Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Simple diode test doesn't jive. What gives?

  1. Nov 8, 2013 #1
    Probably a stupid question but I'm not sure what to search for at this point.

    I'm trying to do a very simple diode test (half wave rectification I suppose it's called) using an ac source. The source is a 440hz signal off of my computer soundcard (generated and played via a .wave file). It's a balanced output soundcard, so I'm just using one of the 2 outputs. Either single balanced output connection shows up fine as an AC wave on my tektronix 465b scope. There's no dc offset or anything. I don't even need to connect the ground scope probe end to the soundcard ground at all. The reference seems to be adequately earth ground (according the the scope's assumption).

    side note: basically I'm trying to use the sound card as a bench signal generator of sine wave output to do some very fundemental testing of caps, diodes, transistors, etc. Could be my first mistake perhaps....

    Ok, but I'm not completing a circuit. I'm just testing voltage. So I add a IN4004 diode on the end (of the positive ac signal connection) and test the voltage. What I get is a smaller ac wave. I don't get just the positive portion of the ac wave.

    And even if I put on a resistive load I see current of 0-30mA (hopefully won't blow my soundcard!). I'm not sure how to measure that correctly. Does it have to be a reactive load???

    I feel lost. Please help me to comprehend!

    THANK YOU VERY MUCH Ahead of Time! I'm sure of the many here who are well versed in electronics can help me get over this simple hurdle! And I hope I'm not posting in the wrong place for this as well. I guess you could call it "homework" but I'm not in any formal school so Ill assume it's ok. Thanks.
  2. jcsd
  3. Nov 8, 2013 #2
    btw the source ac voltage is 0-5VAC (depending on computer volume I set... same with the mA's I mentionedl).

    I think I'm answering my own question a bit here in saying perhaps the diode doesn't prevent voltage, but rather current? And since I haven't completed the circuit no current is actually flowing and thus the diode doesn't behave as axiomatic.

    But I thought I checked it while I completed the circuit with a resistive load attached (probably not good for the soundcard but this is all about learning).
  4. Nov 8, 2013 #3
    When I say "testing" I don't mean I'm trying to verify if compenents are broken. I'm trying to "test" electronic theories using the components themselves.
  5. Nov 8, 2013 #4
    You will have to connect the diode in a circuit to get it to behave as expected. You will also need to connect your scope's reference terminal to the circuit to see the expected wave form.

    The diode will behave like a capacitor if one side it driven by a voltage source and the other is connected to a high impedance. You must develop a forward voltage on the diode to cause it to conduct in a circuit. If you don't complete the circuit, the anode and cathode of the diode will remain at the came voltage and you will simply see the PN junction as a capacitor.

    Connect the anode to your + output on your sound card and the cathode to a resistor and then the other side of the resistor to the - output on the sound card. Measure across the resistor to see the waveform you are expecting. If you reverse the diode you should see the voltage on the resistor reverse. Your sound card should be able to drive an 8 ohm load but I would use 100 ohms or more just to play it safe.
  6. Nov 8, 2013 #5
    ScienceGeyser, thank you for the great information. That gives me a logical path to take.
  7. Nov 8, 2013 #6
    Well ok but (ScienseGeyser) how does a bench signal generator differ. Does it also need the circuit completed and an 8-100ohm imepedance load?
  8. Nov 8, 2013 #7
    Or more simplistically put... how do I start testing/validating ac circuits (without using 120VAC). I've repaired 120VAC ampifiers before but I don't want to use that for testing my basic understanding of basica componenents.
  9. Nov 8, 2013 #8
    A standard signal generator should have the ability to drive a 50 ohm load without drawing too much current. When connected to a higher impedance load, like an oscilloscope, the output will simply supply less current to the load. The output of the signal generator is a voltage source. If you draw too much current, the output will be distorted. The same applies to your sound card.

    Component testing requires a bit more circuit design than just connecting the component between a signal source and an oscilloscope. You will need to consider the characteristics of the device and which ones you intend to measure to know how to hook it up for testing.

    For example: if you connect a 50 ohm resistor between the + output of your signal generator and the probe tip on an oscilloscope and the - output to the reference input on the scope, you will not be able to measure any difference in the amplitude of the signal because the input resistance of the scope is probably more than 6 orders of magnitude higher than the resistor. If, on the other hand, you use a 450 ohm resistor and a 50 ohm resistor in series across the signal source and you connect the oscilloscope leads across the 50 ohm resistor, you will see that the amplitude of the signal on this resistor is 1/10 of the signal amplitude across the output of the signal source.

    Similarly, if you connect a capacitor in the first scenario between the signal source and the scope you will ac couple the signal to the scope but you will not really be able to measure meaningful characteristics of the capacitor. If you connect the capacitor with a resistor, you will be able to measure the time constant by comparing the phase of the signal across the resistor with the source signal. You can calculate capacitance with this information. You could also sweep the signal source frequency to get an idea of the frequency response of the RC filter circuit you have made with this setup.
  10. Nov 9, 2013 #9
    Ok so I think I followed everything you said ScienceGeyser, it makes much sense in regards to the diode functionality as the silicon doesn't really work as planned without current (only partially). It would seem the oscilloscope is high impedence so as to make it as "transparent" as possible to what you are testing, provided you essentially probe a circuit in parallel to it (rather than in series with the scope) so that it can do it's job transparently (interesting to try things anyways to understand them like the diode perhaps). Hopefully I'm on the right page with that.

    And your last paragraph about testing capacitors is essentially what I'm looking to do there: just to actually see some filtering of the input freqencies, etc. as they get used in amplifier circuits, etc. in addition to some basic dc offset filtering, etc.

    I guess one question I have is about so called output impedance of the source. I recently put together a simple L317t regulator circuit for use with various wallworts, etc I have in addition to a 30v power supply I have. How do I design my "output impedance" of that circuit in order to (I'm assuming) make it low or whatever it should be??? I'm guessing that if it is low impedance then that means it has essentially low resistance (as opposed to the way a battery might behave). I guess I'm not sure about all of that.

    Thanks again for the responses.
  11. Nov 9, 2013 #10
    You are correct about the input impedance on the scope. It needs to operate as a voltage measurement device and high impedance is required to make measurements without changing the characteristics of the circuit under test.

    Output impedance is essentially the ability of a circuit to drive another circuit without experiencing any signal loss or distortion (you'll usually see some distortion just before you bake the output). Your LM317 VR circuit has an impedance that is determined by the characteristics of the semiconductor. Typically, changing the resistor values that set the output voltage do not change the circuits ability to deliver current. Having said that, it is important to remember that this type of regulator has to dissipate a lot of power if you run the output many volts below the input. For instance, if you use a 30V supply and regulate it down to 12V, you will have 18V across the regulator. If you are drawing 1A from the regulator, you will be dissipating 18W with the regulator.

    The datasheet for this regulator suggests that you can run the device from 1.2V to 37V and 1.5A with a 41V input. If you read a little further in the datasheet and do a couple calculations with the power dissipation, you'll find that you can't really do that.

    Just for fun, lets say that you have a 40V supply and you regulate down to 2V. This is 38V across the regulator. If you are drawing 1.5A with your load, then the regulator is dissipating 57W. If you look at the thermal resistance of the device, you see that you'll have a 5 degree Celsius rise of the junction temperature for each watt dissipated by the device. Even if you could keep the case of the device at 20C, the junction temperature in this scenario is going to be 305C (Way too high). The operating junction temp for this device is rated at 125C, suggesting a maximum of 21W if the case is kept at 20C. More realistically, this device is good for about 15W which means no more than 10V between input and output at 1.5A.

    I hope all of this is helpful. I sometimes have a tendency to ramble.
  12. Nov 9, 2013 #11
    Thanks for the last response... i'm still working thru it but wanted to post some specs of my sound card:

    Two outputs cross-coupled electronically balanced or unbalanced

    Input Impedance:
    Balanced mode: 24 k ohm, Unbalanced mode: 12 k ohm

    Output Impedance
    Balanced mode: 100 ohm, Unbalanced mode: 50 ohm

    Output Drive Capability
    600 ohm

    +5V @ 400 mA, +12V @ 220 mA, -12V @ 95 mA

    So I've noticed with trying various resistive load sizes that I get some clipping (while under load) of the positive side of the signal once I reach a certain amplitude. I'm trying to understand from those specs what the proper load might be that I should try (assuming that's causing the clipping). Could just be the sound card is not working up to spec.

    I've only seen about 1-4 mA's under load. It will make sound thru a small 8ohm speaker. only when it's distorting (I've had it in series with various resistor sizes from 100 up to 15k ohms.

    I tried a diode (i.e. the diode test I was initially trying) and apparently there's not enough current to forward bias it. Hmmm... Perhaps using the computer soundcard as an "inexpensive" signal generator isn't quite going to work out? (or expensive depending on how you look at it... it's a $400 sound card... I really should hook up my cheap soundblaster one now that I think of it before I wreck this one .. if I haven't already!)
  13. Nov 9, 2013 #12
    RP also - keep in mind the Ground / Guard ' Neg Lead on the scope is probably grounded through the mains, if you go rooting around in a 120V connected amp --- pretty easy to do some damage. Pretty common to have the scope on a separate isolation transformer.
    Oh -- also when energized, around the caps in particular - for example when powering up an old amp that has not been used for a while - please be sure to wear safety glasses - the large filter caps do fail in an exciting manner.
  14. Nov 9, 2013 #13
    Good info on the LM317t. Not suprised reality doesn't match the words. I had read about the heat being a result of the difference between in and out voltage. So it's good to know that it will be the culprit if I start hitting those maxes ever. 10v drop isn't much especially seeing as how I wasn't planning to do a whole lot of a heat sink for it. Perhaps you could have a few in series and step it down gradually somehow. I just liked the idea of having something variable but later it may not matter and I'll just need a 5v for IC's and a couple of others. Likely the 7812/7912 7815/7915 etc aren't much better in the thermal performance area. Probably why those expensive bench power supplies cost what they do... probably unique xformer for every rail.
    Last edited: Nov 9, 2013
  15. Nov 9, 2013 #14
    Thanks for that. I have actually fixed a 100w amplifier (biamped lm3886 i believe) and was in there scoping the caps and stuff. I'll take note of the safety glasses thanks.

    I'm not sure what you mean by "guard" about the scope neg lead. Are you talking about ground potentials being different between the scope ground to mains and the amp ground to mains (same mains)? Could you expand on that a bit. Is it better perhaps to disconnect the scope ground or something? Is the damage you mention as a result of potentially grounded energy raising the voltage of what you're probing?
  16. Nov 9, 2013 #15
    @Windadct: Never mind my questions about isolation transformers. I'm reading up on that now as there seems to be plenty of info about it.

    Thread sidetrack:
    Btw, one thing I discovered about scopes and isolation transformers (in just reading about it after you mentioned it) is that the scope should NOT be on the isolation transformer, but rather the device you're doing the testing on should be on the iso transformer (according to tektronix http://www.tek.com/document/technic...lloscope-measurements-and-operator-protection). I suppose that would depend on if you're using a public iso xformer which has a secondary connection to ground, or a testing one which doesn't have a secondary connection to ground (not good), but tektronix discourages any iso transformer on the scope.
    Last edited: Nov 9, 2013
  17. Nov 10, 2013 #16
    I am not a fan of isolation transformers. In most modern UL listed electronic devices, the earth ground is not allowed to be part of a current carrying circuit and is simply there for the protection of the user. In the US, the neutral side of the line is bonded to ground at the service disconnect and, unless something is wrong with the site wiring, it is at or very near earth potential. The only thing an isolation transformer really does for you is allow your device to float away from ground to some arbitrary and unpredictable voltage that may become dangerous unless you reference it to something else... Like ground.

    I have worked with high voltage power supplies for a long time, and unless you need to have a piece of equipment floating, like the filament supply on an e-beam emission source, you should keep it grounded.

    What you really need, is to understand all of the test points that you choose in your circuit and what the potential of your reference point is going to be. If you need to reference a high voltage, e.g. that filament supply mentioned above, with a scope probe, then you should use an isolated probe. On a lot of newer scopes, you can connect the probe tips to the two points of interest and the references to ground and display the difference of the two signals as an additional trace on the scope. In all cases you need to make sure that your probes are rated for the voltages that you are measuring with respect to earth ground since that is most likely the potential that your body will be at.
    Last edited: Nov 10, 2013
  18. Nov 10, 2013 #17
    Right. I didn't see much point to it from my novice perspective. It seems to me that the scope uses the ground for that probe connector as a convienent way to have a typical reference (i.e. you don't need to connect the reference probe end at all usually) as well as for allowing inner circuit voltage drop testing (on low voltages), as well as a safety ground to the device chassis (three different purposes it would seem). It makes sense to use 2 probes and add/invert them differentially on the scope and avoid the use of "abusing" the safety grounds on the probes for that. Tektronix already says you shouldn't be using the scope to test floating voltages and yet the iso transformer sets up that exact scenario as some sort of benefit yet I don't see it.

    Any suggestions on my post about the soundcard specs ScienceGeyser? I'm wondering if I'm correct about the distortion I'm seeing as well as the diode not forward biasing due to the small current. Thanks again for all the tips ScienceGeyser and Windadct.
    Last edited: Nov 10, 2013
  19. Nov 10, 2013 #18


    User Avatar
    Gold Member

    Looking just at the soundcard, what voltage does it clip at open circuit (can be different positive and negative). They say there is a 50 ohm output impedence, and that it can drive 600 ohms. You said the output was 5VAC --- is that peak to peak?

    Measure clipping voltages at 200 Hz. open circuit, and then measure it with a 1K load. Then measure it with a 100 ohm load. Be sure you are connected in unbalanced mode from 1 output to ground.

    Should be no problem driving a diode in series with a 1K resistor. Measure across the 1K.

    Likewise try a 0.1uF in series with 1K, sweep, and look at 1591 Hz corner (+- component tolerances).

    I'm a little confused about what the issues are.
  20. Nov 10, 2013 #19
    No clipping when it's an open circuit, only under load. I'm measuring the ac voltage on a fluke so I'll assume it's RMS (up to about 5 volts when I get the volume up).

    I'm using the unbalanced mode (one output connected to ground). They've got cross coupling circuitry so I wonder if that's affecting things some how.

    I tried using 100ohm resistor, then a 15k resistor to try to see the difference but it distorted (just the positive portion of a 440hz wave, which I tried at full volumn and a lowered version, they both distort). I haven't measured those two voltages yet (hi and lo resistor values). I wasn't sure what it would gain me. Perhaps I need to be able to calculate the full circuit math via their output impedance and drivable impedance. I'm not sure how to do that yet.

    I'll try the 1k resistor in series with the diode you mention and measure across the resistor. I believe I tried one (IN4001) in series with the various resistors I tried but I don't recall using a 1k so I'll try that specifically.

    Basically I'm just trying to use my soundcard as a signal generator so I can do some experiments while learning some fundemental ac stuff and how basic components can affect things (put in a diode and verify the ac gets half recitified, put in various caps and verify see how the output is affected using different frequencies, etc). Then I was hoping to get enough basic understanding intuitively using this setup and soon after get a standalone oscillator of somesort built up and use that. I just liked this because I could easily vary the frequencies, etc. and see how caps filter the ac,etc. (rather than just reading about it in a book)

    "Likewise try a 0.1uF in series with 1K, sweep, and look at 1591 Hz corner (+- component tolerances)." - you lost me with this line. I realize the .1uF is a cap but unclear what the "1591hz corner" is all about?!

    Thanks for the assistance meBigGuy! I learn something everyday I get into this stuff (or at least try to)!
    Last edited: Nov 10, 2013
  21. Nov 11, 2013 #20


    User Avatar
    Gold Member

    The soundcard is a fine signal generator. You can connect it to an audio amplifier chip if you want more drive later.

    I'm not buying the 5V RMS --- What is the model number of the soundcard.

    I don't understand why a 15K resistor would distort or be different that open circuit for that matter. I assume you mean clip on positive peaks. (the meter readings are not important, lets just talk about peak positive and peak negative voltages from the scope).

    So what is the peak to peak voltage, unloaded, according to the scope and then with 15K load (output to ground). Difference should be very very small. If not, measure the resistor (that's a good use for the meter!)
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Discussions: Simple diode test doesn't jive. What gives?
  1. Simple Diode Circuit (Replies: 4)

  2. Simple Diode Circuits (Replies: 2)

  3. What does the diode do? (Replies: 10)

  4. Simple diodes question (Replies: 20)