Measuring 2 imput voltages!

Hello guys.I am working with arduino for quite a while now,and i have come across a need to measure 2 separate transformers output and make a compare.Practically i have 2 transformers that output 400V(yes four hundreds volts),and i want to use 2 analog inputs to measure the 2 voltages and activate an output if one of the 2 are off or not ± the voltage they need to be(if one is lower than 350V activate the relay).
Now i have done all the coding and it is working with 5 volts input.Now i was thinking to do 2 resistor dividers for each input voltage,but the problem is that i cant have a neutral ground between the 2 sources.I need to keep the 2 sources completely separate,no common ground.So practically the voltage divider is not gonna work.
Anyone have don this before or have any sugestions? Any help is appreciated.Thanks!

One solution is to have an external ADC with SPI interface connected through optical couplers to the main controller.

I hope you know what you’re doing; you know there minimum separation between conductors etc, right?

400V AC? Get a camera and take the battery out of the smoke alarm first!

OTOH you could google "AC voltage comparator".

I thought more about it,and i think it could be done using 1 bridge rectifier for each supply line,then connect together the - so it will result in a common ground,and then the resistor dividers for each line to get from 400 to 5Volts.What do the experts think about this?

It was you who said cannot have a common ground... be careful with that, be careful also with where you get power to feed the arduino, since you may be creating conduction paths between the high AC and mains and whatever. And remember that AC peak is AC rms x 1.414.

How about connecting a small transformer to each thing you are trying to sense, and convert it down to a more manageable voltage? That will also isolate the low voltage side from both transformers so you don't need to worry about inadvertantly coupling the transformers together.

Well i think i made a mistake when i told in the first post that the 2 voltages are AC voltages.They both are already DC voltage,one is exactly 415V and the other is oscillating a bit from 385 to 395 volts.They both come from 2 different motor drivers that are driving 2 DC motors.The motors have an fixed 380V DC supply for the excitation of the stator.Practically i want to make a circuit with an small LCD,that will tell me if the 2 motors run at the same speed all the time.I though that reading the 2 voltages and compare them will achieve that,and i dont want to modify the drivers,neither to open them to see if i can find an common ground.
I also had another idea,to couple an dynamo to each motor end shaft and to compare the 2 voltages,that i could combine them with common ground.The only problem is that i have to add the 2 dynamo and the trouble to make the mechanical arrangement into the costs.
I hope that now i was clear enough about my problem,so maybe somebody have an better idea of how to achieve this.

Regards Daniel.

If you want the motors to run at identical speed then measure RPM

Are you just trying to tell whether the motors are nominally 'on' or 'off', or are you trying to do analogue measurement of the voltage and/or speed?

I am trying to do analog measurements of the 2 voltages.I think that i will have enough resolution to see the voltage from 0 to 400V if it will be scaled right to 0...5V with the divider.That will give almost 0.4V of resolution and i will be more than happy with that.My main concern is that i want to be sure that i will not combine by mistake the 2 voltages.
As for the other reply,how can i measure RPM and not measure voltage??

If you want to know how fast the motors are running, you might want a tachometer not a voltmeter. If you know there's a consistent relationship between your voltage signal and the motor speed, you could measure the voltage instead if that was more convenient (although it doesn't seem to be more convenient in this case). In general, though, the motor speed varies with load as well as voltage so measuring the voltage may not be a good way to determine speed.

Yes i know it is better an tachometer.But like i sad i dont need very accurate measurement.I know from the motor specifications that @ 400V it spin @ 3000min-1 .I only wanted to compare the 2 motors to see if sometimes they are out of sync by more than 10%. Most of the time they spin around 1000 min-1 both,i only need to see the situation if they both run more than 10% difference between each other.I hope i make myself understand,since i am not a native english man.

Assuming the motors are both the same type, voltage rating and seeing similar loads then they will consume the same power ie current. You could therefore use current to sense the deviation. Simply use a hall type sensor and feed one motor's positive feed cable through the core left to right, then feed the other motor's positive feed cable through the core from right to left. If both motor currents are equal, their fluxes will cancel and the hall sensor will be balanced. If the motors become unbalanced the hall sensor will output a voltage, dependant upon the amount of out of balance and in a direction which illustrates which motor is higher and the other lower loaded.

To measure RPM deviation you simply need a photosensor looking at an illuminated reflector on each shaft. The arduino will count the pulse rates from each and hence inferred speeds. Subtract one from the other and divide by the lower to determine by how much the faster is running.

Inferring speed from voltage or current only works if you know or can predict the load. Perhaps the load is constant, or varies in some predictable way with speed. If so, you have a hope.

If the load isn't constant or predictable but you know the electrical characteristics of the motor then from voltage and current you can work out the resistive drop and the back emf and approximate the speed from that.

But unless you're able to stay within these restrictions, measuring voltage or current is not a good way to determine rpm.

As already mention by other users, forget voltage/current measurements and measure shaft rotation. It's much more easy, precise and safe than you can imagine. Even "equal" motors are never "equal". The "put a reflector/anti-reflector pad on motor shaft and measure with and IR LED/detector" as suggested is very simple and precise.