Page 1 of 2 12 LastLast
Results 1 to 15 of 18
  1. #1
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default Electronic Differential Levels - Theoretical Understanding

    I have been looking into electronic differential levels and trying to understand how they work - specifically the Mahr Federals;

    MahrFederal.jpg


    I believe they are based on a Linear Variable Differential Transformer of which I have an overall understanding of. However, the particular type of LVDT used by MahrFederal utilises a "shading loop" as the moveable item that varies the output voltage/reading. I need some help to understand how this is working.

    This is the system drawing that Mahr provides and shows the workings quite well. All components are connected other than the pendulum that holds the shading loop. My assumption is that as the shading loop moves left or right, it 'bleeds off' some of the magnetic flux to the secondary coil on that side and thus alters the output voltage/reading?

    Federal.jpg

    Am I correct in that? This is a little different to how a true LVDT works.

    The inside of the actual unit looks like this and is slightly different from the drawing;

    photo2.jpg


    The things that puzzle me further are;
    • the primary coil appears to be around the middle leg of the core but that leg does not connect to the other side of the core so how is the magnetic flux flowing?
    • the shading loop appears to be around a brass (?) bar so how it it doing anything seeing as brass is not magnetic?
    • I cannot actually see the secondary coils in the picture. I would have expected them to be visible around the 2 sides legs of the core.
    • The cross sectional area of the lower horizontal part of the core is smaller. This would reduce the magnetic flux flowing through that area but I cannot work out how it fits into the puzzle


    I did consider just building a simple bench model of the core and playing around but even a simple model is a lot more work than just asking the brainz trust on here!!!!

  2. #2
    Join Date
    Mar 2011
    Location
    Southern Flinders Ranges
    Posts
    1,536

    Default

    The shading coil is altering the phase angle. Shifting the phase angle (the position on the waveform) varies the output voltage.

    I have to be up at 4am so you only get the short answer this evening, I’ll have a look at your other questions tomorrow.

  3. #3
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default

    I look forward to your fuller reply! Looking into phase shifting whilst I wait. Although I do not understand WHY the phase shifts I believe I can understand what happens when it does;

    Phase Shift.jpg


    If at a point in time, the primary coil is at -20 volts in its AC wave then if the shading loop is centred, both of the secondary coils would read -20v (or some other consistent figure allowing for the various losses). In the above diagram, the shading coil is towards secondary coil 2 causing a phase shift. Secondary coil 1 will be induced to -20v but as secondary coil 2 is shifted along its wave form it will be induced at something less, say -12v. The net of these 2 readings would therefore be -8v which can be interpreted as a certain level change.

    Am I on the right track as far as the effect of the phase shift?

  4. #4
    Join Date
    Mar 2011
    Location
    Southern Flinders Ranges
    Posts
    1,536

    Default

    Quote Originally Posted by Mk1_Oz View Post

    The things that puzzle me further are;
    • the primary coil appears to be around the middle leg of the core but that leg does not connect to the other side of the core so how is the magnetic flux flowing?
    • the shading loop appears to be around a brass (?) bar so how it it doing anything seeing as brass is not magnetic?
    • I cannot actually see the secondary coils in the picture. I would have expected them to be visible around the 2 sides legs of the core.
    • The cross sectional area of the lower horizontal part of the core is smaller. This would reduce the magnetic flux flowing through that area but I cannot work out how it fits into the puzzle

    Consider that transformers are an inductive device in that the primary induces a voltage into the secondary. You can lay two straight pieces of wire next to each other and induce a voltage and call it a transformer, or if you live in the world of high speed data cabling, it’s called NEXT (next pair cross talk). It’s not a very efficient transformer though, takes a lot of voltage and a very small gap between them. If you form the straight wire into coils, you intensify the magnetic fields and create a slightly more efficient transformer. If you then add a very dense ferromagnetic material (iron usually) into the coils you intensify the magnetic field of coils and up the efficiency of your transformer.
    Magnetic flux doesn’t flow (despite what your high school physics teacher told you), you can only intensify and/or steer(shape) it. The core doesn’t need to be continuous to be effective. In this instance, the core has most likely been designed to create a specific shaped pattern based on application. I’m speculating here, but it’s likely given the nature of the application that the magnetic flux is directed in a sort of bell shape to allow more resolution in measurement.
    Brass is a diamagnetic, meaning it’s not magnetic per se, but will react with a magnetic field if placed in one (place a swinging brass pendulum in a magnetic field and it will slow). I’m surmising that the brass has been used exactly because of this property. This is why I’m suggesting the field has been shaped into a bell pattern as it’s likely dampening the pendulum movement to assist in resolution.
    If your instruments are functioning correctly then I would suggest that the two secondary coils are present and accounted for and just a bit shy and don’t like showing themselves to the public maybe.

    Quote Originally Posted by Mk1_Oz View Post
    I look forward to your fuller reply! Looking into phase shifting whilst I wait. Although I do not understand WHY the phase shifts I believe I can understand what happens when it does;

    Phase Shift.jpg


    If at a point in time, the primary coil is at -20 volts in its AC wave then if the shading loop is centred, both of the secondary coils would read -20v (or some other consistent figure allowing for the various losses). In the above diagram, the shading coil is towards secondary coil 2 causing a phase shift. Secondary coil 1 will be induced to -20v but as secondary coil 2 is shifted along its wave form it will be induced at something less, say -12v. The net of these 2 readings would therefore be -8v which can be interpreted as a certain level change.

    Am I on the right track as far as the effect of the phase shift?
    Nailed it, and bonus nachos for the diagram.

    Hopefully that helps

    Nice looking bits of kit, I haven’t seen analogue ones like that since trade school.

  5. #5
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default

    You're an absolute legend.......thank you so much for taking the time.

    I wish I did own those instruments but alas that is a borrowed internet picture. If I had $4000 spare I would purchase a second hand unit........but I don't..........so I will make some!!!!!!

  6. #6
    Join Date
    Sep 2009
    Location
    Newcastle
    Posts
    341

    Default

    There is a thread on practical machinist about DIY differential levels

    https://www.practicalmachinist.com/f...levels.291287/

  7. #7
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default

    Quote Originally Posted by pippin88 View Post
    There is a thread on practical machinist about DIY differential levels

    https://www.practicalmachinist.com/f...levels.291287/

    Did you notice the similarity in the picture between mine and that thread?

  8. #8
    Join Date
    Aug 2021
    Location
    Sydney
    Age
    46
    Posts
    116

    Default

    Professional audio is protected from common mode external noise by differential audio inputs. The theory of this was explained to me 20 years ago using a schematic involving a transformer input on a balanced audio transmission line. After 20 years of experience in pro audio I’ve seen an awful lot of audio transformers but very few of them used for cancelling common mode noise, it’s far cheaper to use operational amplifiers for for that function. They are also used for many other things involving differential voltage reference for threshold set points (e.g. turn on a relay when a voltage level reaches a chosen set point) and so I’m surprised to see such interest in older transformer based circuits. Op amps are also great for feedback in current and voltage drive circuits such as PWM based switch mode power supplies however single chip systems probably do this in a micro controller these days. I’ve set one up recently to set a constant speed in a motor controller irrespective of load. Many applications!

  9. #9
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default

    I believe that modern digital spirit levels and the like use an MPU6050 chip to measure level. These have nowhere near the resolution for measuring the flatness of a granite surface plate which requires sub-arc second resolution. Not sure what the electronics of the more modern differential levels are like - doubt if the manufacturers give up the details of their $20,000+ equipment readily!

    I believe modern electronic amplification and noise control is able to get better resolution, accuracy and repeatability from such old style equipment though. I will need to research this side of things quite well in the future if I start this project.

  10. #10
    Join Date
    Mar 2011
    Location
    Southern Flinders Ranges
    Posts
    1,536

    Default

    That’ll be quite a mind bending undertaking if you have a crack at building a set. The electronics to differentiate the voltages at the levels you’re looking for will need to have a bit of thought put into it. Happy to help where required

  11. #11
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default

    Quote Originally Posted by racingtadpole View Post
    That’ll be quite a mind bending undertaking if you have a crack at building a set. The electronics to differentiate the voltages at the levels you’re looking for will need to have a bit of thought put into it. Happy to help where required
    The bit that is totally doing my head in at the minute is the amount of movement of the shading loop. For these to measure 0.1 arc second, the sideways movement of the loop is only 0.00003937mm (if my quick and dirty maths is correct based on a 127mm long foot and a 100mm long pendulum). Mind boggling stuff - the amplification of the signal will be very important me thinks.I have 2 options;1. buy a brand new large surface plate that I can rely on out of the box2. buy a much cheaper used one then spend time and effort making a set of levels and a lapping plate to measure and lap to specI love the challenge of making things but maybe this is a bridge too far?

  12. #12
    Join Date
    Mar 2011
    Location
    Southern Flinders Ranges
    Posts
    1,536

    Default

    Yeah the amps and feed back loops are going to have to be next level precise. Power supplies are going to need to be no noise, not just low noise. You won’t be buying stuff from Jaycar for that.. you’ll be into precision component manufacturers for toleranced parts.

    I guess it depends on how soon you want a working surface plate. The cost of buying a new calibrated one is the price of not having to melt your head for potentially a very long time trying to get a home brew level set up working and calibrated to repeated measuring. A bridge too far.. that comes down to personal circumstances and preferences. For me, electronics is how I keep a roof over my head and the bills paid so I try not to have anything to do with it away from work. Your circumstances are different and you’ve likely got far more patience than I do….

    The irony of it is you’ll likely need a surface plate for some of the construction..

  13. #13
    Join Date
    Jul 2016
    Location
    Melbourne
    Age
    35
    Posts
    1,522

    Default

    Yeah the only reason to undertake that is because you want to make the level. In any other way buying the plate is the only way.

    Yes you can make them with laps and unlimited time, but its pointless and when you are done you have two enormous useless laps hanging around.

    When you consider how cheap a certified plate is at almost any size.

    Sent from my SM-G973F using Tapatalk

  14. #14
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default

    Quote Originally Posted by caskwarrior View Post
    ...When you consider how cheap a certified plate is at almost any size...
    Cheap is subjective! lol

  15. #15
    Join Date
    Apr 2019
    Location
    Adelaide
    Posts
    589

    Default

    On that cheap front I have been researching the cost of new plates;

    Vertex VSG-15 1000 x 1500 x 200 Grade A $8,000+ including freight
    Mitutoyo 2000-1000-250 @ $7,494Mitutoyo 1600-1000-200 @ $5,754
    Standridge Granite (USA) 1829 x 914 x 203 (2nd hand) $4,400 + freight (waiting on horrible quote)
    Starrett (NZ) assume he meant a large one at around $3,000 + freight but not sure if from US or NZ

    2nd hand plate might be $2,000-$3,000 (guesstimate) plus Repeat-o-meter $1000(?) plus autocolimator/electronic levels $1600 min(?) + cast iron lap $750(?) + a heap of learning and sweat. New is not going to be a world more expensive so def seems the go. Unless of course I could find a large 2nd hand plate of known A grade condition.

Page 1 of 2 12 LastLast

Similar Threads

  1. OT understanding an old engine spec plate
    By bob ward in forum METALWORK GENERAL
    Replies: 5
    Last Post: 30th Nov 2018, 04:28 PM
  2. Differential indexing
    By Michael G in forum METALWORK GENERAL
    Replies: 1
    Last Post: 26th Oct 2016, 02:58 PM
  3. A better understanding of rust
    By neksmerj in forum METALWORK GENERAL
    Replies: 18
    Last Post: 25th Oct 2015, 06:52 PM
  4. Oil levels and VSD's
    By Stustoys in forum METALWORK GENERAL
    Replies: 10
    Last Post: 23rd Jun 2011, 06:51 PM
  5. Reputation Levels
    By simon c in forum FORUMS INFO, HELP, DISCUSSION & FEEDBACK
    Replies: 7
    Last Post: 20th Jan 2005, 03:56 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •