0
\$\begingroup\$

Total noob here, please be gentle.

Back in 2010, this thread was started regarding audio delay curcuits: Audio delay using discrete components

After six years or so, I am wondering (hoping?) whether technology has caught up with this question, and that some expert in here (endolith?) can provide a simple and easy solution for building a delay circuit that fits my needs/fits the bill. After some research online, it seems that the minimum standards for this type of circuit, is a delay of approximately .015-.02ms. This is enough of a interruption in the microphone-speaker-loop, to eliminate most feedback issues, while still maintaining continuity between the speaker's mouth and the sound leaving the speakers.

Any takers? Thanks in advance for any help and advice!

\$\endgroup\$

closed as too broad by Voltage Spike, brhans, Daniel Grillo, Dave Tweed Aug 29 '16 at 21:03

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ I thought most modern feedback killers worked by blocking the high frequency of the feedback being sent to the speakers. You could in theory block feedback with a graphic eq. \$\endgroup\$ – Terry Gould Aug 29 '16 at 19:26
  • 1
    \$\begingroup\$ @magicchance It seems like you are asking for a design, this forum is more specific design problems where you post a design and then ask for help. It's also a good idea to do some research before posting as google.com/… yields scads of results. \$\endgroup\$ – Voltage Spike Aug 29 '16 at 20:03
  • 1
    \$\begingroup\$ Possible duplicate of Audio delay using discrete components \$\endgroup\$ – Voltage Spike Aug 29 '16 at 20:04
  • \$\begingroup\$ @laptop2d - I resent your tone. Not only have I googled this issue before coming here today, I have been researching this exact problem for at least 2 years! In that time I have used the google machine dozens of times, at least 3 times on this issue just today! Yes, on google I have come across some possible solutions to this question. And there are even some commercially available pieces of audio equipment I have seem, such as the Behringer Feedback Destroyer. But clearly, for reasons I do not feel I need to explain, these items must not completely fit the bill or my needs. \$\endgroup\$ – magicchance Aug 29 '16 at 20:11
  • 1
    \$\begingroup\$ You need to lose the attitude. Remember, you are asking the other users here for the favor of an answer. You won't get it if you antagonize the people who cared enough to comment on your question. Specifically, you DO need to explain exactly why what you have found on the Internet does not "completely fit the bill or my needs". \$\endgroup\$ – Dave Tweed Aug 29 '16 at 21:03
2
\$\begingroup\$

... it seems that the minimum standards for this type of circuit, is a delay of approximately .015-.02ms.

  • The speed of sound is 340 m/s at standard temperature and pressure.
  • Sound will travel 5.1 mm in 0.015 ms.

Move your loudspeaker 5 mm further away from the microphone and you'll have the same result without the complexity of further electronics.

Maybe there's something wrong with your research?


[From OP's comment:] "I can tell you that simply moving one of my pieces 5mm is not going to fix the problem."

The trouble is that that's all your 0.015 ms delay is going to do - it will make it seem that the speaker is 5 mm further away. You won't cure feedback - you're making a reverberation unit. You're missing some basic physics understanding somewhere along the way.

Your link to Audio delay using discrete components is a question about generating an audio delay so it's not clear why you think this will assist you with feedback reduction.

See what information you can find on constructive and destructive interference of sound waves. You will find that at certain frequencies and distances the sound will interfere constructively enhancing possibility of feedback whereas at others it will cause destructive interference reducing feedback.

Another area worth looking at is noise cancelling microphones.

\$\endgroup\$
  • \$\begingroup\$ As I said, I'm a total noob, but I can tell you that simply moving one of my pieces 5mm is not going to fix the problem. The type of interruption circuit I need will delay the mic from "hearing" it's own sound/noise as it leaves the speaker, just enough to prevent the loop from repeating. \$\endgroup\$ – magicchance Aug 29 '16 at 19:52
  • 1
    \$\begingroup\$ Transistor's point is : what on earth makes you think a small delay will accomplish that "breaking the loop"? \$\endgroup\$ – Brian Drummond Aug 29 '16 at 19:55
  • \$\begingroup\$ @BrianDrummond - As of 6 years ago when the message I linked to above was written, that was the way this sort of thing was dealt with. And not in a great way, as that older thread alludes to. As my post says, I am here inquiring whether technology has caught up with this question. For such a simple question at that, it sure does leave a lot of educated people scratching their heads it seems. \$\endgroup\$ – magicchance Aug 29 '16 at 20:15
  • \$\begingroup\$ No, the delay means the mic still hears itself, just a little later. Feedback suppression has historically been done with a pitch shifter, which coincidentally can be implemented using similar circuits to delays. The linked Q&A has no mention of using the delay for feedback suppression. And as Richard hints, DSP offers much more attractive options for feedback suppression than pitch shifting. \$\endgroup\$ – Brian Drummond Aug 29 '16 at 20:19
  • \$\begingroup\$ I didn't say suppression, I said interruption. As in "delay". As in, to delay the speaker output by .015ms to prevent the mic from exactly "hearing" itself. If this is the same thing, I apologize. I've already stated I am a noob, so obviously some of these definitions are going to be lost on me, until someone starts understanding my weakness in this area. \$\endgroup\$ – magicchance Aug 29 '16 at 20:21
1
\$\begingroup\$

Technology has caught up? No, technology has marched on. Analog and switched capacitor circuits are being replaced all over the place by DSP. Are being? Probably more like have been, many years ago at this point. Many of old 'classic' synthesizer and audio processor chips are long obsolete. DSP is more reliable, smaller, lower power, etc. And it is subject to Moore's Law, whereas analog circuitry is not, or at least not to the same extent. Capacitors, resistors, and inductors take up a LOT of space, so great pains are taken to remove them and replace them with transistors as much as possible. The only times you see these analog parts on chips these days are when you have no other option at very high frequencies (GHz) for things like radio transceivers, PLLs, VCOs, high speed serializers and deserializers, etc. It's far more efficient these days to throw down a tiny delta-sigma ADC and DAC and some DSP than to build analog circuitry, especially at audio frequencies.

\$\endgroup\$
  • \$\begingroup\$ Thanks for the tutorial, but can you also help direct me to what you are referring to? IE, is there an easy way to make what I need based on your above information? \$\endgroup\$ – magicchance Aug 29 '16 at 19:49
1
\$\begingroup\$

The best feedback cancellation uses two mics to cancel far field noise often used in major media events for public speakers.

Others use a variable f notch filter where the gain peaks due to various resonances and 180 deg phase shift. ( or aka parametric equalizer)

The obvious check is to ensure speaker polarity improves feedback for the casual user.

The low budget user might use an electret mic with a good wind screen and rear open for far field cancellation and used near the cheek or corner of lips to prevent puff noise.

It depends if the mic is used with a high budget or low and for near and far field sensitivity.

Other options are hall acoustic profiling and design of reflective and absorptive materials for concert use with expensive sensitive stage mic's or improved location of speakers.

\$\endgroup\$
0
\$\begingroup\$

It is not clear what the basic requirement/need are here? Are you seeking solutions for feedback-reduction? Are you asking for artistic effects like simple delay (or complex delay/reverb/echo)? Are you trying to use delay (or pitch-shifting?) as a method of feedback-reduction? Please clarify your question.

Or perhaps it doesn't make any difference if it is feedback-reduction or signal delay, because here in the 21st century, all of those tasks are handled completely in the digital domain by DSP. DSP has become so inexpensive and easy to implement that you can find it in everything from sophisticated scientific instruments to children's toys.

Note that pitch-shifting was never very tolerable, especially to musicians, so it died at rather an rather early age.

\$\endgroup\$
0
\$\begingroup\$

There's a near trivial way you can experiment with different delays, and that's to use the sound card of a PC.

It would be good for you to try out what can be achieved with delays, as I'm not sure your 0.02mS will do all you expect it to do. Better to program hardware you have, before building custom hardware.

Bear in mind that most feedback incidents involve a single tone, which as you increase the delay goes through multiple cycles of being in and out of phase.

PortAudio would be a good library to start with, it has bindings for most languages, C++ would be closer to the metal, though I prefer to use Python.

\$\endgroup\$
  • \$\begingroup\$ With a PC it will be very hard to get delay as small as that- normal audio latency is tens of milliseconds. \$\endgroup\$ – pjc50 Aug 29 '16 at 22:41
  • 1
    \$\begingroup\$ @pjc50 it would be impossible, even if you could do an assembler loop reading a single sample and writing a single sample. But it would allow the OP to experiment with other delays easily and find that none of them were useful, before wasting his time on hardware. As somebody pointed out, 0.02mS is 20uS, is 6mm distance. When doing audio, I can get feedback in a very large auditorium, or a small one. Interestingly, 0.02mS is not far off the sampling period of 48kHz. \$\endgroup\$ – Neil_UK Aug 30 '16 at 6:00

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