Full Wave Bridge Rectifier

[Legion]

Finally got all my gear today; scope, DC lab power supply and function generator. So I figured I'd go back and hook up some circuits that until now I could only analyze on paper.

I setup a half wave rectifier and it worked as expected. But when I put together a full wave rectifier I got almost the same results as the half wave. I'm not measuring the input signal and rectified signal at the same time. I measured the rectified signal, saved it as a reference, disconnected the scope and then measured the input signal on its own (not hooked up to the circuit) and overlayed it on top of the reference waveform.

Here's the circuit I built:


I should see twice the frequency of positive humps, but I don't. Instead I get this:


The white waveform is the rectified signal. The blue waveform is the input signal from the function generator. There are a few things going on that I don't understand.
1.) The input waveform amplitude is way off. The function generator is showing 2.03V output. But the scope shows just under 500mV pp.
2.) The rectified signal has only half the frequency it should.
3.) The rectified signal has a strange voltage drop at the right side of each hump.
4.) The rectified signal has an amplitude of ~1.5V. But shouldn't it be 2.03V - 1.4V = 0.63V because of the two 0.7V drops through the diodes?

If it helps I can attach a picture of the actual breadboarded circuit.

[w2aew]

It's because your signal generator and scope have a common ground. This shorts half of your bridge. In order to see the full wave rectified signal, you can put a 1:1 transformer between the function generator and the bridge.

[Legion]

It's because your signal generator and scope have a common ground. This shorts half of your bridge. In order to see the full wave rectified signal, you can put a 1:1 transformer between the function generator and the bridge.

Hmmm. I don't have a transformer. I'll have to pick one up.

Can you elaborate on how having a common ground shorts half the bridge?

[edy]

I looked it up and some have suggested it is because you are measuring your function generator output and rectified signal output using the same scope which shares a common ground between channels, so you are shorting out both sides of your circuit. The transformer isolates your function gen side from your rectifier side. Or you could try not measuring the function gen at the same time as your rectifier (use only 1 channel at a time and disconnect the other).

http://img827.imageshack.us/img827/3841/bridgew.jpg

In this image, you can see a scenario scope hookup but the common ground between channels seems to be the culprit.


 

[Legion]

I should clarify then. I'm not measuring the input signal and the rectified signal at the same time. I tried that and it didn't work. Instead, what I did was measure rectified signal on it's own, save it as a reference. Then I measured the input signal and overlayed it on top of the rectified reference. I made sure both were using the same volts/div (500mv/div) and time/div (200us/div).

[w2aew]

I should clarify then. I'm not measuring the input signal and the rectified signal at the same time. I tried that and it didn't work. Instead, what I did was measure rectified signal on it's own, save it as a reference. Then I measured the input signal and overlayed it on top of the rectified reference. I made sure both were using the same volts/div (500mv/div) and time/div (200us/div).

Even still - the output of your function generator is likely a ground-referenced output. In other words, the sinewave is generated with respect to ground.  Thus, when you connect it to your bridge, one of the two connections is your sinewave voltage, and the other is ground.  When you connect your scope, one of the two connections you make to the bridge is also ground.  Thus, there are two different points in your bridge that are both connected to ground - thus they're effectively connected together (i.e. shorting the components between them).

[Legion]

I should clarify then. I'm not measuring the input signal and the rectified signal at the same time. I tried that and it didn't work. Instead, what I did was measure rectified signal on it's own, save it as a reference. Then I measured the input signal and overlayed it on top of the rectified reference. I made sure both were using the same volts/div (500mv/div) and time/div (200us/div).

Even still - the output of your function generator is likely a ground-referenced output. In other words, the sinewave is generated with respect to ground.  Thus, when you connect it to your bridge, one of the two connections is your sinewave voltage, and the other is ground.  When you connect your scope, one of the two connections you make to the bridge is also ground.  Thus, there are two different points in your bridge that are both connected to ground - thus they're effectively connected together (i.e. shorting the components between them).

Thanks! I get it now. Can you tell me what kind of 1:1 transformer I'd need? I was looking on digikey and mouser and there are thousands of them.

[440roadrunner]

It's because

By  the way,  mr  W2AEW,    I just  enjoy the  heck out of your  YouTube vids.

At  65,  I'm way  behind on  so many things

[commongrounder]

If you can deal with low frequency stuff to start out with (i.e. audio frequencies), you could just get one of those audio isolation transformers that are used to eliminate ground loops in car and home audio systems.  They typically have RCA in and out plug/jacks.  I use one every so often to isolate a DUT input from other test equipment.  There will be a maximum limit to the output, depending on the one you get, but they are inexpensive.

[Legion]

I tried a few things to be able to get a good measurement. None have been successful.

First I tried hooking up the scope probe ground to a common node with the function generator's output. I connected the scope ground to negative terminal of the circuit which is also where the negative side of the function generator connects. I thought this would prevent having a route to ground half way through the circuit. This didn't work. Instead I just got a slightly different half wave rectified signal.

For the second attempt, I found a document from some college's website that talks about grounding issues with scopes. They described this exact situation and recommended hooking up both scope probes. Ground each to the common negative terminal of the circuit. Then hook up each probe on opposite sides of the resistor and subtract their measurements. This also did not work. Below is the image. Are isolation transformers or clipping the ground plug off my function generator the only options?

[Legion]

I tried making my own "isolation transformer" by clipping the ground pin off the power cord of the function generator. It kinda worked, sorta. I get this output:



The frequency is double the input's, so that's good. But it looks like there's two signals even though I only have one probe hooked up. And there's still those flat parts of the signal that make it look like half wave rectification.

[sarahMCML]

Try using both scope probes in differential mode when measuring the signal across across the load, as follows:

1) Set whichever channel has an invert function (usually channel 2) to 'Invert'.
2) Set  channel selector to 'Add Ch1+Ch2'.
3) Remove, or connect together the probes ground leads, and leave disconnected from your circuit.
4) Ensure that both probes are set to the same Volts/Div sensitivity.
5) Connect the probes across the load resistor, and see what you get!

Overlay that on your original input side trace.

Hope this helps.

[Legion]

Try using both scope probes in differential mode when measuring the signal across across the load, as follows:

1) Set whichever channel has an invert function (usually channel 2) to 'Invert'.
2) Set  channel selector to 'Add Ch1+Ch2'.
3) Remove, or connect together the probes ground leads, and leave disconnected from your circuit.
4) Ensure that both probes are set to the same Volts/Div sensitivity.
5) Connect the probes across the load resistor, and see what you get!

Overlay that on your original input side trace.

Hope this helps.

Thanks Sarah! I think that worked, or at least it's the closest so far. Here's what I got:



The yellow waveform is channel 1, blue is channel 2 (inverted) and purple is their sum. Also, unlike my earlier post, I've restored the ground plug for the function generator. There's still some weird stuff going on though.

1.) Why are the yellow and blue signals not sine waves?
2.) Should a full wave rectified signal have little 0 volt sections between each hump? Or (ideally), should each hump connect at a non-differentiable point (kinda like McDonald's arches)?
3.) Every other hump of the summed signal has a slightly higher amplitude. Is this normal for full wave rectifier bridges?
4.) The summed signal should have an amplitude of ~0.6V (2V input - 2x0.7V drop from the diodes), but it looks closer to 1V. This could just be from variances in the diodes, but I'm wondering if it's something else.

[Rudane]

Your input signals aren't very large. Diodes do not forward bias until you go above their forward bias threshold voltage. This can vary depending upon the diodes, but the rectifier datasheet will tell you what that voltage is. So, the output will be 0 volts until you go above the threshold and then the output voltage will climb to it's maximum. That's what you are observing.

[Legion]

Your input signals aren't very large. Diodes do not forward bias until you go above their forward bias threshold voltage. This can vary depending upon the diodes, but the rectifier datasheet will tell you what that voltage is. So, the output will be 0 volts until you go above the threshold and then the output voltage will climb to it's maximum. That's what you are observing.

These are 1N004 diodes, so they should start conducting at 0.7V.

[Rudane]

Everything looks good except those glitches in the waveforms. As to the sine wave question; they're rectified sine waves now, so they look exactly like what rectified sine waves should look like. Every other hump larger just means the diode voltage is slightly different for that pair. In my experience 0.7 Volts is just a theoretical number, I've never seen a diode actually reach that number, it's always lower. I think that's why the rectified waveform is 1 Volt.

[Legion]

Everything looks good except those glitches in the waveforms. As to the sine wave question; they're rectified sine waves now, so they look exactly like what rectified sine waves should look like. Every other hump larger just means the diode voltage is slightly different for that pair. In my experience 0.7 Volts is just a theoretical number, I've never seen a diode actually reach that number, it's always lower. I think that's why the rectified waveform is 1 Volt.

Thanks Rudane. You're right. I checked the diodes and their voltage drops are much less than 0.7V. There is also a small difference in voltage drop between the positive half cycle diode pair and the negative half cycle diode pair. This discrepancy matches what is shown in the purple waveform.

[sarahMCML]

Hi again,

As you have now seen, at the lowish AC voltage that you are applying to the circuit with your function generator your 1N4004 diode drops have a significant effect.
If you have any such, you could try four small Schottky diodes instead of the 1N4004's, or old germaniums, and note the difference!
If you have a low voltage mains transformer (anything between 6 to 12 Vac) or an equivalent ISOLATED AC output Wall Wart, you could use that instead of the function generator.
Just make sure that your load resistor can take the power that you will put through it!

Hope this helps.

[Zero999]

I tried making my own "isolation transformer" by clipping the ground pin off the power cord of the function generator.

That's very dangerous. The ground pin is required for electrical safety and should never be removed.

[RLSprouse]

I have been frustrated by this same issue, trying to understand why my scope would not show the full wave rectified signal you always see in tutorials and textbooks.

So, I am curious...  if I power my scope through a mains isolation transformer like  this one:

http://www.amazon.com/gp/product/B00006HPFH/ref=pd_luc_rh_sbs_01_01_t_img_lh?ie=UTF8&psc=1

would that solve the problem and prevent similar issues in the future?

  ~ Russ 

[Zero999]

That will not make any difference because the oscilloscope's input and signal generator's output will still be connected to the mains earth conductor.

You need a signal generator with an isolated output or an oscilloscope with an isolated input.

[mikerj]

That will not make any difference because the oscilloscope's input and signal generator's output will still be connected to the mains earth conductor.

The point is that by using an isolating transformer you can safely break the mains ground connection.

[mrkev]

Why not just use basically any transformer at the generator output. If you are in electronics, you almost certainly have at least one, type doesn't really matter that much.
If it's from switch mode converter, use frequency of few kHz. If it's mains transformer, use 50Hz (up to about 100Hz).
just use small load and measure the input current and output voltage first...

[Jay_Diddy_B]

Hi group,

No need for a transformer. You can use the scopes waveform math function to show you the desired result.

Here is a model that I have constructed in LTspice.



V1 and R2 represent the 50 Ohm output on the function generator. The output is connected to the junction of D1 and D3.

The junction of D2 and D3 is grounded. This connection, if this was a typical transformer power supply, would be floating.

If you ground the bottom of D4 by attaching the ground clip on the scope, you will short D4.

If you measure the waveforms at CH1 and CH2, with respect to ground and then use the maths function on the scope do perform CH1 - CH2 you will see the waveform that you are looking for.
This is called a differential measurement, You are measuring the difference between two points.

If you have a 4 channel scope you can also look at the waveform at the point marked ch3.

These are the waveforms that you should see.



Regards,

Jay_Diddy_B

[Jay_Diddy_B]

Hi,

I did the same measurement that I modelled using my DSOX3034, with the built in Wavegen.

Here are the measured waveforms:



Pink trace is the maths waveform CH1 - CH2

Regards,

Jay_Diddy_B

[RLSprouse]

Thanks, guys, this is all extremely enlightening...  My scope only has two channels, so I will try to figure out how to use the math function and get the two halves of the rectified output to display.

  ~ Russ