Grid quality meter, included in product

That uses the ADS131M04, which is a new part not yet at any distributors. However, TI has a few hundred parts at their store. We assume it'll be more widely available soon. Meanwhile their are many alternate qualified and suitable, maybe even better suited AC mains ADCs, including many from Analog Devices. But you mention code. The question is, does that code give you exactly what you need? Or could your programmer better use the source as an example to write something more suited for your needs.

Yes, I could reuse the HW and do my own code

It?s not terribly complicated

Ac mains voltage rms Dev from ideal social sine Power factor Peak voltage Peak deviation Surge peak and time Burst peak and count Yada yada yada

Historical data Storage in EEPROM Comm interface Special HW that does not die from surges and measures correctly during abno rmallity

Hmm, seems it won?t be a 2 day dev job ?

Cheers

Klaus

I doubt that TI thing measures transients.

An analyzer as you describe would generate a *lot* of data fast.

Power factor implies that you want to measure your load too. Why not measure power?

I'd hate to meet an anti-social sine! Heh!

bnormallity

Yes, a lot of data. Would need to sample continously and evaluate continous ly, so as not to miss a transient

We should also measure power, but is far from enough

We have failures from the field that we cannot determine where they come fr om. So the idea is to have correlated grid measurements with field failure data

Cheers

Klaus

abnormallity

usly, so as not to miss a transient

from. So the idea is to have correlated grid measurements with field failur e data

We can measure burst, but not with good quality

Surges are possible with a multi MSa ADC (8/50us pulse)

Cheers

Klaus

You need an analog peak detector for short voltage transients. See AoE III, page 255.

There are commercial line quality analyzers around. Rent one of them maybe.

Are the field failures near the power inlet, or signal, connectors? Is there lightning in the area?

Yes, a peak detector and a S/H would be needed

Burst can come often, with very high rate (100kHz) and continously. So that will be a worse situation with respect to data than surges, which are rare

Cheers

Klaus

g abnormallity

ously, so as not to miss a transient

from. So the idea is to have correlated grid measurements with field failu re data

We have some of those.

But the failure rate is low, in the 100'reds of ppm, so we need a large pop ulation to get errors. Renting 100'reds of analyzers would be economically unsound and a logistical nigthmare

Cheers

Klaus

Are you sure you need to record the amplitude of every pulse in a burst, isn't it enough to know the most powerful, and the rate? Then you can simulate that in the lab, and design a fix.

Search for the "Landis + Gyr" and "E850" to get an impression of the features of a common grid meter.

Wim

I might not need to actually record the bursts, since they are typically not fatal, they just reset the micro

But would be nice to have that included

Cheers

Klaus

Thanks.

It seems they do not record surges and bursts

Cheers

Klaus

If you need one that records surges and similar transients (1us resolution), search for "Siemens" and "Q200" instead.

Regards, Dimitrij

It's not clear if the PQDIF format can contain much 1us information. Can you clarity for us?

Restarts are recorded anyway.

Wim