# 120V/208V Voltage sense

Joined Oct 9, 2012
14
I am trying to build a hall effect based voltage sensing system.
http://www.ebay.com/itm/New-5A-20A-...Pi-UNO-/112275500486?var=&hash=item1a24242dc6
I am asking for suggestions on a different source for the hall effect sensor. My current design only has a 42.5mA peak assuming 120V RMS, and the sensor full scale is 5A.
I am asking if anyone knows of a chip that maybe has full scale of +/- 50mA to get better resolution.

I don't want to decrease my resistance as I would like my voltage sense system not to consume to much power or get to hot.
The goal is to feed the isolated signal to an Arduino or RasperryPi without having a Potential Transformer. I targeted my resistors to have about the same current draw as the magnetizing current of 3 PTs

thank you for your time and have a good ay

#### Attachments

• 11.7 KB Views: 16

#### tcmtech

Joined Nov 4, 2013
2,867
Why not use transformers?

It's a far more efficient and cheaper than the proposed method you have here.

Joined Jul 18, 2013
23,913
120/208 voltage sense?
The ACS712 is a current sensor not voltage.
Max.

#### EM Fields

Joined Jun 8, 2016
583
I am trying to build a hall effect based voltage sensing system.
http://www.ebay.com/itm/New-5A-20A-...Pi-UNO-/112275500486?var=&hash=item1a24242dc6
I am asking for suggestions on a different source for the hall effect sensor. My current design only has a 42.5mA peak assuming 120V RMS, and the sensor full scale is 5A.
I am asking if anyone knows of a chip that maybe has full scale of +/- 50mA to get better resolution.

I don't want to decrease my resistance as I would like my voltage sense system not to consume to much power or get to hot.
The goal is to feed the isolated signal to an Arduino or RasperryPi without having a Potential Transformer. I targeted my resistors to have about the same current draw as the magnetizing current of 3 PTs

thank you for your time and have a good ay
https://sensing.honeywell.com/index.php/ci_id/49804/la_id/1/document/1/re_id/0

#### EM Fields

Joined Jun 8, 2016
583
120/208 voltage sense?
The ACS712 is a current sensor not voltage.
Max.
Since E = IR and since current in a series circuit is everywhere the same, I think that what he's trying to do is to determine the voltage on each of the 3 phase legs by measuring the currents through the known resistances.

#### EM Fields

Joined Jun 8, 2016
583
I am trying to build a hall effect based voltage sensing system.
http://www.ebay.com/itm/New-5A-20A-...Pi-UNO-/112275500486?var=&hash=item1a24242dc6
I am asking for suggestions on a different source for the hall effect sensor. My current design only has a 42.5mA peak assuming 120V RMS, and the sensor full scale is 5A.
I am asking if anyone knows of a chip that maybe has full scale of +/- 50mA to get better resolution.

I don't want to decrease my resistance as I would like my voltage sense system not to consume to much power or get to hot.
The goal is to feed the isolated signal to an Arduino or RasperryPi without having a Potential Transformer. I targeted my resistors to have about the same current draw as the magnetizing current of 3 PTs

thank you for your time and have a good ay
Are your only requirements determining whether the source voltage is 120V or 208VAC and that the signal providing that data be isolated from the mains?

Last edited:

Joined Oct 9, 2012
14
EM Fields, you are right. The resistors set the current and the device provides an insulated signal to the Arduino.
The PTs that I use at work typically are designed to get 120V on the secondary.
I could use the 4160V PT with a 35:1 ratio and that would give me about a +5 to -5 signal.(peak of 120Vrms is about 170)
A MV PT costs 700 bucks each.
I could possibly use a general control transformer that goes from 480V to 12V for about the same ratio, I am not sure what the accuracy and linearity are on a control transformer. I figure at lower voltage I wouldn't saturate, but the magnetizing current might make it non linear.
Still that chip is not that expensive, and control power transformers cost more than you think.

Either case the PT is a source (separatly derived system), and the Hall effect was using an arduino sourced voltage and its return signal would be the same reference.
(EDIT, There is a fault current out of a transformer that could damage things, particularly when the utility goes above rated voltage. The Arduino is power limited)

Last edited:

Joined Oct 9, 2012
14
I am trying to make a 3 phase voltage transducer that is safe for the Arduino with good linearity and accuracy.
I need more than to know if its 120V or 208V, I want to tell down to 0.1 Volt accuracy.

thank you.

#### tcmtech

Joined Nov 4, 2013
2,867
I could use the 4160V PT with a 35:1 ratio and that would give me about a +5 to -5 signal.(peak of 120Vrms is about 170)
A MV PT costs 700 bucks each.
I could possibly use a general control transformer that goes from 480V to 12V for about the same ratio, I am not sure what the accuracy and linearity are on a control transformer. I figure at lower voltage I wouldn't saturate, but the magnetizing current might make it non linear.
Still that chip is not that expensive, and control power transformers cost more than you think.

You could just use three cheap old style 5 - 10 VA wall wart power supply transformers since all you need is a voltage signal off of them.

If saturation is a concern use 230 vAC rated ones with outputs rated at double what you need.

As for resolution of measurement that's all going to be determined by the A/D converters in the units and what amount of their input bandwidth you use.

#### crutschow

Joined Mar 14, 2008
28,177
I want to tell down to 0.1 Volt accuracy.
Why such exceedingly high accuracy for power line voltage?
Or is it just that you want 0.1V precision (resolution)?

Joined Oct 9, 2012
14
The cheap wall wart power supply's are NOT linear, there seems to be a recurring theme that transformers are assumed to be linear.
Potential transformers are designed to be as linear as possible.
http://www.flex-core.com/low-voltage-potential-transformer-models-460-468
take a look at the voltage circle diagram.
I also am trying to avoid a seperatly derived system, to provide some real isolation.

#### tcmtech

Joined Nov 4, 2013
2,867
The cheap wall wart power supply's are NOT linear, there seems to be a recurring theme that transformers are assumed to be linear.
Potential transformers are designed to be as linear as possible.
http://www.flex-core.com/low-voltage-potential-transformer-models-460-468
take a look at the voltage circle diagram.
I also am trying to avoid a seperatly derived system, to provide some real isolation.

Then explaining what it is you are trying to achieve or monitor and to what degree of precision accuracy is in order and why is necessary.

Also in my experience an unloaded secondary ona transformer has near zero linearity issues beyond those induced by capacitive coupling between the windings which are typically negligible in referencing common utility line power systems. Correct me if I am wrong.

#### crutschow

Joined Mar 14, 2008
28,177
Then explaining what it is you are trying to achieve or monitor and to what degree of precision accuracy is in order and why is necessary.

Also in my experience an unloaded secondary ona transformer has near zero linearity issues beyond those induced by capacitive coupling between the windings which are typically negligible in referencing common utility line power systems. Correct me if I am wrong.
I agree that a small unloaded power transformer should have near ideal voltage linearity from input to output.

And the OP does need to explain what is the purpose of this measurement, not just what he thinks he needs.

Joined Oct 9, 2012
14
I don't know how else to say it but transformers are not linear as perceived and as the link shows that.

In my first post I agree that I didn't state that I wanted 0.1V accuracy, but my original statement was pretty clear.
"The goal is to feed the isolated signal to an Arduino or RasperryPi without having a Potential Transformer."

I also explicitly stated what I was asking
"I am asking if anyone knows of a chip that maybe has full scale of +/- 50mA to get better resolution."

There is an xy problem going on here in reverse. I know what I want and I asked exactly what I wanted to ask.
If you don't have a chip selection that is OK, there is nothing wrong with the OP.
To respond to the constructive comment I got via my inbox.
although they are out of stock that chip might be a better way.
I could use a 300 ohm resistor with the 6K ohm to get a voltage divider for a +/-10V signal and have some room for overvoltage conditions.

#### tcmtech

Joined Nov 4, 2013
2,867
I don't know how else to say it but transformers are not linear as perceived and as the link shows that.

Yes and what are the min/max variations for no load conditions on the secondary?

In my books for any given properly designed transformer with no definable secondary load and not being run near or at core saturation power levels they are well within the limits that some very basic software math could compensate for once a simple 'as built in circuit' transformer Vs a resistor divider based comparison to define the actual real world working circuits variances has been done assuming of course the resistor divider circuit has absolutely zero variants of its own which is highly unlikely as well.

Which if my math is correct a 12 bit ADC would have a at best ~5.27 arc minute resolution which is more than the typical theoretical no secondary load variances a common properly functioning transformer would likely exhibit.

Or am I forgetting something? It's been awhile since I had to do such phasing Vs A/D converter resolution comparison work.

Hence the necessity to define the needed limit of accuracy and precision for whatever measurements are being done Vs the realistic ranges of variances that they system under test may produce or induce in the measurements.

Which BTW even your hall effect device based sensing system is going to add to things and thusly need to be factored for.

Joined Oct 9, 2012
14
tcmtech, PTs that you are discussing exist. and I use them all the time at work. I buy them and some of them are $1K each. I have met the people who design transformers, and they are not all the same. The team at Flex-Core are sharp and experienced, and they put a lot of time and effort into designing a PT vs a control power transformer. In a way, its insulting to them to assume all transformers are the same. Saturation in a control power transformer is a good thing, it provides voltage regulation because its pushed into non-linearity on purpose. The magnetic properties of electrical steel are inherently non linear. I want to bottom line this for you, Stay on topic, the question that was asked was clear your taking this in a direction that you want, not what the OP was asking. #### crutschow Joined Mar 14, 2008 28,177 The magnetic properties of electrical steel are inherently non linear. Certainly the core can be non-linear but that does not affect the input to output voltage ratio of an unloaded transformer, which is determined by the turns ratio and the flux coupling between the primary and secondary windings, not whether the core is linear or not. A non-linear core will cause the magnetizing current to be non-linear with the voltage, but doesn't affect the output voltage since that doesn't affect the flux. #### Sensacell Joined Jun 19, 2012 2,917 I want to bottom line this for you, Stay on topic, the question that was asked was clear your taking this in a direction that you want, not what the OP was asking. With all due respect, if I focused on responding to only what most original posters requested, I would be denying the communal creative process- a process that often leads to better, cheaper and faster solutions. Open your mind to new ways of seeing things. #### tcmtech Joined Nov 4, 2013 2,867 Open your mind to new ways of seeing things. We are hence the repeated asking for explanations of what it is you are doing and to what justifications it has. Right now to us your doing nothing but throwing words and technical terms around to justify something which to us as long experienced electrical technicians and the like have very little reason to believe is a valid issue without solid data to show us where we are wrong or mistaken in our views. Especially so given people like me have in fact worked with transformers in applications where needing to know the phase angle variances from the primary to secondary within +- 1 degree or less were critical and never found a common generic application transformer working primarily in a voltage isolation and sensing application with no discernible load on the secondary to be of any concern. Even with very low end ones I pulled out of power packs I got f a$1 from garage sales at that!

Believe me as someone who has designed grid tie power inverters knowing this sort of details in order to determine zero cross deadbands and other such technical switching device timing and related details in order to keep a power system stable and within operating specs for power co gen work is relevant to a design.
Critical enough that when I designed my first units I literally set up test systems to measure and confirm such issues and I found so little degree of discernible variances in phase angle delays between a cheap sub 10 VA wall wart and a high quality utility grade PT that it was a no brainer to design things around using common small general purpose transformers for such phase timing work. There was simply not enough variations to justify any concerns.

That's why I keep asking for actual data to show me where I am supposedly wrong. Without solid documented proof that a common general purpose transformer has some unknown to me variances beyond those of a PT transformer in your applications I have zero reason to believe you to have the justification to declare it to be a true issue in this application.
Especially since you have yet to show us any documentation to prove that using a common generic transformer or a hall effect based current sensing device in an application to generate a voltage signal does not display similar phasing delay nonlinearity characteristics as well in the given application. Nor have actually defined what it is you are even doing let alone the realistic precision and accuracy required for it despite having been asked multiple times now.

The thing is for us to 'open our minds to anew way of seeing things'we need documented proof (something to actually see that's real and not just imagined theory in your head) showing the realistic as used in service variances between your \$1000 PT units and common generic transformers and your hall effect based sensing concept or its just unjustified excuses to us.

Extraordinary claims require extraordinary proof and thusly no data equals no proof of legitimate concern to people like us. The balls in your court so if you claiming you can make a slam dunk form the three point line you had better get to doing it or we will keep calling your bluff. We want to see your proof and justification in order to learn something new and expand our minds.

#### tcmtech

Joined Nov 4, 2013
2,867
Certainly the core can be non-linear but that does not affect the input to output voltage ratio of an unloaded transformer, which is determined by the turns ratio and the flux coupling between the primary and secondary windings, not whether the core is linear or not.
A non-linear core will cause the magnetizing current to be non-linear with the voltage, but doesn't affect the output voltage since that doesn't affect the flux.

The only aspects of this phenomenon I am aware of where it's a possible issue of concern is in variable frequency applications where the core may either be operating in the upper limits of its applicable frequency band not fixed 50/60 Hz power line operation where every harmonic frequency is easily determined and the non linear variances at said frequency are easily accounted for.
Or where the physical design of the unit creates a odd peak or dip in the overall operational bandwidth characteristics like when doing a sweep frequency test and an unexpected peak or dip shows up in the response curves that fall outside of the anticipated one.

https://en.wikipedia.org/wiki/Sweep_frequency_response_analysis

http://www.edn.com/design/component...sformer-at-a-frequency-it-wasn-t-designed-for

https://www.omicronenergy.com/filea...s/2011-03-PotM-SFRA-on-Power-Transformers.pdf

Still, with out the OP giving a single shred of credible data to justify his concern or any explaining as to what he is doing and why let alone to what degrees of precision and accuracy are justified its all just coming off as a bunch of theory based run around to avoid having to admit the concerns are totally unjustified and or impossible to account for.

Personally I think he's at best theory guy who's enamored with the tern 'hall effect' and unrealistic tolerances because he sound important and look good paper but has very little practical knowledge of what he's doing and thusly can't produce any credible justifiable proof as to why he needs such levels so accuracy and precision in measurement to begin with. Especially for a arduino based project of yet to be defined intent.

Just a guess until shown otherwise.