Voltage divider vs potential transformer based design

Hi Aswin,

I think all the metering circuits I have seen have used a potential divider for the voltage measurement, but I have no idea what specs are relevant for meters or any such applications.

Isolation is needed somewhere, but that could be digital isolation at the output of the energy meter circuit, or somewhere in-between. Power could be from each phase.

Personally, if I was going to design it, I'd make it as three separate modules, so only one could be used for single phase, and have them all have an isolated output.

But then I don't make energy meters, and if you look at existing meters, I doubt they will have such a design, and instead they will integrate/cost-reduce onto a single board.

By the way, you mention 230, 460, 690V, but the voltages are not in-phase, so these will not be seen. The RMS voltage will be (say) 230V (single phase), or 400V (phase-to-phase)

[or multiply accordingly for whatever country it is in]. In the UK, we tend to see higher nominal values, like 240V or 415V, or even higher since that's not a precise value from the grid.

Hi Shabas,

Thank you for your reply!

In India we also have the same single phase 230 3-phase 240/415 (RMS).

To be exact my DAQ systems goes infront of power delivery systems of motors and pumps. They are rated at 230 and 440 vollts.

The reason I don't want any transformers is to make the design less bulky and less prone to noise. But I got into an argument with a CT vendor last weekend. They insisted us on designing the frontend with a stepdown transformer.

I have don't have experience in designing industrial power/energy meters.

According to your experience can I go ahead with the voltage divider model?

Hi Aswin,

Measuring current with a current transformer is normal (although so is using a current sense resistor, at least my home (consumer) energy meter plug uses a current sense resistor for size and cost reduction), a current transformer is more usual for industrial I'm sure, since there will be less power loss.

I just don't see the reason to have to measure voltage with a step-down transformer if you don't want to. Series resistors will reduce the voltage and drastically limit current in a fault condition, and you can have isolation elsewhere.

I'm sure there must be dozens of example circuits (or could even tear-down existing energy meters) to prove this is done in real life and meets some specs of some sort! : )

They insisted us on designing the frontend with a stepdown transformer

There may well be a step-down transformer for powering the circuit, but as for measuring voltage with it, how accurate could it be? I'm not sure normal voltage transformers are designed for precision so then you'd need to calibrate : (

Completely agree with you. For powering up the device I was thinking about TPS54060 (60-V, 0.5-A, step-down regulator with an integrated high-side MOSFET) based solution with 1N4757A Zener.

Something like this.

For power rails isolation, I am gonna use an external power supply for digital rails and above mentioned power supply for the analog rails.

I think the design should depend on your application. If it is for mass production and needs to be cheap then by all means you must investigate resistive voltage dividers and may be even a resistor type current sense.

If it is for low volume industrial applications, and even more so if experimental, then you should consider the use of ready made current and voltage senors from LEM and others - these will cost more (for the parts) but can much more easily be made safe.

Whichever route you go you must observe local rules with regard to safety, including pcb creepage distances, resistor breakdown voltage, transient peak voltages etc etc. When you are planning a resistive divider take a look at Dave Jones excellent video https://www.eevblog.com/2012/10/19/eevblog-373-multimeter-input-protection-tutorial/

This will give you a better idea of the true cost of a voltage divider that is actually safe and reliable.

I would never willingly use equipment with a power supply such as you suggest - 1 little capacitor between me and the Grim Reaper - no thanks - (actually on three phase it needs only one of three caps to fail).

Use a proper isolated and certified power supply for all your internal supplies.

As a general rule, if a mains powered device has any external connections other than to the mains, then those other connections must be transformer isolated from the mains. So your power supply might be OK in an LED light bulb or a meter with only an optical display and no connections. Although people try it is almost impossible to make a safe device without transformer isolation.

MK

Thanks Michael!