Am I expecting too much from an elderly function generator?

Discussion in 'Test & Measurement Forum' started by dansteely, Jan 7, 2018.

  1. dansteely

    Thread Starter Member

    Feb 26, 2017
    31
    1
    Hi All,

    I'm getting back into electronics so please excuse my simplistic question...

    I've recently go a new DSO and have been watching some threads on measuring capacitors, in particular ESR.

    That has lead me to use my 25+ year old Feedback FG601 function generator. For sine waves it's pretty good up to it's max frequency of 1.49MHz.

    Square waves on the other hand at that frequency are unusable. This is being an example. Feedback FG601 - 1.49MHz Direct Out.jpg

    For the capacitor ESR measurement. I need a 200KHz square wave and this is also slurred. (see below)

    Feedback FG601 198KHz Direct Out.jpg

    Am I seeing the limitations of my old equipment? If so is there a way around the issue or is the cure a new modern FG?

    Many thanks an anticipation.
     
  2. bushrat

    Member

    Nov 29, 2014
    209
    58
    Have you tried adjusting your o-scope probe?
     
    simozz likes this.
  3. bertus

    Administrator

    Apr 5, 2008
    19,507
    3,972
    Hello,

    As @bushrat said, is the probe compensation adjusted?
    There should be a signal available on the scope for adjusting.

    Bertus
     
  4. MrChips

    Moderator

    Oct 2, 2009
    18,477
    5,858
    What you are observing on the oscilloscope is known as bandwidth limitation and is not uncommon.
    You need to know and be aware of the bandwidth of the instrument and all signal leads and probes.

    As you can see, there appears to be a problem with fast rising and falling edges. If the equipment has a restricted bandwidth, or there is too much capacitance in your test leads, the signal has a difficult time rising and falling in that short period of time.

    So what do you do about it?

    The first thing is to check the bandwidth of the oscilloscope. Is it spec'd to at least 10MHz or higher? You need this amount of bandwidth to be able to observe a 1MHz square wave.

    Secondly, what type of probe are you using? Are you using a 1X or 10X probe spec'd for at least 10MHz or 100MHz?
    This is one of the reasons why we use 10X probes. If an oscilloscope is spec'd at 10MHz, that is what you get with a 1X probe. When you use a 10X probe and it is set up correctly you can extend the 10MHz scope to 100MHz.

    Finally, is the function generator spec'd to generate a clean 1MHz square wave?

    I suspect that your function generator is fine. The problem is with your scope and probe. Check out both first.
     
  5. dansteely

    Thread Starter Member

    Feb 26, 2017
    31
    1
    Sorry, I think my reply here crossed yours.

    Yes I calibrated the test leads on the 1KHz test point on the scope

    The scope is a Hantek DSO5102P 100Mhz. The probes are marked 100MHz/6MHZ.
    I have a spec sheet for the probes and can supply more data if required..
     
    Last edited: Jan 7, 2018
  6. KeepItSimpleStupid

    AAC Fanatic!

    Mar 4, 2014
    3,269
    584
    This https://www.circuitspecialists.com/blog/oscilloscope-probe-compensation-adjustment/ is still a bunch of gobbly gook too.

    Anyway, I'll make a short attempt.

    "Typical", not all, scope impedances are 1 M || 22 pf The scope probe you choose has to be able to compensate. The scope probe has a small capacitor on it. It could be at the termination or the probe,

    Switchable x1 x10 probes will cover the compensation adj at 1x.

    The none square form is because the scope is seeing some combination of capacitance and resistance. You want it to see only resistance.

    The probe wires have capacitance per foot.

    The compensation adjustment takes advantage of a usually 1 kHz square wave available at the scope.
    You attach the 10x probe and get a stable calibration signal and then adjust the compensation cap for the probe to get the best square edge.

    Now, your probe has a 10 M input resistance and is totally resistive. It also attenuates signals by 10, so 1v/div on your scope now is 0.1 V/div.

    I had a hard time following the posts above, so I added mine.

    Warning: a probe capable of 1 gHz will not likely work on a 10 Mhz scope. Watch for 50 ohm terminations. Get probes to match. ebay has lots of 100 mHz scope probes for low cost. The scope probe bandwidth should be at and can sometimes be higher providing it can be compensated on your scope.
     
  7. spinnaker

    AAC Fanatic!

    Oct 29, 2009
    7,815
    3,666

    1khz is a long, long, long way from 100mhz.
     
  8. KeepItSimpleStupid

    AAC Fanatic!

    Mar 4, 2014
    3,269
    584
  9. dansteely

    Thread Starter Member

    Feb 26, 2017
    31
    1
    Many thanks to all respondents on this.
    I think I get it.. That the DSO test leads electrical characteristics can alter the signal under test and this is more marked at higher frequencies??
    Notwithstanding that..

    Cutting back to the original question about measuring capacitors are in spec (electrolytic's in particular). Is using a scope and a function generator the right way to go? I'm not looking for 100% parameter accuracy here, just real-world pass/fail. Or am I being too simplistic in my ambitions?

    Again, many thanks to you all.
     
  10. philba

    Active Member

    Aug 17, 2017
    959
    302
    If you are just looking for cap value, then use rise or fall time to determine the RC constant. If you are looking to determine ESR, take a look at this video.
     
    atferrari and JohnInTX like this.
  11. Terry01

    New Member

    Dec 28, 2017
    1
    0
    x
     
  12. BobaMosfet

    Distinguished Member

    Jul 1, 2009
    593
    130
    Regarding the probes-- this is just something to be aware of. Your scope doesn't see the signal. Only the probe does that. How it interprets that signal, right or wrong, is what the scope gets. Tektronix has an amazing write-up on scope probes-- kinda deep technically, but if you can find it, it's worth reading. There is a reason some scope probes cost more than the scope itself.
     
  13. AnalogKid

    AAC Fanatic!

    Aug 1, 2013
    7,776
    2,186
    Compare the square wave function output to the TTL output.

    ak
     
Loading...