Which LCR Meter?

Discussion in 'General Electronics Chat' started by tobyw, Mar 27, 2013.

Thread Status:
Not open for further replies.
  1. tobyw

    Thread Starter Member

    Mar 21, 2013
    37
    4
    Could anyone advise me on buying an LCR meter? My budget is about £150-£200 and it will be for hobby use. My main reason for buying right now is to help me experiment coil capacitor values for radio circuits. But I often find that once I have a tool, I realise that I wish I had got one which did xxx or yyy etc.

    Choice so far is narrowed down to these...


    Extech LCR200 - £180
    Di-Log DL9401 - £165
    Agilent U1731C - £140
    Agilent U1701B - £100
    Atlas LCR40 - £90


    At the moment I am veering towards the Agilent U1701B or the Di-Log - mainly because they seem to go down to the lowest capacitance value (0.1pf compared to 20pf for the Extech). Not sure if that is the right criterion to compare on though...

    Does anyone have a view on which product would be best?

    Also - is it worth my paying the extra £60 or so to get it calibrated..? Or would I be fine for hobby use without it.
     
  2. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    You aren't paying to get it calibrated, you're paying for the "certificate" of calibration. I consider it unnecessary for my work, and I've chased down problems below the nanoamp level.
     
  3. The Electrician

    AAC Fanatic!

    Oct 9, 2007
    2,281
    326
    The Agilent U1701B measures capacitance only, no inductance.

    The 20 pF figure for the Extech is the lowest range, not the lowest capacitance it can measure.
     
  4. Metalmann

    Active Member

    Dec 8, 2012
    700
    223
  5. nigelwright7557

    Senior Member

    May 10, 2008
    487
    71
    You are looking in the right direction price wise.

    I bought a cheapie for £25 and it was useless.
    It told me an SMPS inductor was 100uH when it was 10uH !!!!
    Even a cheap USB LC meter told me it was 60uH.
     
    Metalmann likes this.
  6. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Or build a copy of my high-resolution cap meter here;
    [​IMG]
    http://www.romanblack.com/onesec/CapMeter.htm
    is measures down to hundredths of pF and for resolution beats all the other kits on the market. There's only a handful of parts in it.

    Sorry it does not do inductors. They can be difficult to measure as many need a fair amount of current for the core to work properly. You should really measure inductors with a test setup using similar current to your real application.
     
  7. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
  8. tobyw

    Thread Starter Member

    Mar 21, 2013
    37
    4
    That's a very cool project! However, I suspect the bulk of the cleverness is in the code.. probably a bit out of my league for now.

    Shame - it is really the inductance that I want to be able to measure. My current multimeter will do capacitance and resistance to some extent.

    Most of the meters I listed seem to imply that you can just connect the leads to a coil and it will tell you their inductance...is that not how it works? I was hoping I could experiment with windings and measure their inductance to arrive at specific resonant circuits... ie. practical way to understand the equations relating to L and C
     
  9. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    One would hope that an inductance meter would just spit out the answer, but they are limited. A real inductor has resistance, capacitance, and a self resonant frequency that a meter generally will not explain to you. The meter is only a starting point for the rest of your exploration.

    It's like a volt meter saying a battery has 1.5 volts. It also has internal resistance and a limited speed of current production that the meter says nothing about.
     
  10. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    It looks cool, but I'm a bit lost as to how that works. Why is the triangle wave needed? And how is that synced to the square wave?

    Part of the benefit of using a microcontroller is it has xtal perfect timing ability. I think you could measure the inductance quite well using a square wave of a fixed drive voltage, and measuring time period of the squarewave from current X to current Y?

    If you'd like to discuss designing a digital inductance meter maybe we should start another thread? BillMarsden and I had talked about it earlier but it fizzled.

    The HEX code is supplied on my web page, you just program the PIC or you can buy a kit of all the parts complete with the programmed PIC. (Note; I don't make any money from the kit sales, someone else is selling them).

    As far as inductance feature in a multimeter, I have one in a decent multimeter but it's not great for that task. Many inductors measure very differently to their marked value.
     
  11. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    @ THE RB

    The basic principle is that the current through the DUT is stable when the current is suddenly shut off. The energy in the inductor is sent through a fast diode to a capacitor where the operator does: 1/2 CV^2 = 1/2 LI^2

    The wave forms are sync'd because they both come from the same signal generator.
    The square wave allows current, then doesn't allow current. The triangle wave was just a convenient way to dump the charge on the capacitor between measurements.

    The digital interpretation and display are beyond me right now.
     
  12. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Thanks for explaining! So with 1/2 CV^2 = 1/2 LI^2 we could keep some factors fixed (like the current trigger point of an oscillator) and get a direct correlation of voltage to inductance?

    Ideally of course it would be best to eliminate any squares or roots from the final calc as that is what kill resolution and accuracy in the digital part.
     
  13. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    There is always Xl=2PiFL
    therefore, Xl/2PiF=L
    no squares, mostly constants.
    I'll get to that later.

    With my jig, a slow oscillator speed sets the current as, "stable". When you stop the current through the supply transistor, the energy has to go into the capacitor. Simple basis. Easy to see with a 'scope.

    Increase the speed, and you can measure the current at the moment of stoppage, or set a trigger point at a certain current, measure the resulting voltage on the catch capacitor, and get to the same place. Not quite so obvious to the eyes of a human, but doable. Trying to go, "automatic" with this requires you to know the switching speeds so you know how much time and change in current happens from the trigger point to the shut-off of the first transistor. This makes me want to do things slowly, and that shouldn't be a problem considering that a human can't find any added value in getting a reading once every microsecond instead of once every millisecond.

    There is also a temperature stable constant current device (a jfet) that can be switched off, as long as the switching transistor can survive the resulting voltage spike. That is fairly simple. Just design the switching section to survive more than full scale measurement voltage and declare, "over range" when you need to change the catch capacitor size. Somewhere in the over range voltage, a fast spike suppressor saves the switching transistor from the excess voltage. The constant current approach again eliminates switching time errors. Problem is, jfets are impossible to match for a production line quantity. You would have to measure exactly how much the constant current is. MPU to the rescue. And while you're at it, you can calibrate as to exactly what capacitor size is installed by doing V=KI/C. That seems to cover the fact that precision capacitors are expensive, and still not very precise.

    If you want to go the Xl=2PiFL method, you need a stable frequency (not a problem) and the goal is to find Xl. A resistor as a voltage divider, right? This is where a microprocessor becomes the hero by doing a 1/X function quickly.

    Vout = Vin * R/(R+Xl) or Vout = Vin * Xl/(R+Xl)
    Find Xl by doing math on the resulting voltage.

    I can't imagine that I'm telling you anything you don't already know, or couldn't work out for yourself in 10 minutes. Am I right? Or am I so frigging analog that this kind of stuff is a lost art to the youngsters?
     
  14. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Nah, I'm far from a youngster having done electronics since about 1975. ;)

    Part of the issue is people want to measure SMPS inductors etc which often need a certain level of current to make the core work, so it's ideal to test their inductance at a current near their final operating current, or at least high enough test current to get the core working.

    My preferred method would be to test the dI:dT ratio, so for instance to apply a fixed voltage, and measure the time from 50mA to 100mA, assuming 100mA was selected as the test current. So the analogue hardware would apply a fixed 10v (when oscillator ON), and the oscillator would have thresholds at 50mA and 100mA and oscillate between. Then the digital circuit would measure the total "up ramp" time from thousands of accumulated cycles which becomes very accurate when using a xtal for timing.

    So the device would have a rotary switch, and you set the current test range ie 50-100mA, and the display gives a very accurate digital readout of the part's actual inductance when running in that current range. It could also display the frequency, as that could be helpful info too if building SMPS devices. :)
     
  15. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    I'm glad you have a method for that. I let the idea of core hysteresis loose in my head 24 hours ago and an answer has not come out.
     
  16. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Nah it's just an idea of a method so far, not anything usable. Making a high speed oscillator that can apply a pulsed regulated 10v to the inductor raises it's own issues, as does the accurate current trip points which will likely need different current sense resistors switched in, and which need to be outside the voltage regulation loop.

    It's probably easier to put the inductor in a buck circuit at the right current, and measure average inductor voltage, and dI:dT ramp on the 'scope. ;)
     
  17. int1

    New Member

    Apr 2, 2013
    1
    0
    I've bought a Tonghui TH2822A from ebay about US$230.
    Its a real 4 terminal meter, and has very good low Z range. I remember measuring some 0.1% resistors
    , and it was within this error. Not bad for the price.
    There is a 100kHz version for nearly double the price.


     
  18. hutale

    New Member

    Jun 29, 2009
    1
    0
    There is an LCR 2.9 bridge, meter kit with video found at and doc at http://lcr.is-best.net/en/ . This is a LCR/RLC/in-circuit C/ESR meter measuring L/C/R/Z/Rs/ESR/X/Q/D/Phase angle with 100Hz/1kHz/10kHz sine test signals, base on auto-balancing bridge technique. Maybe you are interested in it too.
     
  19. alfacliff

    Well-Known Member

    Dec 13, 2013
    2,449
    428
    I use a sencore LC102, it checks all caps and inductors, and does esr and dielectric absorption tests too.
     
  20. #12

    Expert

    Nov 30, 2010
    16,284
    6,797
    Necropost.
     
Loading...
Thread Status:
Not open for further replies.