My first scope came in, and already I have questions... :)

Discussion in 'General Electronics Chat' started by Austin Clark, Aug 15, 2012.

  1. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44
    I recently got this puppy in the mail: http://www.saelig.com/PSBE100/PSBE100008.htm:cool:
    So far, it's been amazing. Clear screen, sturdy construction, etc;. I had a little hiccup in calibrating it correctly at first (the first time I did it I must have made a mistake, because I got quite a bit of noise. Thank god I decided to try re-calibrating it, else I would have never known the difference) but other than that my only issues are in understanding the proper use and considerations that go with using the scope effectively. For example, probe resistance and capacitance. The capacitance I've known about for awhile, but I never really considered some of it's consequences until I've begun experimenting with it hands-on. For example, if you probe the 1kHz test signal from the scope, you get a nice clean square wave, however, if you put a 1M ohm resistor in series with the signal, the rise and fall time takes a serious hit. I imagine this is because less current is getting to the internal capacitance to charge it to the actual voltage coming in. In general, I theorize that the higher the frequency you wish to measure, the more current the source signal must be able to deliver in order to get a realistic view of the wave. I thought that the only thing to consider was the bandwidth, but it isn't, if your signal source isn't capable of much current, you could have problems displaying even a 1kHz square wave. How is the bandwidth really measured then? What voltage/current is used to spec it?Also, how would I go about calculating the "strength" of this effect.

    I've also had trouble measuring current with my probe. I put a low-value resistor (1Ohm, 1%) in series with the circuit I wanted to probe and probed across that resistor to calculate the current from there, however, the voltage change isn't much to measure, it's not too accurate, and you still had to sacrifice current thanks to that 1Ohm resistor (which is a lot when you want to probe the current through a motor on startup and stuff, even a low resistance will keep the proper current flow from occuring, because it's almost a dead short otherwise.)
    HOWEVER, I haven't really given this a test since I re-calibrated, possibly things have improved since. Still, am I doing this correctly? Is it even possible to measure a fairly precise current draw way?

    I have a few other questions, but they're mostly related to those issues. Any suggestions? Tips? Answers? Etc;? I'm not COMPLETELY lost, but I'd like some reassurance if nothing else :p This is my first probe and no one I know has first hand experience. Thanks :)
     
  2. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,388
    1,605
    The probe you use is also part of the measurement circuit. If you buy a gigahertz scope and use cheapie 30 megahertz probes... you got a 30 megahertz measurement. So get probes of equal (or greater) frequency.

    There is usually a knob somewhere on good probes to null out some of the frequency attenuation. On the scope should be a CAL signal output: chip the probe on to that and adjust for the best square wave.

    Measuring current with a sense resistor as you are doing is a standard and effective technique. For extra credit, look up (high side) current monitor chips that can amplify the signal thru a small resistor
     
    Austin Clark likes this.
  3. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44
    How would I go about testing my probes? Check rise/fall times? How do I know if the rise/fall times are due to input resistance and not bandwidth? and under what voltage/current should I test at? I've looked on the site and can't seem to find if the specs are listed. I'll have to check the packaging it all came in and see if I can find a manual or something later. I assume (hope) they'd give me probes that would be the best for my scope though...

    My probes do have some sort of adjustment on them, but I didn't mess with them at the time because it seemed to be ok testing the 1kHz square wave without that extra resistor, I'll have to give that a shot as well when I get the chance. What exactly am I adjusting, and how does it work/help? Also, how do I know what the rise/fall time of the test signal should be? Probably the manual once again :p

    Sense resistors of 1Ohm are sufficient then? If so, it must have been too noisy for me to see at the time and/or maybe I was just being stupid and had that channel AC coupled (Which is totally possible, beings that I tried that out within an hour of openning it, and had no idea what I was doing :) )
     
  4. MrChips

    Moderator

    Oct 2, 2009
    12,440
    3,361
    The input impedance of the scope + probe is just as important as when using a DVM.

    Most DVM have input impedance of 1MΩ or maybe 10MΩ. If you are going to measure a circuit that has impedances approaching 10% of that then the DVM is going to give an erroneous reading.

    Same goes for scopes.
    Most scopes have input impedance 1MΩ and 20pF in parallel. Hence you are looking for trouble if you are going to put an additional 1MΩ in series with the input. You are essentially creating a 2x attentuator as well as a high frequency rejection circuit.

    Get a 10x probe and use it. This will give you 10MΩ input impedance.
    There is a little trim cap on the 10x probe for frequency compensation. Learn to adjust this properly. Connect the probe to the test output and adjust the trim cap for the best square wave signal showing on the screen.

    For measuring current, be careful with what you are doing. Proper current measurement requires connecting both the probe and the ground clip across the sense resistor.

    YOU CAN ONLY DO THIS IF THE SCOPE OR THE TEST CIRCUIT IS FLOATING.

    Most, if not all scopes are GROUNDED.

    CONNECTING THE GROUND CLIP OF YOUR PROBE TO A CURRENT SENSE RESISTOR COULD DAMAGE YOUR SCOPE OR CIRCUIT UNDER TEST.

    As Ernie hinted, this is called "high-side" current monitoring.

    If you know how to measure "high-side" current, do a quick Ohm's Law calculation and determine what voltage to expect. This could be too low (under 50mV) to measure properly with the scope.

    Another way to measure "high-side" current is to use two probes, one for each channel on the two ends of the current sense resistor and use the Math function to take the difference of the two channels.
     
    Austin Clark likes this.
  5. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    That's the reason Tektronix put a calibration square wave output on the front panel of most of their scopes.

    Good probes have a "compensation trimmer" on them you must calibrate to the scope's input stage to get the proper response on a square wave. You need a sharp square wave signal to use to set this trimmer.
     
  6. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    It's a trimmer capacitor used for "peaking" that affects the front edge of the square wave to square it up at the top.
     
  7. WBahn

    Moderator

    Mar 31, 2012
    17,743
    4,795
    Get in the habit of always checking the probe compensation when you connect a probe (including moving it from one channel to another). You can really mess up your results, even for relatively low frequency signals, if you don't.

    Also, don't just grab any probe and expect it to work on any scope. The range of adjustment on the probe is usually quite small (so that you have good resolution) and therefore it will only work with a pretty narrow range of channel input front ends.

    When you put a series resistor between the point you are measuring and the scope, you have created a new probe and you need to compensate that probe.

    Essentially, what you are doing when you compensate a probe is the following:

    Imagine that there was no capacitance involved. You then place a 9MΩ resistor (R) between the signal you are wanting to measure and the scope input, which has a 1MΩ input resistance (Rin). You now have a 10:1 voltage divider at all frequencies.

    But now what if you do have some input capacitance (Cin)? Say there is 10pF of capacitance. At high frequency, that capacitor starts shorting out the input resistor and your gain falls off. But what if you put a capacitor (C) across the 9MΩ resistor so that it starts shorting out the probe resistor at the same rate? At low frequencies the gain is Rin/(R+Rin) whereas at high frequency the gain looks increasingly like C/(C+Cin). You thus want

    <br />
\frac{R}{R_{in}}\;=\;\frac{C_{in}}{C}<br />
\ <br />
RC\;=\;R_{in}C_{in}<br />

    As an exercise for the interested reader, show that this not only results in the DC gain being the same as the asymptotic high frequency gain, but that the gain at ALL frequencies is constant (i.e., is frequency independent).
     
    Austin Clark likes this.
  8. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44
    Thanks for the replies thus far, however, I'm still confused about a few things.

    For one, how much noise should I expect when I connect my probes ground clip to it's positive "hook"? I feel as though I'm still getting too much noise here. Probably 50 mv p-p at 20mv per division.

    Also, if input impedence changes my effective bandwidth, how are scopes spec'ed? What voltage is used and is it a pure voltage source (Effectively no resistance)?

    Both of my probes don't produce the same output waves when I connect them to a crystal oscillator (about 25 MHz), one gives me a sine wave and the other gives me a sine wave with what looks like there's two waves combining (it gives me two humps per cycle), even when I test them seperately and I've adjusted them the exact same way with the trimmer.

    Lastly, I have two crystal oscillators (one at 1MHz and the other at around 2.5MHz) They both output a fairly nice square wave. However, I have other crystal oscillators at 20+ MHz and none of them give me square waves, only what looks like sine waves. Do crystal oscillators come in both types, or is my scope not able to see the output correctly, because the higher frequency oscillators also give me a heavily attentuated signal, it's several times weaker than the lower frequency ones. Are my probes of a lower bandwidth, or do the oscillators have a hard time charging the internal capacitance of my probes? What type of wave characteristics should I expect from square and/or sine wave crystal oscillators? What sort of rise-times, ringing, etc; if any?

    I do my experiments with a 10X probe setting.

    Thanks tremendously for the help thus far and for the help coming. :D
    Soooo much more I can learn with my scope, I'm super excited. Aaaand, I have a frequency generator coming my way within a month, so be on the lookout for a thread on that, as I'm sure I'll have more questions on that as well.:cool:
     
  9. MrChips

    Moderator

    Oct 2, 2009
    12,440
    3,361
    I have a Tektronix TDS 220 DSO right in front of me so we can compare.
    It is 100MHz, 1GS/s, same as yours.
    I am using a Tektronix P6112 probe which is 100MHz 10x probe.

    With the input shorted to the ground clip, about 4 inches long,
    the noise is about 15mV pk-pk on 20mV/div scale.

    Every square wave will contain higher frequencies other than the fundamental frequency.
    A 20MHz square wave would consist of odd harmonics, i.e. sine waves at 20, 60, 100, 140, 180... MHz

    Since the scope is going to attenuate signals above 100MHz you will essentially be left with the 20MHz and 60MHz signals. Hence don't be surprised to see mainly a 20MHz sine wave.
     
    Austin Clark likes this.
  10. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44
    Wow... I was under the impression that defining square waves as a sum of a series of odd harmonics was just a clever way to look at the mathematics. I figured that with the scope you'd basically just see the square wave with a sloped rise and fall thanks to the limited bandwidth, not a sine wave... Veeery interesting and totally unexpected. So, your probe behaves the same way?
     
  11. MrChips

    Moderator

    Oct 2, 2009
    12,440
    3,361
    I don't have a 20MHz oscillator right now but I can try later. I'm pretty certain you will see close to a sine wave. I can do a Matlab simulation for you.
     
  12. MrChips

    Moderator

    Oct 2, 2009
    12,440
    3,361
    Here is what a 20MHz square wave would look like with only the 20 and 60MHz signals:

    [​IMG]


    And with the 20, 60, 100MHz signals unattenuated and omitting everything higher:

    [​IMG]


    The x-axis is a nominal value in cycles. At 20MHz, 1 cycle is 50ns. Two cycle display shown is 100ns.
     
  13. The Electrician

    AAC Fanatic!

    Oct 9, 2007
    2,281
    326
    What are the rise and fall times of the output from these two oscillators?

    Have a look at their outputs with the probe set to both 10x and 1x; what are the differences?
     
  14. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44
    I'm actually already pretty familiar with this, I've used a program called HarmonicExplorer and messed with this concept a bit.
    However, the thing is, although technically a square wave is a sum of odd harmonics, I don't think your scope can actually filter out those higher frequencies and leave you with, say, a wave like your third example. Why would there be a few waves at the top there? Wouldn't your scope hold steady once it reached the true "High" voltage? I see no reason why it would swing back and forth like that. I think that, the lower your bandwidth, the slower your rise and fall times will be compared to the true rise and fall times and not much else. I also have a 10 MHz crystal oscillator that I've tested and it too gives me a pure 10MHz sine wave... If your theory were correct, should be able to see several of the odd harmonics and get a decent looking square wave, but I don't. (10, 30, 50,70,90MHz harmonics I should be able to see). Is it even possible for crystal oscillators to give you sine waves? (at least, I assume these are crystal oscillators I'm using).

    Also, one of my probes seems to lead the other slightly and have a bit higher amplitude when I compare them when connected to the 20MHz oscillator. I suspect maybe it has a better bandwidth? I have no idea why they would be different much at all. At lower frequencies they perform precisely the same.

    Lastly, I did another test on noise after tinkering with the scope a bit, I found that one of my channels experiences more noise than the other when shorted with ground clips. One gets very little noise, at about 5mV pk-pk, whereas the other channel gets more like 50mV pk-pk, with the same probe...


    Overall, I feel like something's not quite right with my scope. I'll play with it a bit more, but it just seems to act a bit weird.

    I've post some screenshots and/or videos soon, hopefully it'll show you one way or the other.

    I sent the technicians at Saelig an e-mail asking about the probes and some of the issues I've had, and linked them to this thread.



    EDIT: ALSO, the rise time I experience with the 1KHz test signal is 1000nS, and with the 1MHz oscillator I get about 10nS. I assume this is because the test signal truly does have poor rise time compared to the crystal oscillator.
     
    Last edited: Aug 17, 2012
  15. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44

    APROX RISE TIMES:
    -- 1KHz test signal --
    1X - 4uS
    10X - 1000nS

    -- 1MHz --
    1X - 40nS (no ringing or overshoot)
    10X - 10nS (I get a bit of overshoot and what looks like ringing)

    -- 2.4576 MHz --
    1X - 40nS (no ringing or overshoot)
    10X - 10nS (Ringing, overshoot)


    HOWEVER, when I try all of this WITH THE SAME PROBE on my channel 1, I get tons of weird ****. The waves aren't square, and they bob up and down thanks to an overlying low frequency. I get something that looks more like what Mr. Chips was talking about. It really is pitiful. My other probe will do better on that same channel, but it's still pretty bad. My other probe on my better channel is also pretty bad. Basically, I only have one good probe/channel combination it seems. Bizarre.

    EDIT: OK, awesome. I've got consistent results with every probe/channel combo now. I just re-self-calibrated the machine again. I think you may need to recalibrate every time you turn off the power (there's an off/on button, but also a hard on/off switch. A small LED remains lit when I only use the soft off/on button, but turns of when switch the hard switch).
     
    Last edited: Aug 17, 2012
  16. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44
    Sorry for the triple post.

    I've made enormous progress in the last 30 minutes of tinkering <.<

    I did a factory defaults reset of all of my settings, and of course did the calibration again and all of my problems have been resolved!!!! Well, except for one... I'm still getting sine waves where I expect to see square waves.
    In 10X mode,I see more like what MrChips was talking about, and most of the wave is positive (like I want/expect from a square wave oscillator), in 1X mode I see a triangle wave. This leads me to believe that MrChips was on to something, I'm basically seeing a square wave that wants to transition before it can fully rise and fall I think.

    Overall, I'm getting my hopes waaaay up now. Nothing's wrong, just misunderstood :) I hope.


    EDIT: I calculate that I'm getting a bandwidth of 35MHz. I think my probes are dummies afterall. Not 100% sure though. I'll wait to see what saelig has to say. Maybe the crystal oscillators really do have 10nS rise time, and it's not my scopes/probes bandwidth.
     
    Last edited: Aug 17, 2012
  17. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Do you have good decoupling caps on the power pins of your XTAL oscillators? Even then you can't expect them to produce sharp square waves, there is little requirement for that on a 20MHz XTAL osc output pin.

    Maybe the XTAL osc datasheet will have rise time specs for it's output?
     
  18. The Electrician

    AAC Fanatic!

    Oct 9, 2007
    2,281
    326
    The reason I asked you to measure rise times in 1X mode is because the bandwidth is greatly reduced in that mode, so rise times should be long compared to 10X mode. You should be aware that if you want the full bandwidth of the scope, you can't use the probe in 1X mode.

    Typically, the bandwidth and rise times for scopes are better than the spec.

    On eBay you can find low cost scope probes that are considerably better than their advertised specs. For example, these are probably more like 200 MHz bandwidth rather than 100 MHz:

    http://www.ebay.com/itm/Two-x1-x10-...020?pt=LH_DefaultDomain_0&hash=item334f669144

    This probe is specified to have a bandwidth of 4 MHz in 1X mode. The bandwidth in 10X mode is much more than 100 MHz.

    See this thread for more info about these probes:

    http://www.eevblog.com/forum/produc...t-100mhz-oscilloscope-probes-hands-on-review/

    I've attached two images. The first shows the output from a Rigol function generator set to generate a 20 MHz square wave as displayed on a Tektronix TDS210, which is a 60 MHz scope. The Rigol spec says the rise time should be 5 nS.

    The second image shows the same signal as displayed on an Agilent DSO5034A 300 MHz scope. The measured rise time is 4 nS.

    These two images were captured using the low cost probe referenced above which I bought on eBay.

    One way you can generate a square wave with a fast rise time is to get a 74AC74 flip-flop (the AC series is Fairchild's "advanced CMOS" family). Connect the flip-flop to divide by two and feed it with the output of one of your clock generators. The rise time of this part is just a little more than 1 nS when properly wired and bypassed.

    Your scope should be able to display a 20 MHz square wave at least as good looking as the TDS210 image.
     
    Last edited: Aug 17, 2012
    Austin Clark likes this.
  19. WBahn

    Moderator

    Mar 31, 2012
    17,743
    4,795
    Keep in mind that if the mathematics didn't do a good job of modelling the real world, they wouldn't be used.
     
  20. Austin Clark

    Thread Starter Member

    Dec 28, 2011
    409
    44
    I got most of my oscillators in a grab bad, so I'm not sure where I'd find their datasheets. What's a typical rise time for these things?

    I tried putting a few caps of all different capacitances and types and it only helped marginally with the overshoot and ringing.
     
Loading...