PRO EM-50 and PRO 3EM - measurement deviation

[gabay009]

I'm having a problem reading active power with the PRO EM-50. If the cos phi is very close to 0, the power reading is completely wrong. On the other hand, with the PRO 3EM, still in single phase, the power reading remains correct. In my case, my heat pump was stopped :
- Active power: 16W
- Apparent power: 350VA
- Amperage: 1.46A

The EM-50 and 3EM measured the same current and voltage. However, the 3EM didn't calculate the phase shift correctly. As a result, the EM-50 is nowhere near the 2% accuracy between 0 and 1 amp that it claims. Is my model the problem? Is there a specific adjustment I need to make? Why is the 3EM calculating active power correctly? When the cos phi is better, the EM-50 and 3EM have relatively consistent readings, for example for my hot water tank :
- Active power: 18W
- Apparent power: 48VA
- Amperage: 0.2A

--------------------------------------------------------------------------
J’ai un problème de relevé de la puissance active avec le PRO EM-50. Si le cos phi est très proche de 0, le relevé de la puissance est totalement faux. Alors qu’avec le PRO 3EM toujours en monophasé, le relevé de la puissance reste juste. Dans mon cas de figure, j’avais pour ma pompe à chaleur à l’arrêt :
-          Puissance active : 16W
-          Puissance apparente : 350VA
-          Ampérage : 1,46A

Le EM-50 et 3EM mesurait bien le même ampérage et tension. Par contre, le 3EM ne calculait pas de déphasage comme il faut. Du coup, le EM-50 est très loin des 2% de précision entre 0 et 1 A comme indiqué. Est-ce que c’est mon modèle qui a un problème ? Est-ce qu’il y a un réglage particulier à faire ? Pourquoi le 3EM calcul correctement la puissance active ? Lorsque le cos phi est meilleurs, le EM-50 et 3EM ont des mesures relativement cohérentes, par exemple pour mon ballon d’eau chaude :
-          Puissance active : 18W
-          Puissance apparente : 48VA
-          Ampérage : 0,2A

[gabay009]

And even with a cos phi close to 1, the EM-50 still has a significant margin of error, while the 3EM is accurate compared to the value measured by Victron. I'm using the latest 1.4.3 firmware on both units. What do I need to do to get the EM-50 to give me an accurate result ?

--------------------------------------------------------------------------

Et même avec un cos phi proche de 1, EM-50 a une marge d'erreur encore non négligeable alors que le 3EM est lui exacte si on le compare avec la valeur mesurée par Victron. Je suis sur le dernier Firmware 1.4.3 sur les 2 appareils. Que dois-je faire pour que EM-50 me donne un résultat exacte ?

[thgoebel]

You may call a difference of 3,4% a "significant margin". I do not. Tolerances of 3% with power measurement are usual, even with calibrated energy meters used by the electricity companies.

[gabay009]

The technical documentation for the Shelly PRO EM-50 says that the accuracy is 1%. So yes, 3% is unacceptable to me. Especially since if the power factor is close to 0, the EM-50 measurement is completely wrong.

[thgoebel]

Sometimes the specs have to be read thoroughly. For the Shelly pro EM-50, an accuracy of 1% for the voltage measurement and 2% for the current measurement is indicated:

 

Power is U * I * cosφ. Thus, the both accuracies have to (geometrically) added. This gives more than 1% accuracy.

Furthermore, you‘re comparing Shelly pro EM-50 with Shelly pro 3EM. Accuracy for Shelly pro 3EM is 1% for voltage measurement and 5% max. for current measurement:

Hence, your reference device is less precise as your device under test! For power measurement accuracy, the same as above written is valid.

Conclusion: 3,4% deviation with power measurement is lamentable but a common fact.

Edited September 19 by thgoebel

[gabay009]

Indeed, thank you for these clarifications, you're right.

Having said that, I also tried with my Legrand eco-meter and the value was also accurate compared to the 3EM and Victron, whether in the case of 6000W or 18W. So I'm wondering if my EM-50 has a bad calibration.