Oscilloscope oddities

[GadgetBoy]

So, I have this neat DSO138 that I got online, and I've been planning to do some practical demonstrations on AC rectification for my YouTube channel, but I'm running into something a bit weird. So, I have a frequency generator producing a nice 100Hz 8Vpp sine wave (first picture). Now I hooked up the generator to a 1n4007 diode with a 100ohm load resistor, and I don't get a nice half-wave rectified waveform (instead I get picture #2). It's been quite a while since I played with an oscilloscope, so I imagine I'm doing something wrong - can anyone point me in the right direction for establishing a good half wave rectified waveform (and subsequently full wave rectified waveform)?

Thanks!

Gadgetboy

Sent from my ONEPLUS A3000 using Tapatalk

[Mr. Scram]

It looks a bit like it's hitting the breakdown voltage on both sides and conducting again. Are you sure your diode is good? Note that though the DSO138 is a neat little thing, it's limitations can sometimes cause effects to show that aren't really there. It can be difficult to distinguish what's what if you don't have a "real" DSO at hand. Aliasing would be one example where you can get into trouble with the DSO138.

[Ian.M]

Try a 1N4148 + a 10K load as I suspect you are overloading the signal generator.

[GadgetBoy]

It looks a bit like it's hitting the breakdown voltage on both sides and conducting again. Are you sure your diode is good? Note that though the DSO138 is a neat little thing, it's limitations can sometimes cause effects to show that aren't really there. It can be difficult to distinguish what's what if you don't have a "real" DSO at hand. Aliasing would be one example where you can get into trouble with the DSO138.

Didn't think to test the diode, but it's fine. Tried a different diode just in case - with similar results. Tried a 1k resistor, and the waveform is nearly complete with just a little nipped off of the bottom curves. First picture is waveform across the 1k. Second is waveform across the diode.

Sent from my ONEPLUS A3000 using Tapatalk

[JS]

The load might be too much for your signal gen.

You have 4V peak, if your sig gen is 50Ω it should output 2.6V peak into 100Ω, take the 2V- from the rectifier you should be seeing at least 0.6V peaks. Also, the generator needs to source a few mA.

Try taking the load out, or use a much bigger one if the signal isn't that clear. 1k to 10k should do.

Also, with a low voltage AC you won't get a very nice rectified signal, something more like what you are seeing, increasing source voltage, decreasing the load, using lower voltage diodes (Ge or schokty) would make things better.

JS

PS: too many fast replies, just to add, if you can't generate higher voltage you could use a mains AC transformer and take the output from there, sine could be a little distorted as those tend to saturate at the peaks but the picture might end better than with your config.

[GadgetBoy]

Try a 1N4148 + a 10K load as I suspect you are overloading the signal generator.

I just get this.

Sent from my ONEPLUS A3000 using Tapatalk

[GadgetBoy]

The load might be too much for your signal gen.

You have 4V peak, if your sig gen is 50Ω it should output 2.6V peak into 100Ω, take the 2V- from the rectifier you should be seeing at least 0.6V peaks. Also, the generator needs to source a few mA.

Try taking the load out, or use a much bigger one if the signal isn't that clear. 1k to 10k should do.

Also, with a low voltage AC you won't get a very nice rectified signal, something more like what you are seeing, increasing source voltage, decreasing the load, using lower voltage diodes (Ge or schokty) would make things better.

JS

PS: too many fast replies, just to add, if you can't generate higher voltage you could use a mains AC transformer and take the output from there, sine could be a little distorted as those tend to saturate at the peaks but the picture might end better than with your config.

I tried a bigger load resistor, all I got was a sine wave with slightly flattened bottom peaks. Doesn't matter what diode I try, I've tried signal diodes, Schottky diodes, and standard silicon. I don't have any ge diodes on hand, unfortunately.

Sent from my ONEPLUS A3000 using Tapatalk

[JS]

You should be using dc coupled, that shape is pretty much what you will get from this circuit at this levels, with dc coupled you will see it's a rectified waveform with the sharp bottom ripped off for the effect of the diode drops.

If you actually want to capture the text book waveform of a fully rectified sinewave look for a precision rectifier, which involves two opamps and two diodes. You could actually get a better video than what you were expecting, as you can show that diode drop issue and the ideal waveform.

JS

[alsetalokin4017]

Presumably you are going for something like this?

[Mr. Scram]

Note that displaying a fully rectified wave on a regular oscilloscope it troublesome. As far as I know you need an active probe for that, or apply a bit of math function trickery. A half wave should be possible, though.

[JohnPen]

I did a quick set up to check/modify your results.  Using a FY3200S to produce an 8v sine (100 Hz) connected to a germanium diode (OA47) followed by a 10K load.  This was monitored by a X10 probe on a DS1054 unlocked set for DC coupling.  The resulting trace showed a very close to half wave rectification.  4.8 horizontal divisions rectified sine to 5.2 divisions ~zero volts.  The amplitude of the half sine reduces by the diodes forward voltage drop.  This should be close enough to demonstrate half wave rectification in a simple way. Also fit a switch to reverse the diode to enhance the demo.

[rsjsouza]

How are you making the connections? Do you have a diagram?

Using a simple 1N4007 and a 1k resistor connected as below, I get the following waveform:

              1N4007                 DS4014
       ┌────────►│───────┐←───────┐ ╔═════════════╗
       │   ↑             │        │ ║ ┌───┐    ☼  ║
HP3314 │   │             <        │ ║ │   │ º  º  ║
 4Vpp (~)  │             > R      │ ║ └───┘ ☼  ☼  ║
 100Hz │   │             < 1kΩ    │ ║  ch3    ch1 ║
       │   │             >        │ ╚═════════════╝
       │   │             │        └───┘+ │-  │+ │-
       └───┼─────────────┘←──────────────┘   │  │
           │   ↑                             │  │
           │   └─────────────────────────────┼──┘
           │                                 │
           └─────────────────────────────────┘

[Ian.M]

Note that displaying a fully rectified wave on a regular oscilloscope it troublesome. As far as I know you need an active probe for that, or apply a bit of math function trickery. A half wave should be possible, though.

If you are using a bench scope with conventional probes and a bench signal generator, you need a coupling transformer as the signal generator output to the rectifier must be fully floating so the negative side ot the DC output can be grounded without shorting half the diodes in the bridge.  For demo purposes with small signals a 600R line isolating transformer from a dead modem will probably do the job as long as you don't let it saturate.   Otherwise, you need to use differential probing techniques and avoid grounding the negative side of the DC output.

[tester43]

I am guessing: you will not see zero on AC because of minimum forward voltage of the diode.

[JohnPen]

An extra thought if you have a function generator like the FYxxx series which has 2 separate independent outputs. Using both outputs on say 100 Hz, with 8v sine waves on each, set one output offset by 180 degrees.  Connect the 2 outputs to a bridge rectifier's AC inputs,  load the output of the rectifier with a 10K load and monitor rectified 200 Hz with your scope.

[Ian.M]

An extra thought if you have a function generator like the FYxxx series which has 2 separate independent outputs. Using both outputs on say 100 Hz, with 8v sine waves on each, set one output offset by 180 degrees.  Connect the 2 outputs to a bridge rectifier's AC inputs,  load the output of the rectifier with a 10K load and monitor with your scope.

That setup is equivalent to a transformer with a 16V secondary with its center tap grounded.  It will give equal positive and negative DC rails out from the bridge rectifier symmetrical with respect to ground, and you myst not ground the bridge rectifier negative terminal.   If you only want to demo full wave rectification, *DON'T* use a bridge rectifier.  All you need is two diodes, one from each output, cathodes towards the common load.

[DannyTheGhost]

I, as an owner of DSO138, suggest you to use DC mode instead. In AC mode input there is capacitor in series, so you won't see any DC part of voltage (in your case, rectified part of your wave).

[JohnPen]

Oops! My enthusiasm jumped to far.  Agreed I should NOT have used a bridge rectifier especially as my set up does have a common ground connection for my test equipment. Fortunately in this set up it only loads the generator outputs rather badly without any lasting effect.  Real world AC it would be rather different.   The better way to show things would be a small low voltage mains transformer, say 6vac,  and a switched selection for each category of rectification using a 10K load.

[alsetalokin4017]

Well, if one is trying to demonstrate such things for students, or even for a technical YT audience... perhaps using a toy oscilloscope and an unknown FG isn't the best way to go. On the other hand if one is trying to demonstrate the features and capabilities of the toy scope, perhaps characterizing a FWB, or even two diodes, isn't the best way to do it.

Careful study of the scopeshot I posted above will reveal many things, both about the rectification function of the diodes and about using an oscilloscope to display the phenomena.

By the way the scopeshot was obtained using a packaged FWB and an analog FG.

[IanMacdonald]

If your sig gen output is capacitor coupled, placing a diode in series will cause the series capacitor to become negatively charged until no more current flows. Which could explain your lack of a waveform.

[GadgetBoy]

Well, I gave up using the signal generator (Hiland DDS signal generator), and muntzed an old transformer based DC power brick. The results were exactly what I was looking for. I guess I'll be using this from now on, until I can afford more sophisticated equipment (I'm working on a low budget).

Sent from my ONEPLUS A3000 using Tapatalk

[JS]

That seems niceer, single wave rectifier makes a better picture, to go for full wave you see what you was getting, which is not too far off of what I would expect from your setup.

(I'm working on a low budget).

We can tell, and we've all been there, I never got one of those DSOs but I did have a nice sound card as I've worked on audio for a long time. A month or two ago, after being over 15 years into this, I got my first scope, a DS1054Z and is a blast. Only now I can justify the nice toys as I'm working with them and they end up paying for themselves. I wish the videos start paying for your toys, that would be nice!

JS

[GadgetBoy]

That seems niceer, single wave rectifier makes a better picture, to go for full wave you see what you was getting, which is not too far off of what I would expect from your setup.

(I'm working on a low budget).

We can tell, and we've all been there, I never got one of those DSOs but I did have a nice sound card as I've worked on audio for a long time. A month or two ago, after being over 15 years into this, I got my first scope, a DS1054Z and is a blast. Only now I can justify the nice toys as I'm working with them and they end up paying for themselves. I wish the videos start paying for your toys, that would be nice!

JS

The way it was initially setup (before I muntzed it), it was full wave rectified, and that waveform looked great, too.

Sent from my ONEPLUS A3000 using Tapatalk

[GadgetBoy]

In case anyone was curious, here's the lovely full wave rectified signal from the transformer I found.

Sent from my ONEPLUS A3000 using Tapatalk

[JS]

Yes, but now you have 32Vpp sinewave against 8Vpp, the diode drop is much smaller than the signal so it looks much closer to ideal.

JS