Checking Calibration of Oscilloscope..

Discussion in 'General Electronics Chat' started by Dr.killjoy, May 16, 2016.

  1. Dr.killjoy

    Thread Starter Well-Known Member

    Apr 28, 2013
    1,190
    156
    I have a Tek 2215 that I currently slowly rebuilding and upgrading .. But I got the unit broken and after repair I would like to check the calibration of the scope .. I never have done anything like this before and hoping someone could help out.. My questions comes as can I check the calibration of the scope and make adjustments if needed with the tools I have ???


    Where is what tools I have
    Fluke 87 Series II
    Fluke 77 Series II
    Ascel AE20125
    Victor VC2002
    Power Supply
    Heathkit IG-4505

    Here is what I need
    [​IMG]

    Also here are a couple items I thought might help but not
    1000:1 divider
    http://hakanh.com/dl/docs/datasheets/067-0529-00.pdf
    Tunnel Diode Pulser
    http://w140.com/tek_067-0681-01.pdf
    Basic RF Attenuators - Design, Construction, Testing - PI and T style
     
  2. Lestraveled

    Well-Known Member

    May 19, 2014
    1,957
    1,215
    @Dr.killjoy
    The two things you need to verify/calibrate are the time and the amplitude accuracy. An easy way to cal the time base is to buy a few crystal oscillator modules. They are cheap and you can choose the best frequency based on what your scope requires for calibration.

    The amplitude can be done with a power supply and a voltmeter. Use your scope as a voltmeter to measure a voltage determined by a voltmeter you trust.

    My advise: Don't calibrate a function unless you are sure it needs it.
     
  3. MrChips

    Moderator

    Oct 2, 2009
    12,436
    3,360
    +1
    Fully agree.
     
  4. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    It depends on how close to the specs you want to get. The analog scopes I've calibrated only needed to be within 5% or so of nominal values. If you don't have calibrated time/voltage sources and meters, that could be problematic.

    The horizontal calibration procedure is for every range with the fastest being 50nS; your scope can't display a 20MHz square wave. For the faster ranges, you could cobble together a bunch of differentiators to get output equivalent to a time mark generator. From what I read in the 2901 manual, Tektronix set the time constant of the RC networks to be 5-10% of the square wave duty factor.
     
  5. shteii01

    AAC Fanatic!

    Feb 19, 2010
    3,387
    497
    What about calibration tab on the scope itself?
     
  6. RichardO

    Well-Known Member

    May 4, 2013
    1,230
    382
    How much do you want to build versus buy?

    I was able to verify the calibration of a 250 MHz scope (Tek 475) without buying any fancy test equipment.

    It _was_ a lot of work. I built my own 50 ohm terminators and attenuators using male and female connector pairs, tunnel diode and avalanche pulsers, and a time mark generator. I used DC and a 4 1/2 digit DVM to check the vertical amp gains. I learned a lot doing the test equipment so I don't regret the effort.

    The terminators worked very well but were fragile, the tunnel diode pulser had very disappointing performance, the avalanche pulser worked great but was a pain to use because of the fairly high voltage needed to make it work.

    The time mark generator (which abused a PIC microcontroller) worked okay while it was cool but crashed when it got warm. I should have read the calibration procedure before doing the PIC time mark generator. I was real proud of myself that it could generate 100 MHz pulses until I got to step 3 of the timebase calibration. I said to set the timebase to 10 ns per division. Good so far. That is 100 MHz for one pulse per division. But... Then the procedure said to turn on time 10 horizontal sweep expansion. That is 1000 MHz (1 GHz) at one pulse per division! :eek::(

    Fortunately, the scope was in calibration since these homemade tools would not have been up to the task of doing an accurate calibration.

    I can help with any of the calibration stuff you want to build yourself including doc on the pulsers and on the time mark generator. I have new and improved pulser circuits that can give sub-nanosecond risetimes.
     
    Dr.killjoy likes this.
  7. Dr.killjoy

    Thread Starter Well-Known Member

    Apr 28, 2013
    1,190
    156
    Wow and thank you very much .. I would love any information I could get on building the calibration equipment..
     
  8. RichardO

    Well-Known Member

    May 4, 2013
    1,230
    382
    Here is a first step for verifying the bandwith of the vertical amps. It is a pulser with a fast falltime.
    74F14_Pulse.png

    The I built this circuit using a DIP 74F14 on a solderless breadboard measures a risetime of about 1.75 ns. (2.5ns risetime on a scope/probe system with a 1.7ns risetime.) You could get better performance using surface mount parts on a PCB. I should have taken a picture of the circuit to show you how I did built it. I will do that later when I have a chance.

    This should be fast enough to test your 60 MHz scope. The "slow" falltime of the pulser will make the scope appear a bit slower than it really is. This is because the risetime of a system is the square root of the sum of the squares of the actual risetimes.

    p.s. If you feel the need for something faster, I have made a circuit board with a pulser that has a falltime of less than 1/2 ns. :D
     
    Dr.killjoy likes this.
  9. Dr.killjoy

    Thread Starter Well-Known Member

    Apr 28, 2013
    1,190
    156
    Sorry I haven't answered the thread but I just realized the Heathkit will pretty much cover the Calibration generator and Time-mark generator and I could use signal generator could be used as the Leveled Sine-wave generator I think but sure .. I just got a pretty good deal on a Fluke 1900A frequency counter .. I do have one concern here as to when I am done with checking calibration what can I use the frequency counter and Heathkit for and are they worth keeping ???

    @RichardO
    Thanks you very much and still interested
     
  10. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    The voltage and time outputs on that generator are by decades and it won't be convenient to calibrate a scope that uses a 1-2-5 sequence. Shortest time interval is 1uS, so you can't calibrate the fastest time base ranges.

    EDIT: Just noticed the 1-2-5 slide switch for time out making it more convenient, but still too slow for a 60MHz scope.
    I'd keep them. Even though the scope calibrator isn't very good, it's better than nothing. It's always handy to have a frequency counter. You never know when you'll need it again.
     
    Last edited: May 22, 2016
  11. Dr.killjoy

    Thread Starter Well-Known Member

    Apr 28, 2013
    1,190
    156
    I wasn't sure whether to start a new thread or just continue this one...
    So mods please modify as see fit..


    I am still working on getting more parts and such.. I got the Freq counter in mail the other day and only issues I could find was the main fuse was blown. When I replaced it the unit ran fine and I got a chance to check the scope calibrator and my function generators to see what was in spec or not.. But I have some questions or need advice on some other parts .. In the manual it say I need the parts listed below.. Also need this stuff for my frequency counter,scope,and function gen.. Where can I get cheap bnc patch cables and does it matter to buy cheap or quality ? I am looking to get some male & female bnc connetor from china for a couple projects and was wondering if any experience with cheap bnc vs quality bnc ??Does anyone sell the parts below cheap ??


    Need parts
    Bnc patch cables
    50ohm terminations
    Dual-input coupler
    T-connector
    Probe to bnc adapter
     
  12. shteii01

    AAC Fanatic!

    Feb 19, 2010
    3,387
    497
    While looking for o-scope probes I was not sure if the cheap Chinese probes on ebay were any good. The tests that people ran show that those probes are good for 100 MHz or less. Based on that I would say that BNC patch cables from ebay should be fine for signals that are around 100 MHz.
     
  13. dl324

    Distinguished Member

    Mar 30, 2015
    3,242
    619
    How cheap are you looking for? Just checked Jameco:
    3 foot cables are $5
    50 ohm terminators $1.65
    T connector $1.60
    Barrel connector F-F (2 input coupler?) $1.25

    I'd buy quality because you'll likely be using them for decades.

    A probe to BNC adapter will probably be spendy. They're convenient, but I just stick the probe tip in the connector. Taking care not to bend/break the probe tip.

    Be aware that there are 50 and 75 ohm BNC connectors; they look identical on the exterior.
     
  14. RichardO

    Well-Known Member

    May 4, 2013
    1,230
    382
    Scopes and other test equipment is normally 50 ohms. Video is usually 75 ohms. As dl324 said, the look the same so get them from a reliable source and if you have both keep them separate to prevent later confusion.
     
  15. RichardO

    Well-Known Member

    May 4, 2013
    1,230
    382
    I have built the circuit on a solderless breadboard. As shown, I was not happy. The falltime was about 2.5 to 3 ns and the waveform was not "clean".
    74F14_pulser.JPG
    I made 3 changes to the circuit. The first was to get rid of the 150 ohm series terminating resistors. This got the falltime down to about 1.5ns but the circuit must now be plugged directly onto the scope input BNC. The output will not properly drive a coax cable without the 150 ohm resistors.

    I added a 2.2 kohm resistor to the output to get a higher amplitude pulse. I reduced the frequency from 10 MHz to 3 MHz to allow the 2.2 K resistor longer time to charge the input of the scope.

    I am now happy with the results and I think it will perform quite well if you build it. The circuit should be built on a prototyping PCB with traces the same as on a solderless breadboard. The capacitor and jumpers that go over the 74F14 on the solderless breadboard should be installed on the copper side of the PCB.

    If you decide that the pulse is not fast enough then you can build one of these:
    http://forum.allaboutcircuits.com/threads/fast-risetime-falltime-pulsers.124518/#post-816114

    They may need some simple modification because, although the pulse is fast, the pulse width is quite narrow.
     
  16. RichardO

    Well-Known Member

    May 4, 2013
    1,230
    382
    Get a copy of an ARRL manual. It has _lots_ of great information. It does not have to be a new one... mine is from 1992. :)

    The ARRL manual has tables for "T" and "pi" networks to make high frequency attenuators and how to make your own step attenuator. I built this one:
    ARRL_attenuator.JPG
    I put some fast pulses into it today and it works quite well giving good clean rise and fall times in the few nano-second range.
    (The trick is using large slide switches and really good shielding between attenuator stages.)
     
Loading...