Shop OBEX P1 Docs P2 Docs Learn Events
Torroid Ferrite Core Eliminates Servo Jitter — Parallax Forums

Torroid Ferrite Core Eliminates Servo Jitter

ercoerco Posts: 20,244
So those things actually do work! Pretty dramatic demo. I just ordered ten small cores for $3: http://www.ebay.com/itm/22mm-x-14mm-x-8mm-Power-Transformer-Ferrite-Toroid-Cores-Green-10Pcs-GY-/262033919263

Comments

  • WhitWhit Posts: 4,191
    Guess it time to start junking for old radios!
  • Duane DegnDuane Degn Posts: 10,588
    edited 2015-11-12 22:31
    Some of that explanation didn't make sense. He said something about the HXT900 servos being sensitive to noise on the PPM control wire but I doubt the ferrite core is doing anything with the signal wire.

    I hope someone who knows more about this sort of stuff jumps in but I thought adding the ferrite core resisted changes in current so it smoothed out the current spikes of the digital servo.

    It's still a good trick (if you don't mind a big metal ring added to your project) but it would be nice to learn how it reduces the jitter.

    I'd also like to know if there are any negative side effects from using the torriod. What happens to the voltage levels as the ring inhibits the current from changing? It would be interesting to compare oscilloscope traces of both the signal lines and the power lines, with and without the toroid.

    It seems like the majority of servo jitter problems are caused by power supply issues but not all jitter is caused by power supply problems. Servo jitter can also be caused by bad pots.

    Those HXT900 servos have (IMO) low quality pots (I don't know if they're worse than other servos but I don't think they last as long as they should). I have to replace servos on my small hexapod every so often because the pots wear out. They usually wear out in the "home" position since the wiper rubs this area of the resistive film more than other areas. As the pot starts to wear out, the servo has a harder time finding its correct position which can cause a small jitter. Once the jitter starts, it becomes a destructive feedback loop as the increase in jitter wears the resistive film out even faster than under normal use.

    I still like the HXT900 servos. I really like how they have a full 180 degrees of travel. I'll need to take the servos with worn out pots and use erco's method to convert them to CR servos. I hear HXT900 servos make nice CR servos.
  • WhitWhit Posts: 4,191
    edited 2015-11-12 22:49
    OFF TOPIC (a bit) - I was really thrown into a bit of jitter myself when it wasn't erco voice on the demo! HA!
  • I don't know him that well, I thought it was Erco, the voice fit the face.
  • That's great to know. I saw someone make a toroid to filter out motor noise from an fpv camera.
    I'm going to grab a few from dead electronic equipment.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2015-11-13 16:06
    http://www.goldmine-elec-products.com/products.asp?dept=1050

    Here you go... new ferrite toroids in all sorts of shapes and sizes.

    How do ferrite beads work? http://www.jameco.com/1/3/ferrite-beads
  • Hams use em all the time. I used to tear apart old CATV matching transformers for the cores. Always helped with some noise reduction (common mode?)
  • ercoerco Posts: 20,244
    I mainly recall seeing them on power wires from old-school heavy iron core wall warts. Haven't seen them on the lightweight new switching types. They have a magical effect in that video, I find it hard to believe that just 2 turns would do much of anything.
  • Used them on home brew low level signals and power leads. Ferite cores were also used in old transistor RF amps and signal converters.
  • potatoheadpotatohead Posts: 10,253
    edited 2015-11-14 17:31
    Re: Two turns...

    Isn't the "magic" the fact that the ferrite is extremely permeable?

    A couple of turns is enough to act like a low pass filter, and as mentioned it's great for pushing the noise floor down, while leaving the signal intact. More turns would distort the signal, due to hysteresis impacting it more directly.

    Long ago, when I was doing HAM radio, the ferrite was a staple in many things I did. Purpose then was to manage RF noise. I also used them a lot in anything having to do with a TV, which being AM in the video signal, basically displays any noise it sees... I still do this. My old gaming consoles have a larger ring, and two to four turns to clean up the RF output. Basically, strip the processor and logic noise from the video signal going out. They work great.


  • "Used them" ... "Seen them" ... but do you know what is taking place?

    By adding a Ferrite core you effectively increase the inductance of the wire at the location of the Ferrite. More turns = larger inductor AND/OR larger Ferrite core = larger inductor.

    An inductor is always in opposition to a voltage change (think like a spring or shock absorber) , so by adding inductance, any sudden spikes are attenuated.

    You could just as easily add an inductor in series and get the same result, but think about it for a moment, the construction of an inductor ... With an inductor you have the advantage of multiple turns on a small core ... with the method described in the top of this post you have a disadvantage of a few number of turns around a large core ... either way you slice it your increaseing the inductance to the wire by a specific amount. The aspect ratio of core to wire is all you are changing here.

    ... Rather than just throwing some value at the problem, I'm a little more OCD than that, we should really understand what it is we are doing and WHY it works when we do it a particular way.

  • potatoheadpotatohead Posts: 10,253
    edited 2015-11-20 16:44
    That's a great video above. It's basically what I was taught a long time ago about these things. Enjoy.

    I do think it's interesting how the inductor doesn't get the love it used to these days. Is it just the requirement for mass in the project?

    Edit: This guy has a great channel. Recommended. I've been watching some basics this morning. He's clear, fun, and insightful.



  • Very good! Explains it better then I could!
  • Do we want to add a ferrite bead to the Vcc on our prop projects?
    As long as I'm stocking up, I may as well get everything.
    I'm thinking Part no. 1844580 from loopy's jameco link.
    Is this what we want?
  • Heater.Heater. Posts: 21,230
    spud,

    Inductors get a lot of love these days. Have a look in the power supply of your PC, it's choke full of inductors. ("choke" get it?). Switched mode power supplies need them. Transformers, mains chokes, the buck/boost inductor...Same goes for all the other millions of SMPS in wall warts and so on.

    Then there are the surface mount inductors, hardly distinguishable from SM caps or resistors, quite amazing.



  • potatoheadpotatohead Posts: 10,253
    edited 2015-11-20 17:28
    Indeed they do!

    I was thinking about radios mostly when I wrote that. Recent ones have unfavorable AM characteristics, and frequently FM is sub-par too. Just yesterday, the peeps at work decided to spiff up our lobby. It's big, boring, and well, just a lobby. So we are dropping a pool table in there we got for free, adding some cool lights, and they wanted a stereo.

    So I went and got them an actual "stereo", peak gear from the late 70's, from a little hole in the wall shop run by a guy who really cares about this stuff. (These guys have vinyl, and can appreciate a good cartridge and stylus. I know that I will enjoy whatever they decide to play on it.)

    Turntable, receiver, etc... The FM receiver in that thing is insane good, as is the AM one. In a building full of computers, that AM section can pluck out a station and do so with what I would say is "unnatural" good noise rejection compared to just about anything else I've put my hands on recently. And it has the classic, ferrite antenna bar. Nothing beats those things on AM, save for long wire, or other well tuned and constructed antenna, and of course, an inductor to match the whole thing up for optimal reception.

    And it does so while offering up a good 8Khz of bandwidth too. --lost on the many US stations clipping it hard at a mere 5Khz, which is a crime when it comes to the human voice. 10Khz is perfect, 7-8Khz makes all the difference in the world, but I seriously digress...

    So many units today are 3-5Khz, which is awful, and seem way more susceptible to noise.

    A look under the hood shows a TON of inductors. Big ones. :) As I've watched those go away, or get much smaller, so has overall radio performance changed.

    I wish I understood those kinds of circuits better. I get the distinct impression that the overall possibilities involving the more massive inductors have their place, given the performance I see from circuits we see today. That's all.

    But, as low-pass, noise filters, their use is obvious enough. I like 'em, and use as directed.
  • This is a little late but an inductor in series does perform like a low pass filter, but since inductors resist current change, they produce voltage spikes.

    Generally, you'd want to have a capacitor in parallel with one end tied to ground and the other tied to the power line. With a cap in this configuration, it'll be a low pass filter and it will provide instantaneous current since capacitors resist voltage change.

    Ideally, you'd want to use both of these with the inductor in series driving (before) the capacitor in parallel. The value of the inductor should be substantially lower than the value of the capacitor (you're wanting an overdamped circuit but trying to get close to a critically damped circuit).
  • ercoerco Posts: 20,244
    Thanks for that useful info, Howard. It's never too late, some discussions go on for several years. Welcome to the forums!
Sign In or Register to comment.