STPM34 Measuring 240Vrms without Neutral

[HaydenG6]

Hello,

We are designing an EVSE for a North American 120/240-V split-phase supply. The EVSE is connected to L1 and L2, with a nominal voltage of 240 V RMS between them. Neutral is not available at the EVSE.

We would like to use the STPM34 to measure the L1-to-L2 voltage and calculate the load power and energy. Current sensing is implemented separately using current transformers and is not shown in the attached schematic.

Our proposed voltage-sensing circuit uses the primary differential voltage channel as follows:

• L1 is connected to VIP1 through three series 560-kΩ resistors.
• L2 is connected to VIN1 through an identical resistor chain.
• Each input has a 1-kΩ resistor and filter capacitor connected to the local analog ground, iGND_A.
• iGND_A is the local ground of the isolated STPM34 metrology domain. It is not connected to neutral or protective earth.

The intention is to create a symmetrical differential divider so that the voltage between VIP1 and VIN1 is proportional to the L1-to-L2 voltage. The metrology ground would float approximately at the midpoint between L1 and L2.

I reviewed the STPM34 datasheet and the available STPM34 evaluation-board schematics, but I could not find a reference circuit showing how to measure a line-to-line voltage when no neutral connection is available. The evaluation-board implementations appear to use neutral as the metrology reference, which is why we considered the symmetrical differential arrangement shown above.

Could you please advise on the following?

1. Is this a valid and recommended method of measuring a 240-V line-to-line voltage with the STPM34?
2. Could movement of the floating midpoint due to unequal parasitic leakage, resistor tolerances, or line-to-ground voltage imbalance affect measurement accuracy or violate the input common-mode limits?
3. Must VIP1 and VIN1 each remain within the specified input range relative to GNDA, in addition to the VIP1–VIN1 differential voltage remaining within range?
4. Would ST instead recommend referencing the isolated metrology ground directly to one line and using the standard single-ended voltage-divider arrangement?
5. Is there an STPM34 reference design or recommended circuit specifically for a two-wire, 240-V split-phase application without an accessible neutral?

Thank you.

[Didier HERROUIN]

Dear HaydenG6,

We do not have a reference design for your use case.

Nevertheless, your schematics are valid (as long as STPM34 is isolated from the rest of your circuit as you mentioned). I would just recommend to decrease the 1k resistors value to ~700 ohms , because with 1k+1k, the differential signal at 340Vpeak may exceed the +/-300mV ADC input range specified in the datasheet (when the L1 and L2 are 180° phase-shifted).

VIN1 and VIP1 must remain inside the maximum voltage range defined by the spec (+/-0.7V relative to iGND_A), but according to your schematics, this is the case if I am not wrong.

Regards,

Didier

 

 

[Richard Li]

​@Didier HERROUIN 

L1 L2 is AC, without rectifier, after resister, the signal to float GND sometimes is negative, This violate IO input voltage range.

If IO ESD diode can handle this negative, it may not damage, but no good.

Please don’t confuse AC and DC.

Please remember, in north American 120/240-V split-phase supply., L1 and L2 always 180 degree phase different. 

[Richard Li]

For EVSE, NEC Article 625 no need Neutral, but your design for STPM34, it need DC power, the GND still need connect to Neutral, AGND connect to GND, you can’t float AGND.

So any time your two inputs always got wrong power phase.(Negative voltage)

If you really want measure two phase voltage, use safety design, add one transformer.

[Didier HERROUIN]

Dear Richard Li,

From what I understand, the Neutral is not available and the STPM34 is supplied by an isolated power supply (generating an isolated_3.3V and an isolated_GND). The floating ground, thanks to the resistors bridge, should always be centered between L1 and L2. 

[Richard Li]

​@Didier HERROUIN 

L1 L2 is AC, without rectifier, after resister, the signal to float GND sometimes is negative, This violate IO input voltage range.

If IO ESD diode can handle this negative, it may not damage, but no good.

Please don’t confuse AC and DC.

[Richard Li]

Sorry forgot have to add a rectifier.