Why do oscillation dampers use signal conversion through a sufficiently massive electrical circuit (with resistors, capacitors, diodes) to create antiphase, instead of simply shifting the signal in time through the timer and controller calculations (and then releasing it through mechanical oscillations by the inverse piezoelectric effect)?

In theory, it is possible to slow down the signal by increasing the magnetic flux by winding or by mutual induction through the passage of current in the opposite direction.

I have just started to study this topic, so the question may be very primitive or contain errors of understanding.

  • $\begingroup$ Can you show an example of what you're referring to? When I was searching, I found plenty of passive dampers, and plenty of active dampers using micro-controllers. What use cases are you looking at where they specifically use networks of components? There may be a domain-specific reason for it, like longevity. $\endgroup$ – Cort Ammon May 26 at 2:41
  • $\begingroup$ Model of active damper. It is usually controlled by a feedback loop. Controller choosing from next types: PID, LQG, PPF. For passive models electrical circuit needed for energy dissipation. For active models, if I understood right, - for correcting signal parameters. However I want try to create more simple and reliable model for study project. $\endgroup$ – Dashwind May 26 at 4:11
  • $\begingroup$ worldwide.espacenet.com/patent/search/family/024028837/… $\endgroup$ – Dashwind May 26 at 4:28

We use those networks of components in engineering because its actually rather hard to do the time delay circuit you describe. If your input is a sine wave, you can detect its frequency and determine the delay needed to provide an anti-phase output. However, if your input is more complex, like a sum of multiple sine waves, or even worse... real inputs, it can be very difficult to tease them apart and get the phase behaviors correct. Those networks surrounding an op-amp are known to have very specific phase characteristics, regardless of the frequency spectra.

You can also design those networks to the specific dynamics of your object. Sometimes the best damping isn't precisely a simple anti-phase relationship, especially if your materials have non-linear behaviors.

You can, of course, use a microcontroller for this. There are digital filters which are designed to emulate the analog filters in those electronic networks. However, they are expensive. In a business setting, you never use a 70 cent micro-controller for what you can do with a 20 cent op-amp. They are also much easier to manufacture and have much better longevity.

If you are doing a study, you may wish to have flexibility. It is much easier to modify an algorithm than it is to re-spin a circuit board. For study purposes, a microcontroller may be the correct tool for the job.

One thing to consider is the output. A microcontroller is not well suited for outputting the analog signals you need to stabilize the system. You need some sort of D2A (Digital to Analog) device. All of them cost money, which is why you don't see them in industry unless there's a real advantage. DDS based converters, in particular, are very effective for this kind of output, but are comparatively expensive.

Also, remember that most of these devices produce signals, but not power. They tend to be unable to drive loads. You need an amplifier attached to them to take care of the load. This often involves a similar network of resistors and capacitors to properly bias the amplifier, so you are likely to find those networks appear again, in the end.

  • $\begingroup$ Thanks for answer! Of course, I thought about it and it is only appropriate for harmonic vibrations. With smooth fading, if the delay is minimal, then the difference will not be critical. At least, it seems to me, the efficiency will still be higher than in systems with similar networks, an example of which I described above. $\endgroup$ – Dashwind May 27 at 5:57
  • $\begingroup$ So I'm looking for information on the Internet, but it's so hard to find something like that about time delay systems. At least exemplary implementations, where one could read much more about the components of the system, research partially determines the choice. If you have any advice, it would be great. $\endgroup$ – Dashwind May 27 at 5:58
  • $\begingroup$ Why to complicate the electrical network management system in those versions through feedback or adaptive filtering. Since the vibrations are finite, and not like, for example, an engine running constantly in time. Each time there is a change suitable for not just a correlation, but a significantly different one. Well, for example, on the racket - the ball hit the spot, and the ball hit the bottom of the strings, or even the rim. Isn't it advisable to just count one result over and over again instead of wasting resources on calculating the error as well? $\endgroup$ – Dashwind May 27 at 5:58
  • $\begingroup$ @Dashwind Typically you tie it to the dynamics of the system. The event, such as a ball hitting a racket, is a single event, but it propagates across the racket according to the dynamics of the racket. This network permits them to prevent things like ringing which might occur with a simpler approach. And it also lets them use all of the tools that they have at their disposal. From experience, someone who understands control theory can use their tools to get the right answer faster than I can use my limited tools to make a simple answer. $\endgroup$ – Cort Ammon May 27 at 14:50
  • $\begingroup$ You might be interested in the Inverse Z Transform. It is a digital equivalent of the Laplace transform that you see used to make those analog component network, so it can do (almost) the same thing in software. $\endgroup$ – Cort Ammon May 27 at 14:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.