2
$\begingroup$

So I was operating a Laue machine recently which generates x-rays and fires them at a sample, and I noticed that there was a high-pitched whine coming from the instrument (nothing abnormal, just something I noticed). Both the current and the voltage could be set separately; typically values are 10 mA and 20 kV. I noticed that if you change the current, the frequency increases or decreases with it. A frequency spectrum app on my phone showed that the frequency was roughly proportional to the current being applied, so, for example, at 5 mA there'd be a 2000 Hz sound and at 10 mA it'd be 4000 Hz. (I don't remember the exact frequencies. I do remember there were lots of harmonics excited.) The voltage, on the other hand, only caused a small change of the frequency even if doubled.

My question is how would such a sound be generated?

My best guess is that the vacuum tube where the electrons are accelerated has its current adjusted by moving either the anode or cathode, and that forms a variable capacitor which happens to modulate the resonance frequency of the circuit such that the frequency and current increase together. Is this reasonable?

$\endgroup$
  • $\begingroup$ I'd have to say the answer is: just about anything. Somewhere in there is a tank circuit whose resonant frequency depends on the current (or as a dependent variable) voltage, and it's makking something vibrate mechanically. <-- not that that's a lot of help. $\endgroup$ – Carl Witthoft Nov 16 '13 at 17:28
2
$\begingroup$

Prime suspect is the switching power supply that generates the bias voltages for the tube. As the anode voltage is changed to increase the cathode current (the anode doesn't physically move), the power supply must work harder. Some switching power supplies operate at a fixed frequency, and so wouldn't display this behavior, but others have a variable frequency that increases with the load. (Also, the frequencies you report look sort of reasonable for a switcher.)

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Most switching power supplies operate outside the audio frequency range because the noise is easier to filter and it won't directly pollute any audio circuitry. However magnetostriction can cause audio noise at lower frequencies than the operating frequency of the electronics. $\endgroup$ – user6972 Oct 24 '14 at 20:12
  • $\begingroup$ @user6972, I concur about power supply switching frequencies in general; it all depends on the supply. However, high voltage supplies tend to run at lower frequencies because of the high effective capacitances presented by the high voltage transformer. Moreover: 1) I claim magnetostriction as part of the power supply, since the transformer is an element in it. 2) a steady sound usually requires a driving signal at the same or a related frequency, so just saying "magnetostriction" without identifying an underlying frequency source is not an answer, imho. $\endgroup$ – Art Brown Oct 24 '14 at 21:00
  • 1
    $\begingroup$ Obviously there has to be energy oscillating, I didn't see that as part of the question. Depending on saturation levels and free induction decay rates the non-linear effects often shift the mechanical vibrations to a variety of frequencies. My point was the rate at which the magnetostriction occurs can be less than the driving frequency. $\endgroup$ – user6972 Oct 26 '14 at 2:08
  • $\begingroup$ @user6972, ok, got it. $\endgroup$ – Art Brown Oct 26 '14 at 2:13
2
$\begingroup$

Typically it is the ferrite cores in inductors/transformers that resonate mechanically, or through magnetostrictive effects that produce a high pitched whine. Switching PSUs are the main culprit. It can also occur when the EM fields interact with steel components in the PSU.

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

I appears that you are asking two questions. 1) what can generate a high pitch noise, & 2) why is the frequency of the noise directly proportional to the current applied?

The general answer to 1) is, anything electromagnetic (inside or outside the vacuum tube) "loose," that resonates in the range of the mentioned frequency.

The answer to 2) can only be a guess, and its as follows: $$ E = P \ (Energy = Power) $$ $$P = VI \ (Power = Voltage\times current) $$ $$ E = hf $$ $$ \ therefore, \ f = \frac {VI}{h} $$ Since V (voltage) and h (plank's constant) are constant, this equation shows that the frequency (f) is directly proportional to the current (I).

| cite | improve this answer | |
$\endgroup$
0
$\begingroup$

To convert one voltage to a higher voltage one often uses a transformer or a ladder network (or a combination of these).

enter image description here

The above from Wikipedia shows a Cockroft-Walton multiplier - commonly used to step up an alternating waveform. It acts like a bucket brigade - the charge on each capacitor is being passed on to the next capacitor, at a higher voltage, after each cycle of the input.

For this circuit to work you need an alternating input voltage - which is most efficiently generates from a DC voltage by creating a square wave (less switching loss), which is sometimes stepped up part of the way with a transformer.

Finally, the output voltage of this multiplier depends on the current drawn - to maintain a certain value you must carry enough current up the ladder and since the amount of charge transferred during one cycle of the circuit is constant for a given output voltage, you need to increase the frequency linearly with output current to regulate properly.

Finally - the currents in the transformer can be quite high, and this will generate forces and this displacements in the circuit that can be perceived as audible sound. A square wave typically contains the 3rd, 5th etc harmonics of the fundamental, although even terms can appear due to circuit nonlinearities and high frequencies can be attenuated due to finite switching speed and material properties.

There are many other potential causes - but many of the above considerations will apply to those too.

Fun fact - the Cockroft-Walton generator was not invented by Cockroft and Walton although they did won the Nobel for work they did with it (particle accelerator to cause nuclear changes in atoms). See

| cite | improve this answer | |
$\endgroup$
  • $\begingroup$ Certainly the more frequency content you put into a transformer the more likely you'll hit a spot that causes a mechanical resonant frequency. But I wouldn't say the clipping/distortion of AC waveforms are the cause. $\endgroup$ – user6972 Oct 20 '14 at 8:16
  • $\begingroup$ This doesn't require resonance. OP stated the pitch is rising continuously and there are harmonics. That is entirely consistent with a system that has the components I describe including a "badly designed speaker" that concerts currents into mechanical vibrations. If the driving force was a sinusoidal wave you would see fewer harmonic components. But a switching supply isn't a "distorted sine wave"- it is intended to be a square wave. $\endgroup$ – Floris Oct 20 '14 at 9:04
  • $\begingroup$ I'm not talking about an electrical resonance. In order for a sound wave to be heard through the mechanical structure a resonance has to physically occur. $\endgroup$ – user6972 Oct 24 '14 at 20:16
  • $\begingroup$ I don't know why you say that. A well designed loudspeaker makes sound - but it has no resonance (otherwise some frequencies would be louder than others). Maybe you and I have a different definition of resonance? $\endgroup$ – Floris Oct 24 '14 at 23:59
  • $\begingroup$ A speaker of any type is specifically designed to physically resonate mechanically at the frequency it is driven at. This is why tweeters have a small diameter and bass speakers have large diameters. A "loudspeaker" attempts to have a wide bandwidth and resonate from about 1200hz to 5600hz all in one physical design. When a coil or transformer generates an audio sound it is because it is physically moving at a rate that supports vibration due to the physical construction of the device in the audio range, otherwise you wouldn't hear it because it wouldn't vibrate. $\endgroup$ – user6972 Oct 26 '14 at 1:58

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.