Mains 120 Voltage through full bridge rectifier could you explain the following?

[JJ_023]

So I am taking 120V mains and running it through a full bridge rectifier.

This is what I get from my multimeter.

The following readings are at the output of the bridge rectifier.

On VDC reads 114

On VAC reads 35

On AC + DC reads 124

On dual display DC reads 114 and AC reads 50

I assume the last 2 readings are not the same because of a phase shift.  Am I interpreting this correctly?

[IanB]

Have you put a load on the output of the bridge rectifier? You won't read good voltages unless you make current pass through the diodes. Any kind of small load like a low power bulb will do.

When you say the last two readings are not the same, do you mean 35 V AC does not match up with 50 V AC?

[MrAl]

So I am taking 120V mains and running it through a full bridge rectifier.

This is what I get from my multimeter.

The following readings are at the output of the bridge rectifier.

On VDC reads 114

On VAC reads 35

On AC + DC reads 124

On dual display DC reads 114 and AC reads 50

I assume the last 2 readings are not the same because of a phase shift.  Am I interpreting this correctly?

Hello there,

You can't really measure just the output of the bridge you first have to measure the input of the bridge.  That's because the input line could be something other than 120vac. The normal range is 108vac to 132vac so you could possibly have anything between those two.  Normally you would get something close to 120vac, but you should use your meter for that anyway so you will know how your meter interprets that voltage.  It could also vary significantly over the course of one day.

We can look at some theoretical data, but your meter may show some unusual readings because meters respond differently to different harmonics, and there are a bunch of harmonics present in the output of the bridge.
Most notable, you do not see the same frequency on the output as on the input because the fundamental becomes two times the input frequency.  That means the lowest frequency, with a line input of 60Hz, is 120Hz.  Most meters can respond to that though, but you never know without checking the spec's of the meter.  The amplitude is about 72 volts peak which means about 50.9vrms, but there are also harmonics that add up to more, plus the DC component.  The average DC however is around 108v.

Normally you have some sort of load on it too, and also usually some capacitance that can filter the bumps into a more smooth DC output.  With little load you would see the peak voltage which would be close to 170v peak when you have a capacitor to filter the output.
You have to be careful with the selection of the capacitor though, it has to be able to handle the full peak of 170v and preferably with a rating 250v or more.

Different meters respond differently to frequencies not in the 60Hz to maybe 400Hz so it's hard to say what the meter will read exactly.  It is however possible to calculate all that if we know the input voltage for sure.

[JJ_023]

Have you put a load on the output of the bridge rectifier? You won't read good voltages unless you make current pass through the diodes. Any kind of small load like a low power bulb will do.

When you say the last two readings are not the same, do you mean 35 V AC does not match up with 50 V AC?

I have not put a load on the rectifier.

That is exactly what I mean.

[JJ_023]

Hello there,

You can't really measure just the output of the bridge you first have to measure the input of the bridge.  That's because the input line could be something other than 120vac. The normal range is 108vac to 132vac so you could possibly have anything between those two.  Normally you would get something close to 120vac, but you should use your meter for that anyway so you will know how your meter interprets that voltage.  It could also vary significantly over the course of one day.

The input at the bridge rectifier is about 122 V.

[Ian.M]

Without knowing the exact make and model of your meter, we have no way of knowing what its going to do when presented with full wave rectified, unsmoothed mains.  We suspect that on DC it will read fairly close to the average, but on AC it may have issues with the very non-sinusoidal waveform.  e.g. assuming its a true RMS meter, what's the maximum crest factor its specified for?

I've attached a simple LTspice simulator test jig to help you visualise the waveforms and see what may be going on.

N.B. the test jig simulation uses a floating AC source, e.g. in real life the secondary of an isolation transformer, so its outputs can be measured with respect to simulator ground.  Your mains supply Neutral is grounded, so without an isolating transformer, both DC terminals of the bridge rectifier have a large pulsing voltage with respect to ground.

Do *NOT* attempt to connect any ground referenced test equipment across the real bridge rectifier output!

[bdunham7]

On AC + DC reads 124

On dual display DC reads 114 and AC reads 50

I assume the last 2 readings are not the same because of a phase shift.  Am I interpreting this correctly?

I'm not sure what you mean, what phase shift?  TRMS AC + DC = (AC² + DC²)^0.5 and this works out for these numbers.

Your overall numbers as a whole may be a bit off because the input capacitance of the meter is smoothing out the rectified DC a bit--the only load is the 10M or so input impedance of the meter.  A very small load (10k or so) would give you numbers more in line with basic theory.  122VAC into a bridge implies about 109VDC (less any diode drops) output.  This is because the AC can be RMS, while the DC is always average.  It would help to know what meter you have.

[ejeffrey]

For an ideal rectified waveform the AC+DC true RMS voltage will be the same as the input signal.  For real diodes it will be a bit lower due to Vf

However this is only the case if the output tracks the input, which requires the diodes to stay conducting.  Otherwise the voltage at the zero crossings will "float" and not quite reach zero.  This will increase Vrms and decrease Vac.  The logical extreme of this is a linear power supply where the capacitor stores enough charge that the voltage only drops a few volts between peaks and Vrms of the rectified signal is nearly vmax of the input.

My guess is that the discrepancy is because the AC and dual display modes are a different load. Since you have no designed load you just have the multimeter input impedance.  In dual display mode the input is configured as for DC, and is a straight 10 megaohm load.  In AC volts mode the input is capacitively coupled. I guess that difference accounts for the differing result.

I would guess that if you put a 1 megaohm resistor across the rectifier outputs the discrepancy will go away.

[MrAl]

Hello there,

You can't really measure just the output of the bridge you first have to measure the input of the bridge.  That's because the input line could be something other than 120vac. The normal range is 108vac to 132vac so you could possibly have anything between those two.  Normally you would get something close to 120vac, but you should use your meter for that anyway so you will know how your meter interprets that voltage.  It could also vary significantly over the course of one day.

The input at the bridge rectifier is about 122 V.

Hello again,

Ok then you would multiply the theoretical values by 122/120.

I have a question though.  What is going to be the purpose for measuring this quantity?  Normally you don't really measure that until you have some load or some filter capacitance or something else like that.  Is this a study of some type?

[CaptDon]

The output of the bridge is pulsating d.c. and as such it isn't A.C. or smoothly filtered D.C. so your measurements are basically meaningless. If you place a suitably rated capacitor across the D.C. side of the bridge it should charge to 1.41 times the RMS A.C. input voltage, in your case 122 X 1.41 =172 V.D.C. A bridge rectifier is seldom used without a filter capacitor on the D.C. output side so your experiment is basically bogus.

[JJ_023]

Without knowing the exact make and model of your meter, we have no way of knowing what its going to do when presented with full wave rectified, unsmoothed mains.  We suspect that on DC it will read fairly close to the average, but on AC it may have issues with the very non-sinusoidal waveform.  e.g. assuming its a true RMS meter, what's the maximum crest factor its specified for?

My oversight, I should have included the make and model.

I have no idea what the maximum crest factor of this meter is.

This is the meter that was used.

https://www.cem-instruments.in/product.php?pname=DT-9979

I've attached a simple LTspice simulator test jig to help you visualise the waveforms and see what may be going on.

N.B. the test jig simulation uses a floating AC source, e.g. in real life the secondary of an isolation transformer, so its outputs can be measured with respect to simulator ground.  Your mains supply Neutral is grounded, so without an isolating transformer, both DC terminals of the bridge rectifier have a large pulsing voltage with respect to ground.

Hey thank you for doing that.  I am actually getting more into LT Spice and Q spice.  It would be awesome to see how you did all that.   

I am actually going to be looking into creating a 60 V DC source and adding some ripple and noise to it in Spice.  I’ve seen a couple of ways of doing it in my research.  My understanding is you can write the parameters in a text file and have Spice emulate that. 

[JJ_023]

On AC + DC reads 124

On dual display DC reads 114 and AC reads 50

I assume the last 2 readings are not the same because of a phase shift.  Am I interpreting this correctly?

I'm not sure what you mean, what phase shift?  TRMS AC + DC = (AC² + DC²)^0.5 and this works out for these numbers.

Your overall numbers as a whole may be a bit off because the input capacitance of the meter is smoothing out the rectified DC a bit--the only load is the 10M or so input impedance of the meter.  A very small load (10k or so) would give you numbers more in line with basic theory.  122VAC into a bridge implies about 109VDC (less any diode drops) output.  This is because the AC can be RMS, while the DC is always average.  It would help to know what meter you have.

Thank you for that.

The meter used was a  https://www.cem-instruments.in/product.php?pname=DT-9979

On AC + DC setting it reads 124V

On dual display DC reads 114 and AC reads 50

That is the part that is not making sense to me as 114V + 50V do not add up to 124V.

[JJ_023]

For an ideal rectified waveform the AC+DC true RMS voltage will be the same as the input signal.  For real diodes it will be a bit lower due to Vf
However this is only the case if the output tracks the input, which requires the diodes to stay conducting.  Otherwise the voltage at the zero crossings will "float" and not quite reach zero.  This will increase Vrms and decrease Vac.  The logical extreme of this is a linear power supply where the capacitor stores enough charge that the voltage only drops a few volts between peaks and Vrms of the rectified signal is nearly vmax of the input.

This stuff I know.

My guess is that the discrepancy is because the AC and dual display modes are a different load. Since you have no designed load you just have the multimeter input impedance.  In dual display mode the input is configured as for DC, and is a straight 10 megaohm load.  In AC volts mode the input is capacitively coupled. I guess that difference accounts for the differing result.

I would guess that if you put a 1 megaohm resistor across the rectifier outputs the discrepancy will go away.

This is similar to what I was looking for is an answer as there is a discrepancy between the 2 settings on the multimeter.

Thank you. 

I will try putting a 1 Meg resistor and see what happens.  Thank you for the suggestion.

[JJ_023]

Hello there,

You can't really measure just the output of the bridge you first have to measure the input of the bridge.  That's because the input line could be something other than 120vac. The normal range is 108vac to 132vac so you could possibly have anything between those two.  Normally you would get something close to 120vac, but you should use your meter for that anyway so you will know how your meter interprets that voltage.  It could also vary significantly over the course of one day.

The input at the bridge rectifier is about 122 V.

Hello again,

Ok then you would multiply the theoretical values by 122/120.

I have a question though.  What is going to be the purpose for measuring this quantity?  Normally you don't really measure that until you have some load or some filter capacitance or something else like that.  Is this a study of some type?

I was just honestly playing around there is no particular purpose. 

I have  a bridge rectifier that is set up to go into mains directly.  So all I really have to do is just plug it into the wall and everything is insulated and there are jacks on the output for the multimeter.  Because of me playing around  I now have this question I am trying to figure out with this particular meter.

[JJ_023]

The output of the bridge is pulsating d.c. and as such it isn't A.C. or smoothly filtered D.C. so your measurements are basically meaningless. If you place a suitably rated capacitor across the D.C. side of the bridge it should charge to 1.41 times the RMS A.C. input voltage, in your case 122 X 1.41 =172 V.D.C. A bridge rectifier is seldom used without a filter capacitor on the D.C. output side so your experiment is basically bogus.

I understand everything that you are saying.  I am just trying to figure out the discrepancy in the meter.  As someone already suggested I am going to put a 1Meg load and see if it evens out.

[bdunham7]

That is the part that is not making sense to me as 114V + 50V do not add up to 124V.

It does though, if you add them properly as per the formula I posted.  114² = 12996, 50² = 2500.  Then 12996 + 2500 = 15496, the square root of that is 124.5.

[macboy]

Maybe the LTSpice sim shown below will illustrate something for you.

After the bridge rectifier, you will have the absolute value of the sine wave of about 170 V peak. The meter will measure the average of this for the DC value.
The meter will insert a capacitor for AC measurement to block the DC component, then measure the RMS value of the result. It is also possible that the DC is removed mathematically instead, if it is a high speed sampling type (even some cheap meters are), but the result is the same.

This sim shows both the DC-coupled average and AC-coupled RMS. Note these measurements in the "error log".

I didn't attach the .asc because it is a trivial circuit.

If the meter is showing something too low in only AC mode (vs AC+DC), then try manually setting the range to the next higher one. Note the AC-coupled waveform (green) is exceeding -100 V even though the RMS is only around 50 V. If the meter is internally clipping the waveform due to not selecting a high enough range, then it will return the wrong value.

[JJ_023]

That is the part that is not making sense to me as 114V + 50V do not add up to 124V.

It does though, if you add them properly as per the formula I posted.  114² = 12996, 50² = 2500.  Then 12996 + 2500 = 15496, the square root of that is 124.5.

So I redid everything with a little bit more information.

Same exact meter which has a multiline display and the setting where it could measure both AC and DC on screen at the same time.

Before the bridge rectifier coming out of the Mains.

On AC Setting = 124V
On DC Setting = 0 V  (very minute DC fluctuations)
Then utilizing the setting of the multimeter that combines AC + DC = 124V
Then utilizing a separate setting that is multiline shows the following.  AC = 124 DC = 0

So far so good. Repeating the same exact procedure after the bridge rectifier.

On AC Setting = 38 V
On DC Setting  = 111 V
Then utilizing the setting of the multimeter that combines AC + DC = 126V
Then utilizing a separate setting that is multiline shows the following.  AC = 52 DC = 111

========================================
So I understand the math and everything else that was explained up to this point.  What I am still baffled by is on one setting (after the bridge rectifier) on AC it shows 38 V and on a different setting it shows 52 V AC.

[JJ_023]

Maybe the LTSpice sim shown below will illustrate something for you.

After the bridge rectifier, you will have the absolute value of the sine wave of about 170 V peak. The meter will measure the average of this for the DC value.
The meter will insert a capacitor for AC measurement to block the DC component, then measure the RMS value of the result. It is also possible that the DC is removed mathematically instead, if it is a high speed sampling type (even some cheap meters are), but the result is the same.

This sim shows both the DC-coupled average and AC-coupled RMS. Note these measurements in the "error log".

I didn't attach the .asc because it is a trivial circuit.

If the meter is showing something too low in only AC mode (vs AC+DC), then try manually setting the range to the next higher one. Note the AC-coupled waveform (green) is exceeding -100 V even though the RMS is only around 50 V. If the meter is internally clipping the waveform due to not selecting a high enough range, then it will return the wrong value.

Thank you for the explanation as well as the time you put into the circuit.  I know how the circuit that I am utilizing works and what comes in and what comes out, it was the discrepancy between the readings that I could not understand.  I apologize to everybody for clearly not being able to express my question.

So I redid the experiment based on your suggestion going to a higher voltage setting and it now reads 51 V that is on the 500V setting. On the 50V setting (which it auto arranged to and then I confirmed it by manually putting it into that range) it reads 37. 

This is complete nonsense.  lol

[bdunham7]

This is complete nonsense.  lol

It's a deficiency in the autoranging system of the meter, or perhaps a failure of the autoranging to take into account deficiencies in the AC signal input system.  Things like this separate the proper instruments from the pretenders.

[TimFox]

Side note:  a typical voltage for a DC motor in US industrial catalogs is 90 V, which is approximately the mean value of the full-wave rectified unfiltered result from 110 to 120 V rms AC voltage.

[JJ_023]

This is complete nonsense.  lol

It's a deficiency in the autoranging system of the meter, or perhaps a failure of the autoranging to take into account deficiencies in the AC signal input system.  Things like this separate the proper instruments from the pretenders.

I think it is a deficiency in the autoranging system because if you go up you get the right reading.  I checked it with 2 other meters.

The meter itself is 1 of my favorites.  The display is so easy to read, you can customize the colors so that it's vivid and bright in sunlight.  It's waterproof.  It's fairly rugged.  It has a rechargeable lithium-ion battery and lasts for ever on a single charge. 

[bdunham7]

I think it is a deficiency in the autoranging system because if you go up you get the right reading.

They're two sides of the same coin.  If a -100V peak makes it inaccurate on the 50V range, that implies that its full-scale crest factor capability is less than 2, which is suboptimal but not unheard of.  To compensate for that, the autoranging should have peak-detection to bump the range up when the peaks are beyond what the input can handle.  All meters eventually fail at this if you increase the CF to a high enough value, but this example is fairly poor performance IMO.  But now you know how to deal with it.

[JJ_023]

I think it is a deficiency in the autoranging system because if you go up you get the right reading.

They're two sides of the same coin.  If a -100V peak makes it inaccurate on the 50V range, that implies that its full-scale crest factor capability is less than 2, which is suboptimal but not unheard of.  To compensate for that, the autoranging should have peak-detection to bump the range up when the peaks are beyond what the input can handle.  All meters eventually fail at this if you increase the CF to a high enough value, but this example is fairly poor performance IMO.  But now you know how to deal with it.

I agree its poor performance and I honestly think it's unacceptable.  In hindsight I should have checked with my other meters, I mean I know how the circuit works I just couldn't figure out what the issue was.

Thank you and everybody else who gave their opinion.

[the_cake_is_a_lie]

A full bridge rectifier is an unregulated linear power supply. It's invalid to measure unregulated power supplies without a load. As in, connecting straight to the multimeter is invalid. I have a 15W 9VAC power supply that merely contains a transformer. Under no load, it measures 13-15VAC. My 15W 9VDC unregulated linear power supply measures 12-15VDC under no load. Measuring across a high wattage / low ohm resistor that draws ample current, the AC and DC power supplies read correctly. At least within 10% but the voltages on power supplies are nominal in any case.

Been said but rms voltage of AC with DC offset is sqrt(AC² + DC²). 114VDC and 35VAC yield 120Vrms, which is close enough to 124Vrms. AC measurement jumping up to 50VAC on dual display doesn't mean anything when you aren't measuring under load. As in, an unregulated 1A power won't measure correctly outputting under 1mA.

Note that switching mode power supplies are inherently regulated and can measure correctly under no load, i.e. directly connected to multimeter.

Phase shift between voltage and current affects the power readings, not the voltage readings.

Also been said but you should check the input voltage. I live in the US and my nominal 120Vrms supply out of the outlet is between 121-125Vrms when I measure it.