Why is the accuracy of multifunction testers so low

Hi Alan

We have run into this kind of issues a couple of times with customers in rural locations (for different makes of EVCP) - but to the credit of our local DNO (National Grid), they would investigate and act when we raised a query such as yours.

I suspect your own MFT may well be more accurate than the numbers the EVCP itself can report, however the readings of the EVCP may help to give you the extra evidence to persuade the DNO to come out and check for themselves with good diagnostic kit, without you having to spend extra time or money. As the EVCP has been tripping out and (I imagine) is connected to the internet, can the supplier help you by providing some data on the tripping events (or even - a time trace of the voltages)? I am not too familiar with the Tesla offering - can the voltage trace even be downloaded from the customer's app / web portal somehow? Armed with such information I would hope the DNO will be more helpful. 

Whatever the absolute precision of the voltages involved, the big variations you noted yourself during the span of only an hour do suggest that the local supply is not very stable and liable to exceed the permitted supply parameters at times - so really the DNO should be dealing with this one. Just to convince yourself it is not your kit playing up, does it produce more stable readings at home / in a built-up area with a presumably more stable supply?

Could you argue that typically a meter will read either high or low?  After all if it is say 5 volts - or more - out, then it will likely be in the same direction for the low and high voltages; it is not going to be reading low at the lower limit and high at the upper limit so as to give the DNO grounds to ignore it. Is it?

Clive

errata - I thought I had posted this a few hours ago, but it sat here unloved, given other answers that have come in , it is now slightly out of order, but may still be informative .

Portable Multi function testers  are a compromise, as there is neither the space nor the budget to incorporate  all the bits for a lab grade instrument, that also has to be battery powered, and rugged enough for regular van  transport, and work on a cold day or a hot one.

Consider a 'good' lab bench meter may be one of these, but skip to pages 7(DC) and 8 (AC) to see how the accuracy may vary with temperature and over one year after cal.

And the price tag is not that inexpensive - expect to pay over £1k for one of those.
PDF
A really 'precision'  meter is a few times that times that...

https://www.keysight.com/us/en/product/34470A/digital-multimeter-7-5-digit-truevolt-dmm.html

a snip at 4k

example from Le Croy  Idare not ask the price

But additional accuracy is harder than it looks.

Mike

Hi Alan 

Just a suggestion - Could you hire a voltage logger? and gather the voltage values (max/min etc.) over a period of a week or so. If there is anything untoward you have the data to prove to the DNO. 

Kind Regards,

CPC 

I see the Fluke 1662 MFT quotes Accuracy 45 Hz – 66 Hz 0.8% + 3digits. I suspect similar from other manufacturers are also available.

(That said: I will say in answer to your question, having had the misfortune in the past of having to design an AC meter for a specialist market, yes it is very difficult and expensive to guarantee that accuracy! And yes to be high accuracy it has to be slow to average out the rubbish - it's the irony that you can have fast and inaccurate or slow and accurate, which absolutely is a pain with voltages that are jumping around.) 

On your original point, it does feel a bit as if you're in a vicious circle here if you have to prove what the voltage swings are before the DNO will prove what the voltage swings are...