Audio devices without audio passband

[sonic tac]

Hi, if your audio gear (devices, recorders, amps, any audio stuff, etc.) doesn't have any filter above 20kHz. What noises can be produced in your audio. Like can it get Mhz or GHz signal? How? Can you show waveforms how the
high frequency signal can affect the audio? Thanks.

[ef37a]

Hi, if your audio gear (devices, recorders, amps, any audio stuff, etc.) doesn't have any filter above 20kHz. What noises can be produced in your audio. Like can it get Mhz or GHz signal? How? Can you show waveforms how the
high frequency signal can affect the audio? Thanks.

The HF response of audio gear is a matter that has been argued about ever since it became possible to record frequencies beyond 20kHz!

There are those that insist a mic pre e.g. must have a response to 200kHz or even higher to capture "all the harmonics of music signals". My personal view is that this is largely marketing bllx. There are very few capacitor microphones that have a response past 20kHz and many have 'drooped' even earlier. There is no MC or ribbon mic that goes past about 18kHz to my knowledge? Why then would you want a pre amp with a response into the low RF spectrum? Best start rolling off at around 50kHz and keep the ever increasing RF muck out!

There are of course special situations where ultrasonic signals need to be detected (James can probably give us chap&verse on that?) but for music? Not necessary IMHO.

Of course, when recording with a digital system your response will be limited to just under the sampling rate so why make the rest of the system open to RFI?

Anyone who has worked in the "real world" of knockabout domestic audio and PA will side with me. The coming of the Silicon planar transistor caused us all a couple of decades of grief!

Dave.

[ajay_m]

The main issue in the absence of filtering would probably be the capability for wide bandwidth electrical noise causing intermodulation distortion e.g from switching PSUs and led lighting (the pwm hash from these can really affect AM radio reception so the energy spectrum below a few MHz is significant).

It's instructive to put a finger on an oscilloscope probe and see not just the expected 50Hz mains signal but a huge background of wide spectrum noise from radio signals plus all the digital hash modern gadgets generate. This crap you definitely do not want getting into your preamps.

[ef37a]

Indeed Ajay and mobile phones are particularly penetrating. The triode input stage, if designed with half a brain, is largely immune to reasonable RF fields but a phone within about a mtr will "chirrup" I found.

I do wonder if pro studios ban MPs? Lock them in a metal box? Since one bleep could ruin hours of prep!

Dave.

[Wonks]

There was this forum thread back in 2016, where a decent wide frequency range was discussed, with the reason for going up to 200kHz was for reasons of reducing temporal smearing and getting as linear a phase response as possible in the audio band.

https://www.soundonsound.com/forum/view ... hp?t=48630

[James Perrett]

It may be worth pointing out that, from previous posts, the OP is working on a sensor based project rather than an audio project. For some reason they are trying to use audio gear rather than potentially more appropriate devices intended for wideband measurements.

[Hugh Robjohns]

Hi, if your audio gear (devices, recorders, amps, any audio stuff, etc.) doesn't have any filter above 20kHz...

Everything has a defined bandwidth. The only question is what that bandwidth is. It would be a foolish designer who didn't impose some degree of input RF filtering.

What noises can be produced in your audio. Like can it get Mhz or GHz signal? How?

Signals in the MHz and GHz bands are around us all the time, from radio and TV transmissions, Wi-Fi, Bluetooth, mobile phones, car and garage door openers, remote controlled switches and so much more.

Properly designed equipment should keep such signals out, but if they don't there is a risk that non-linearities in the circuitry will intermodulate with them, producing anything from demodulator signals to weird noises to a significantly increased noise floor.

[Hugh Robjohns]

I do wonder if pro studios ban MPs? Lock them in a metal box? Since one bleep could ruin hours of prep!

I've never been in one that has banned phones. If the gear and cabling is designed right it shouldn't be a problem.

That said, I normally activate airplane mode on my own phone when working on location and ask those I'm working with to do the same, but that's to avoid unwanted phone ringing and pinging mid-session rather than unwanted electrical interference noises.

One of my standard review tests is to place my phone on the equipment and listen for unwanted noises. It's only happened a couple of times — once was when reviewing KH310 monitors. It later transpired a previous reviewer had opened the back panel and dislodged a screen connection. Oops.

And I did once get some mobile phone interference during a concert recording rehearsal in Tewkesbury Abbey. A cellist had placed her handbag containing a phone right on top of a mic cable, and I could hear occasional low-level chirping. I swapped the cable out for a starquad which cured the problem completely.

[merlyn]

Perhaps worth bearing in mind that if you can hear interference, what you're hearing is below 20kHz, otherwise you wouldn't hear it. We don't hear the actual kHz, MHz or GHz signal, but some product of that in the audio band.

There are live recordings of Jimi Hendrix where the radio comes through. Not static, but an actual broadcast. I would imagine the broadcast was AM (amplitude modulation) and Hendrix's stack of amps is acting as an aerial and a filter that filters out the high frequency carrier, leaving the radio broadcast.

[James Perrett]

I would imagine the broadcast was AM (amplitude modulation) and Hendrix's stack of amps is acting as an aerial and a filter that filters out the high frequency carrier, leaving the radio broadcast.

I'd guess that he may have been using pedals containing germanium transistors. They are more likely to translate the radio signals to audio - in the same way that a germanium diode does in a crystal set.

There are also other techniques for translating high frequency signals to audio such as heterodyning. This is used for a musically in the Theremin.

[Philbo King]

You might consider looking into 'software defined radio' circuits.

https://www.digikey.com/en/articles/lea ... ined-radio

[Wonks]

A lot of solid-state audio devices contain op-amps in the audio circuit and depending on the model and configuration of the op-amp, a lot of these will stop the passage of signals in the MHz region.

In an open loop configuration, op amps have a flat gain response up to a certain frequency, and then the gain drops off as the frequency rises. Using feedback in the op amp circuit lowers the maximum gain, but increases the flat response area of the op amp, but the fall-off characteristic remains the same.

https://www.electronics-notes.com/artic ... sponse.php

Different op amps will have different frequency cut-off characteristics, but most selected for audio circuits will block the passage of those Mhz and GHz frequencies (or drop them to inaudible levels) without the need for dedicated filters.

It's far from the only way these signals can be blocked/reduced, but it's certainly why one reason why audio devices, like say a mixer, have a limited bandwidth.

Some op amps can pass much higher frequencies than others, so you can't assume that because a device has an op amp that it can't pass a MHz signal, but for those selected for normal audio use, it's unlikely.

[merlyn]

Once electromagnetic interference is in the system, filtering isn't going to make any difference. The part of the EMI that is a problem are the components in the audio band. Filtering with a cutoff at say 1MHz won't remove the components in the audio band, only the high frequency carrier. Better to stop EMI getting in in the first place, so use balanced cables.

[James Perrett]

Following on from Wonks' post, I've just seen that Rod Elliot recently posted an article on gain vs bandwidth a few months ago.

https://sound-au.com/articles/frequency-vs-gain.htm

It looks well worth reading if you want to know about extended bandwidth audio circuits.

[Nazard]

Thanks for the link, James, a good read with morning coffee!

The op-amps I have used the most are, in descending order: NE5532/5534, LM4562, TL072, OPA604 and AD797. The latter only in very low noise RF receivers and then an overkill. The OPA604 is clearly not one of Mr Self's favourites, but I used them in my first revision of the original Quad 405.

[Martin Walker]

Following on from Wonks' post, I've just seen that Rod Elliot recently posted an article on gain vs bandwidth a few months ago.

https://sound-au.com/articles/frequency-vs-gain.htm

It looks well worth reading if you want to know about extended bandwidth audio circuits.

That is indeed a fascinating read, with lots of useful examples. Thanks James!

[sonic tac]

No one has mentioned this. If I'm very interested only in the 1mV audio signal magnitude range. Can strong RF signals swamp/overload the amplifier so that low signal and low frequencies can't be seen (how low)? What part of the audio (what magnitude and frequency) can strong RP signals really suppress, just to have an idea?

Thanks.

[James Perrett]

RF signals won't swamp the amplifier if the system has been properly designed.

[Hugh Robjohns]

As James says, if the whole system is designed properly, RF won't cause any problems at all. But that 'whole system' starts with the design of the sensor, its connecting cable, the preamp, its output connection, the overall gain structure, the converter, and its connection to the computer. Everything matters, and messing up one element will compromise the overall results.

[sonic tac]

Let's take the example of the system not properly designed. I just want to understand how exactly RFI noises can drown milliVolt signal.

Supposed you used the INA321EA chip (not optimized for audio). And you plug microphone into it. And you don't use any Butterworth low pass filter. How exactly can the RFI drown the 1 milliVolt signal? What exactly happens when an amplifier is swamped/overloaded with RFI?

[Hugh Robjohns]

I just want to understand how exactly RFI noises can drown milliVolt signal.

Unwanted RF interference gets into the front end of the circuitry through poor screening or poor grounding. It causes various intermodulation effects due to non-linearities in the circuitry, potentially resulting in severely corrupted signals and an unusably high noise floor. The specific results obviously depend on the specifics of the circuit design and its implementation.

Good, standard-practice circuit design techniques to restrict the input bandwidth appropriately (which requires knowledge of the source sensor's bandwidth, sensitivity, dynamic range and other relevant parameters), along with good constructional techniques of the unit itself, combined with appropriate cable interfaces, will all help to keep RFI out and prevent problems.

If you're not sufficiently competent in electronic design and construction it might be worth soliciting the assistance of someone who is, or read a good book on the subject like the Art Of Electronics (https://artofelectronics.net/)

[James Perrett]

If you're not sufficiently competent in electronic design and construction it might be worth soliciting the assistance of someone who is, or read a good book on the subject like the Art Of Electronics (https://artofelectronics.net/)

I'd echo this - I had often seen the Art Of Electronics recommended and we had copies where I used to work, but I only bought my own copy recently. Reading it now, I wished that I had bought it earlier. While it contains a great deal of theory, it is also very practical with plenty of useful ideas.

P.S. In my early days in oceanography I worked on gear that Winfield Hill (one of the AOE authors) had designed. It was a cassette based data logger but it was the most solidly built cassette mechanism that I have ever seen. I seem to remember that it was cast from a single lump of metal with the capstan directly driven by a stepper motor. The replay unit was similarly impressive - it was all computer controlled and could read a whole C60 length tape in just a few minutes.

[sonic tac]

Months ago I asked whether to get the Behringer U-Phoria UMC204HD or Focusrite Scarlett or the E1DA Cosmos ADC and APU. I was suggested the latter because it was supposed to be made for quality recording with low distortions. So I bought them including the latest Scaler. But do you know the frequency response of them is very high.



APU Frequency Response Range Pre-AMP Preamp: 10Hz-1MHz+ -3db@60db gain, 10Hz-5MHz+ -3db@34db gain

Scaler Frequency Response Range .2Hz-570kHz @ 26db gain +.5/-3db. It can stretch up to .2Hz-7.11MHz @ 0db gain

ADC Frequency Response Range 5Hz-180kHz @ 32Bit/384kHz

What are the frequency responses of say the Behringer, Focusrite or even the M-Studio? Are they 0 to 20,000 Hz?

I don't know if I made the right decision buying the E1DA. Couldn't Radio Frequency Interference (RFI) distort the audio signal esp in 0.5mV level? Can it affect the ADC (without amplifier inside it too?)

My signal is only 0.5mV level mainly. Can the RFI swamp them such that the 0.5mV can no longer be seen?

Also what separate audio bandpass filter (0 to 20,000 Hz) module can I get that I can add to the E1DA to filter the RFI so it won't drown the 0.5mV signal? Should it be before or after the say APU or Scaler? Thanks.