault scenario, is it?

I'm starting to feel like I'm going to have to find someone with the proper equipment to look at this for me rather than trying to do it myself; I can't justify spending 500 euros on gear just to work out why my LEDs are flickering  :(
 

Offline Seekonk

  • Super Contributor
  • ***
  • Posts: 1966
  • Country: us
Re: Diagnosing problems with solar mains supply
« Reply #4 on: November 25, 2018, 07:47:55 pm »
Any flickering problem you can see will easily pass through a transformer. I would tend to think this is a battery problem. As a good diagnostic, I would consider two cheap digital voltmeters that are self powered.  As you likely have a battery bank, connect one meter across half of the bank and the second on the other half.  The two voltage should be equal.  Above telling you when there may be a bad connection under load, it will indicate when the batteries are out of equalization.

Stuff like this can be handy to diagnose a problem.  My system is microprocessor controlled and I added a feature that measures five timed voltages when the refrigerator turns on. As a fairly constant short term 120A load it gives me a good indication of connection resistances and battery health.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf