30 minutes ago, TinkerBoy said:

Sjoe waar het jy so n OU foto gekry .... 

2 minutes ago, The Terrible Triplett said:

As dit jou happy maak, dink maar so ja, of rook nog van daai goed wat jou so laat dink ... (rofl)

Eish so lekkka  

Interesting that the Axpert takes the quiescent current from the grid when in bypass (did I get that right?). Nothing wrong with that, it's actually quite cool. The Victron always takes it from the battery, even when in bypass (aka grid mode). Of course this distinction only matters if (like me) you use only solar to charge the batteries and the charger in the inverter is disabled, because then it means that at night the inverter continues to consume 2 watts (for my particular model) from the bank even when in bypass. If you actually use the charger in the inverter... then there is no real distinction.

 

6 hours ago, PaulF007 said:

 

@Chris Hobson what is your experience in this matter?

Kurt from the AEVA measured consumption @ 46W. I am not too fussed about the inverter's self-consumption. 46W amounts to 1.1kWh which under normal circumstances can be generated by a single 300W panel. So we going to turn the world upside down for 1 panel's worth of energy. Think back to the good-old days of incandescent bulbs, the power consumed by inverter is the equivalent to leaving a light on. We lose far more energy through not having a big enough load for most of the day. How many of our battery banks are charged by midday? I am just grateful to have power during the day that is not costing an arm and a leg.

 

9 hours ago, PaulF007 said:

there isn't much you can do about it , apart from maybe going Blue hey

Roughly half that consumption is coming directly off the panels so roughly 600wh from the batteries. Nearly  1A discharge on a 48V battery bank. I am not going to buy an inverter because it has ½ the quiescent draw of an Axpert. The saving does not justify the cost. I will buy a Victron or another inverter because of its true hybrid capability and therefore can do things that the Axpert cannot do.

Agreed @Chris Hobson the only reason why I am looking at the "losses" is for the sun does not play its part. Also to check if there is a consistency between the Axpert users. Mine measures about 70w , @viper_za seems to run at 50w , yours seems to be the same as @viper_za .

We have to keep in mind that these inverters are "cheap" in comparison so it is fair to assume that you will get fluctuations in quality and as such losses. Mind you I count my beans under a magnifying glass so every cent is important and so my inverter was VERY expensive.

This reminds me of the joke about relativity (sorry this one is best done in Afrikaans)

" Een haar op n man se kop is relatief min 

maar

Een haar in jou sop , by die Spur is relatief BAIE".

 

 

 

  

Right I have some more info and a theory so here goes.

I took the full load off the inverter (after lighting some candles and parrafien lamps , the SWAMBO instantly had a headache ) and then got the reading from the BMV nou it gave me 50w like the rest of the guys. So 

• My cables could be the problem although they are about 500 mm long and is 16mm2 and according to calculations there should not be any problems there The cable from the batts to the DC box is a 90mm2 
• The inverter use about 20w to 30w to do the DC to AC thing and mine is just not as efficient as some of the others. This is what I am leaning towards
• The fuse box and DC isolator and those bits and pieces have a high resistance and I am loosing some watts there.

Any thoughts on this?

The moment I add a load there is an extra loss of 20w so my loss is about 70w in total

@PaulF007 I like that you need to find out all the little watts, wanting to know what is going on and why.

Just looking at what the BMV readings is not the true 100% accurate value. As we have seen BMV says X, so inverter must by X. No it is not so.

When you read the inverters values you get say X-1. All developers who have developed software to read devices have seen these subtle differences. Together they make up the +-85% efficiency of the entire system.

Obviously some of us ignore these small values, they are not important, just add another panel ... but look after the pennies and the pounds will look after themselves. You as a financial person understand that extremely well.

Way I see it, in today's world it is so easy to throw away 20w here, I mean, 20w serias!? But it is more than that. It is 20w here, another 10w there, and some more over there, and one day you sit and say WTF ... no, not that WTF ... the Wine Tasting Friends WTF.

One by one you will find these little bliksims and understand your load requirements so much better, getting a leaner and meaner system in place = better ROI in the end.

Respect.

 

7 hours ago, TinkerBoy said:

So all of this is about 20W ?? LOL

Not so much 20w as for a few more lessons learned , suspensions confirmed and now have even a much better understanding of my system and what it can and cant do.

For instance I now know that it is no use to see if I can bring my base lower than 50w at night (When everybody is asleep) as the load on the batteries will stay about 100w on average. 

When the sun does its thing I don't care much for the losses as they are for free bat the moment the sun goes down and I am paying for those cycles then I notise ...

 

8 hours ago, TinkerBoy said:

How do you know there is an extra 20 watts being used by the Inverter ? The load changes every second.

True but every second the difference stays the same.

All of this makes me think of the urban legend of the old Tannie that noticed 50c charged extra on her insurance.

She called the company but the service center dismissed it but she was adamant and kept on pursuing the matter. After a while they started to get fed up with her and agreed that she was overcharged and they will refund her in kind.  But she wanted to know why she was over charged. Eventually this "complaint" ended up on a table of a penny pusher and he started to make work of it.

It turned out that a programmer within the company wrote a small program the skimmed 50c from every client n the company and paid it to a account somewhere. There were more than 500 000 customers........ and all of that for 50c  

1 hour ago, TinkerBoy said:

With what software and or meters and what is your methodology of testing. 

I am using the data directly from the Inverter and the BMV I haven't checked the load with something else.

What I did was 

1. Removed all loads from inverter
2. Checked the difference between load at the inverter and load at BVM, 0-50 = 50 watt loss (Confirmed that that is about true for every one with a Axpert)
3. add a 3w load (LED Globe) 
4. Inverter showed 3 - 4 watt load  - BMV showed 73 w load 
5. I know that with no load the BMV showed 50w so when I add 3 w it adds 20 watts to the BMV
6. Confirmed it with other loads ( 20w and 30w ) with the same results 

But if there is a better way of doing it that could be used by other Axpert users , we are all ears.

Kyk, ek is dom. What am I missing?

When Edmund asked the same question a while back, actually using the exact same batts and Axpert, it was also nitpicked. Why?

 

Paul has set a baseline - No load. BMV says 50w is used. Then he adds a 3 watt load. Inverter says 3w is drawn. Yet BMV says 73w, up from 50w.

Pray tell me why that is not a very good very simple very basic way to set a baseline to see what is potting? A plan that every person can use to confirm what happens without buying more measuring devices that will only confirm that DC to AC conversion comes at a price? What is that price? Do this quick check to get an idea, costs you not a sent.

What is coming out it seems to me is that some Voltronic models may be more efficient than others, for whatever reason.

And less we forget, what I am taking away here, if you run small loads on a large inverter, make sure the load justifies the conversion losses for over time, like 20w penalty to power small loads is high AND it adds to battery costs over time.

Another little sum to do in order to cost the actual ROI on a solar system.

20 minutes ago, The Terrible Triplett said:

very simple very basic way to set a baseline to see what is potting?

This is what I am aiming towards and if all the Axpert manne is using the same methodology I can not see why it could not be a good bench.

Is it scientific? - Most definitely not  

Does it cost money - not a cent (unless you go and buy a inverter just to check )

Is the finding useful -  well to me it is ( Even if the measurement is out by 30% as long as it stays consistent you can work with it.)

 

 

Mmmh. So couple of things.

1. First, at such very low loads, inverter efficiency is rather terrible. This is mostly due to the current the inverter itself requires to run, so I am pretty certain that is NOT what you're seeing. If the inverter needs 50W to work, then with the LED lamp on it should need 53W to work, and you would then say the effiency is 3/53 ~= 6% efficient. Here the delta is 20W.

2. The current meters in inverters aren't precision stuff either. Even in the flagship blue inverters, at low wattage, it gets it very very wrong. I actually think THIS is what you're seeing. See this for example.

3. LED lamps often have very poor power factors and 20VA for a 5W lamp is not uncommon. Comes back to measuring precision again.

You need to put a precision instrument (at least one that tries) on the output, especially if you're going to use small <20W loads :-)

@plonkster just out of quriosity, how accurate is the is the BMV?

Also this video that looks at a Kill-a-watt meter. These things use A-to-D converters on the inside, and the digital side is often limited to a certain number of bits, for example, with your smaller Arduino type MCUs the AD converter is 1024 bites wide, so it can measure roughtly a thousand discrete steps. Map that over a range of 5KW or more, and the smallest unit it can measure is 5W. For my efergy, for example, the smallest unit is around 18W. The slightest bit of error and your 20W lamp shows up as 36W. The interesting bit about the Kill-a-watt is that it can switch "ranges" in software to get better readings lower down.

That realisation is at 28:28 in the video.

The MCU in your inverter is likely not much different and will suffer the same issues.

 

But.... I went back and read what you said, that the BMV is also showing 73W. Spec says it's accurate to 0.01A, or 10mA, at 48V, that's accurate to about a watt. You really need a better way to accurately measure the output though. Simplest solution might be to use a 60W incandescent lamp as load. Those things are fairly accurate to a few watts and have a power factor of 1. Because they have a PF of 1, you can even measure AC current with your DMM and multiply it by the voltage, thereby saving on using a real power meter.

Now we are getting somewhere!

I'm actually worried that the outcome of all this is that the Axpert inverter manages a quiescent current of only (!) 50W by doing some kind of shutdown trick, but the moment it cannot fully shutdown (because the indicator lamp on the tumble dryer draws 2 watts) you see its real usage. I'm worried because at this point team Voltronic really don't need another bad data point... :-P

2 minutes ago, plonkster said:

I'm actually worried that the outcome of all this is that the Axpert inverter manages a quiescent current of only (!) 50W by doing some kind of shutdown trick, but the moment it cannot fully shutdown (because the indicator lamp on the tumble dryer draws 2 watts) you see its real usage. I'm worried because at this point team Voltronic really don't need another bad data point... :-P

 still cheaper

Hi Paul

From your observations it appears the Axpert uses 50w to power it up with no load. The additional 20w maybe associated with load detection circuitry. I am not clever enough to look at the wiring diagrams to see if there is a portion of the inverter that switches off under no load.

 

Just now, Chris Hobson said:

Hi Paul

From your observations it appears the Axpert uses 50w to power it up with no load. The additional 20w maybe associated with load detection circuitry. I am not clever enough to look at the wiring diagrams to see if there is a portion of the inverter that switches off under no load.

 

Between actually measuring it, or asking @Coulomb , we'll figure it out :-)

Gents please don't get too fixated on the 20 w that number also varies in either direction. I have also setup a feed in EmonCMS so that I can monitor the "losses" overnight and plot them so that I can get a better "feel" for what is happening in real time. Obviously daytime data will be out completely but I will ignore that. 

I have to agree, 20 or 50w is not a lot on 5000va ... but who uses 5000va at night constantly?

Yes, I also agree the rands and cents on 50-70w is near to nothing.

What irks me is 50w for inverter just being on = 5 x 10w spotlights ... on the entire night.

Yes but 5x spotlights is a moerse lot cheaper than 4 x inverter