Can I use an audio transformer as AC voltage sensor

[Piet de Pad]

Dear reader,
I have a question about using audio transformers as 120Vac 60Hz line voltage sensors.

I know how to measure AC line voltages with dedicated AC voltage transformers like the ZMBT101B or the ZMBT107. To me, these transformers are nice but bulky for the small PCB's that I make. So I thought can I use the 600:600 Audio transformer transformers for this which exist in various sizes. They also exist in SMD formats.
My knowledge of transformer tells me that this should be possible as long the core is not saturated, I'm not so worried about the phase difference between input and output signal. It might be that the pri-sec isolation voltage is an issue.

Below you see some examples of these transformers, The two left are SMD audio transformers. The third from the left is a traditional 600:600 transformer, and the right is an AC voltage sensor. The schematic shows how I want to use it and that is including a series resistor. The transformer seer only 1mA of current.

My question is: does anyone have a good reason why it is better not to do this or better why it is possible but with some points to consider. Let me know in the comments.

Thanks

 

[KlausST]

Hi,

I didn´t even check function.

For me: safety first: I doubt that small (SMD) audio transformers can satisfy safety regulations.
You know there are rules for "functional isolation" and "safety isolation". You need to follow "safety isolation" regulations.

This is the first thing you have to check.

Klaus

 

[Piet de Pad]

Hi,

I didn´t even check function.

For me: safety first: I doubt that small (SMD) audio transformers can satisfy safety regulations.
You know there are rules for "functional isolation" and "safety isolation". You need to follow "safety isolation" regulations.

This is the first thing you have to check.

Klaus

You're jumping to conclusions, I may follow the isolation rules. If you call a ball a football, that doesn't mean you can't play and other game with it. You are right, you can't use it for Clay pigeon shooting

 

[barry]

You're jumping to conclusions, I may follow the isolation rules. If you call a ball a football, that doesn't mean you can't play and other game with it. You are right, you can't use it for Clay pigeon shooting

What a snot.

You specifically asked “does anyone have a good reason why it is better not to do this”. Klaus answered your question, and you respond with an arrogant post.

I, for one, will ignore your future posts.

 

[KlausST]

Hi,

Thank you for showing the world your attitude. So now everyone knows.
This way I´m sure many people feel invited to respond to your question.

Klaus

 

[D.A.(Tony)Stewart]

It is rated for std 3kV insulation voltages and is current limited if R1 doesn't fail.

The only problem I see is R1 Voltage Insulation rating will not be 3kV but rather more likely 250V.

There may be some minimum magnetization current and distortion effects .

Creepage from dust will significantly degrade the performance to insulation breakdown to lightning noise.

R1

 

[Piet de Pad]

It is rated for std 3kV insulation voltages and is current limited if R1 doesn't fail.

The only problem I see is R1 Voltage Insulation rating will not be 3kV but rather more likely 250V.

There may be some minimum magnetization current and distortion effects .

Creepage from dust will significantly degrade the performance to insulation breakdown to lightning noise.


View attachment 188948
R1

Thanks Tony for your thoughts on this.
So your concerns are possible linearity or distortion of the input reduced input voltage, dust collection in the open transformer, and the Insulation rating.
I have seen that insulation ratings are 500Vdc and up till 1500Vdc, but I agree with you that it might be an issue definitely to keep it in mind. If I connect these transformers to the neutral, the voltage will not be higher than 1Vac. When I add an RC snubber over the primary winding, it's also protected against peak voltages on the AC line.
Linearity I consider that less of an issue, but it is good that you mention it. These transformers have a wide bandwidth to trans audio signals. I assume that it will do that without to much THD.
The dust collection can be resolved by covering it. There are types that are sold and which are closed.
R1 will be implemented with two SMD R1206 resistors. May be I take a TH type.
I appreciate your ideas on this thanks.

 

[D.A.(Tony)Stewart]

That won't be bullet -proof as power meters can pass thru up to 8kV , but good luck.

 

[FvM]

I presume the primary ground symbol actually means neutral line. If so you shouldn't make assumptions about maximum voltage at this circuit node. Even if you do, safety regulations don't. It's treated as <= 150 V mains circuit with respective expected surge voltages and isolation requirements.

If you target to reinforced insulation (secondary side exposed), test voltage is 2700 Vac for overvoltage category II (standard for equipment inside buildings). Meters have usually Cat III requirement.

Standard 1206 resistors are typically rated for 300 VAC test voltage, two resistors don't provide sufficient surge capability.

 

[crutschow]

I would put one resistor in series with the hot side and one in series with the neutral side.
That way, even if the hot and neutral are accidentally reversed, the current is still limited to a safe level from either transformer input to earth ground.

 

[D.A.(Tony)Stewart]

Induced Lighting impulses are common mode. Using the PE gnd must be Neutral on your schematic and as suggested paired input resistors.

Otherwise it causes greater mismatch on impedance and converts a Vcm into Vdiff so it should only be used on primary side with a CM choke Y caps / aka line filter.

Perhaps not relevant here, but I remember solving a HIPOT problem where PowerOne supplies had moved from S.D. to Mexico with overlooked changes in safety gaps on passive primary parts. They all passed HIPOT at the factory yet failed in our shop.

I reasoned that our 48V secondary and 0V= PE grounding causes the differential switching transformer leakage to become AC grounded and all the way back to AC input we had multiple failure nodes on 1K units /mo. So I put a "stop ship" on their product until fixed. I also decided to modify our HIPOT tester to add current limiting between HV cap and probe but still permit > 100 uA. Then the failures became non-destructive.

 

[Piet de Pad]

I would put one resistor in series with the hot side and one in series with the neutral side.
That way, even if the hot and neutral are accidentally reversed, the current is still limited to a safe level from either transformer input to earth ground.

I like that tip [U]crutschow[/U] two resistors in series to reduce the voltage between line and neutral and secondary. I take that as a design idea.

--- Updated Feb 26, 2024 ---

I presume the primary ground symbol actually means neutral line. If so you shouldn't make assumptions about maximum voltage at this circuit node. Even if you do, safety regulations don't. It's treated as <= 150 V mains circuit with respective expected surge voltages and isolation requirements.

If you target to reinforced insulation (secondary side exposed), test voltage is 2700 Vac for overvoltage category II (standard for equipment inside buildings). Meters have usually Cat III requirement.

Standard 1206 resistors are typically rated for 300 VAC test voltage, two resistors don't provide sufficient surge capability.

Hi FvM That is correct it's not the earth it symbolizes level earth in practice the neutral.

--- Updated Feb 26, 2024 ---

Induced Lighting impulses are common mode. Using the PE gnd must be Neutral on your schematic and as suggested paired input resistors.

Otherwise it causes greater mismatch on impedance and converts a Vcm into Vdiff so it should only be used on primary side with a CM choke Y caps / aka line filter.

Perhaps not relevant here, but I remember solving a HIPOT problem where PowerOne supplies had moved from S.D. to Mexico with overlooked changes in safety gaps on passive primary parts. They all passed HIPOT at the factory yet failed in our shop.

I reasoned that our 48V secondary and 0V= PE grounding causes the differential switching transformer leakage to become AC grounded and all the way back to AC input we had multiple failure nodes on 1K units /mo. So I put a "stop ship" on their product until fixed. I also decided to modify our HIPOT tester to add current limiting between HV cap and probe but still permit > 100 uA. Then the failures became non-destructive.

I got your point on the paired resistors and the CMchoke thanks

 

[D.A.(Tony)Stewart]

I got your point on the paired resistors and the CMchoke thanks

I would define the CM choke function with the X,Y caps as -100 dB @ 1MHz in both CM and DM modes. Line filter ground current 0.5mA max and >3 kV Hipot rating.

 

[Piet de Pad]

Hi all,
I see that the main concern you have mentioned is the isolation voltage between the primary and secondary windings. I was not too worried about that, people die today for may stupid reasons, why should I be more Catholic than the people who kill us like politicians ......... however. Your responses have convinced me to pay more attention to this isolation voltage, and I may have found a solution for this problem.

I found a 100mH common mode choke coil winded on a EE8.3 core which measures 9x9mm. Who runs 1mA though the primary side of these chokes get 53mVp on the secondary side. An amplifier with 50x amplification which adds an offset of 2.5V gives a nice AC voltage between 0 and 5V ready to be sampled by a microcontroller. I haven't found the specification for these transformers yet, but they are widely used as chokes in led drivers. The primary and secondary coils are separated in the winding bobbin of the magnetic core, so I expect that the 1500V or more isolation voltage is no longer and issue. I added the schematic of the amplifier that adds the amplification and offset to this text. The same circuit can be used for ac current measurements.

I hope someone may like it.
Regards

Fig1 the amplifier Fig2 the common Choke EE8.3

 

[D.A.(Tony)Stewart]

Hi all,
I see that the main concern you have mentioned is the isolation voltage between the primary and secondary windings. I was not too worried about that, people die today for may stupid reasons, why should I be more Catholic than the people who kill us like politicians ......... however. Your responses have convinced me to pay more attention to this isolation voltage, and I may have found a solution for this problem.

I found a 100mH common mode choke coil winded on a EE8.3 core which measures 9x9mm. Who runs 1mA though the primary side of these chokes get 53mVp on the secondary side. An amplifier with 50x amplification which adds an offset of 2.5V gives a nice AC voltage between 0 and 5V ready to be sampled by a microcontroller. I haven't found the specification for these transformers yet, but they are widely used as chokes in led drivers. The primary and secondary coils are separated in the winding bobbin of the magnetic core, so I expect that the 1500V or more isolation voltage is no longer and issue. I added the schematic of the amplifier that adds the amplification and offset to this text. The same circuit can be used for ac current measurements.

I hope someone may like it.
Regards

View attachment 189004

View attachment 189002


Fig1 the amplifier Fig2 the common Choke EE8.3

another example of unsafe primary resistors without specs and a 35 ohm coil at 60 Hz attenuates nicely but not at it's highest frequency range. Yet that may reduce CM impulses by 60 dB . maybe you should stick with a voltage transformer until you know what you are doing.

 

[Piet de Pad]

another example of unsafe primary resistors without specs and a 35 ohm coil at 60 Hz attenuates nicely but not at it's highest frequency range. Yet that may reduce CM impulses by 60 dB . maybe you should stick with a voltage transformer until you know what you are doing.

I do know what I'm doing, but you maybe don't like it.

 

[D.A.(Tony)Stewart]

RegardsView attachment 189004

View attachment 189002


Fig1 the amplifier Fig2 the common Choke EE8.3

CM chokes that look like this photo may be rated for 500V, just like flame-proof 1 W resistors.
The 100 mH acts like 40 ohms reactance depending on grid frequency variations.
Most of the differential V protection will be your flame-proof resistors in series. More in series might be cheap protection.
If the 60 k resistors, Pd = 120 mW. The case temp. rise will be 125'C times the ratio of Pd/Prated.

Any common mode lightning noise test should cause some concern. Generally a Hipot test for 60 sec or 10% more for 1 second is mandatory for all products applied to both Line & Neutral with respect to PE gnd. Test Engineers often elect to use DC HIPOT at 1.414 times Vrms as permitted. These CM chokes although clever, cheap and small, this design should have protection from potential arc failures in this design.

 

[dick_freebird]

Question whether the windings are individually good for 500V
end-to-end, or the 500V is winding -to-winding. Looking back
disassembling old time low power audio boards for parts, the
600:8 ohm transformer primaries sure were fine wire and thin
varnish, with no extra interlayer insulator, so I would not be too
trusting.

You might be better off with the smallest wall wart (that's still
transformer style) xfmr you can find. At least somebody would
have tested it.

 

[Piet de Pad]

I

CM chokes that look like this photo may be rated for 500V, just like flame-proof 1 W resistors.
The 100 mH acts like 40 ohms reactance depending on grid frequency variations.
Most of the differential V protection will be your flame-proof resistors in series. More in series might be cheap protection.
If the 60 k resistors, Pd = 120 mW. The case temp. rise will be 125'C times the ratio of Pd/Prated.

Any common mode lightning noise test should cause some concern. Generally a Hipot test for 60 sec or 10% more for 1 second is mandatory for all products applied to both Line & Neutral with respect to PE gnd. Test Engineers often elect to use DC HIPOT at 1.414 times Vrms as permitted. These CM chokes although clever, cheap and small, this design should have protection from potential arc failures in this design.

use 1/4Watt resistors 200Vmax so Pd will not be the problem. What i see is that the HiPot are 1500V, but I admit the specs are difficult to get.

--- Updated Mar 4, 2024 ---

Question whether the windings are individually good for 500V
end-to-end, or the 500V is winding -to-winding. Looking back
disassembling old time low power audio boards for parts, the
600:8 ohm transformer primaries sure were fine wire and thin
varnish, with no extra interlayer insulator, so I would not be too
trusting.

You might be better off with the smallest wall wart (that's still
transformer style) xfmr you can find. At least somebody would
have tested it.

There will be hardly any winding to winding voltage. The primary side has only one volt. The question is the primary to secondary HiPot value. I see that many audio transformers are rated 500V. That will work but not be enough.