Can I measure an audio signal using a 15 meters BNC Cable?

[leinho]

Hello,

I would need to use a 15 meters BNC Cable (RG58) in order to measure an audio signal that goes directly to a pair of earbuds. This means, the test point would be exactly before the balanced armature speaker located in the earpiece. The idea is that I would leave a BNC jack in the DUT and connect this directly to an oscilloscope.
I need this setup in order to check signal integrity during EMC Testing and we would prefer to place the oscilloscope outside the chamber.
By the way, the device is battery operated that play sounds on earbuds.

[vincentvmac]

It is better to use twisted pair cable such as LAN cables which are less capacitive but remember to insert ferrite rings along the cable to prevent it from affecting EMC testing results.

[Kire Pûdsje]

I currently work for a test lab.
In general audio devices are measured by guiding the sound through a hollow plastic tubes to a speaker/microphone outside the chamber. Since you mention 15 meters distance I guess you are talking about a professional lab with a huge room. Any test lab should have this setup with tubes available for you. Another way is to place the oscilloscope directly underneath the turntable with an HD video camera directed at the screen. Again any test lab should have this camera setup available for you.
I do not know what standards Uruguay has, but in general 10% distortion is the limit for EMC testing, 15 meters of coaxial cable should not be a problem.

[leinho]

Oh, I did not know about the plastic tubes. Yes, I am talking about leaving the oscilloscope outside the chamber for monitoring and in order to not be affected by the immunity testing (I would need this setup only for this test).
10% distortion for what unit?
Actually, the testing will take place in the US, they seem to be happy with the 15 m BNC cable but will test how it works before performing the test. I believe it should behave like an RC filter and at these frequencies, it should not matter.

I might also need to leave a router with working internet connection during the test. I will consider that but I do not think I can use it for the audio.

[Kire Pûdsje]

Unfortunately, personally I am not really too well versed for FCC part 15B.

In Europe, the method using tubes for devices with microphones and/or speakers is even prescribed in some of the standards:
From e.g. the ETSI EN 301 489-05 v2.2.0
"The audio signal is, after loopback in the test system, picked up at the Ear Reference Point (ERP) by an audio
transducer provided with a non-metallic acoustic tube. The transducer signal is feed to an audio distortion
analyser located outside the test environment."

[leinho]

Thanks Kire. Actually, the test lab is going to test for EN301489 but have not said anything about this. Our target markets are Canada, US, EU and UK.
That line that you are referring is something I was thinking of doing but seems kind of crazy of having a tube of 5 meters attached to the earbuds. 5 meters is the length needed to locate the monitoring equipment outside the chamber. 15 meters is to reach their monitoring room.
The audio distortion analyzer is provided by the manufacturer, right?

They have the camera and I could leave it inside the chamber with a 1 meter BNC Cable, for example.

[Kire Pûdsje]

5 meters can still be done. Just do not use the silicone type, you need the harder plastics.
As said the test lab should have this available. Also if the test lab is accredited for telecoms audio measurements, they should have the audio analyzer available. (In general part of a modulation analyzer).
Just talk to the test lab they should be able to help you.

The bigger labs are more formal, with a clear separation between coordinators and button pushers. The first just handling the paper work. The latter just knowing how to start a test script. In the smaller labs, people tend to be more involved in all aspects of the work.
(Disclaimer: working myself in a location with 20 persons, although we have a logo of a big organisation on the front wall).

As for the 301 489, this is part of a bigger series. I do not know your product, the most likely candidates in general are the 301 489-3 or the 301 489-17. In both standards, the tube option is not explicitly mentioned. US labs are more strict, as the FCC is more prescriptive. This is what they are used to. For CE you have more freedom in testing, since in the end with the risk analysis, you as a manufacturer are responsible for the compliance.

Since the 301 489 series is currently not harmonized, be sure that the latest draft standards are used, because due to the RED transition, as soon as the EU decides to harmonize these standards, you need to apply them immediately, without the usual 1 year transition period. Also be aware of several changes the last couple of weeks in the blocking measurements, especially relating the new drafts of the EN 300 220 and the EN 300 440 (check whatever radio standard you are using). Finally I do not know what safety standard you are using for the RED, but be aware that the EN 60950 will be removed in 2018. To be prepared, use the EN 62368.

[vk6zgo]

Oh, I did not know about the plastic tubes. Yes, I am talking about leaving the oscilloscope outside the chamber for monitoring and in order to not be affected by the immunity testing (I would need this setup only for this test).
10% distortion for what unit?
Actually, the testing will take place in the US, they seem to be happy with the 15 m BNC cable but will test how it works before performing the test. I believe it should behave like an RC filter and at these frequencies, it should not matter.

I might also need to leave a router with working internet connection during the test. I will consider that but I do not think I can use it for the audio.

I would have thought that coaxial cable would have minimal effect at audio frequencies.
I doubt that it would even look like a transmission line .

Plastic tubing, on the other hand, may exhibit acoustic resonances at that sort of length.
I would be interested in how they work around this.

[Kire Pûdsje]

Coaxial cable would have no effect at all on the audio. At 100 pF/m and 15 meter, as long as the drive impedance is less than 5k, nothing should happen. For radiated EMC and telecom tests the coaxial cable could work as an antenna, hence the usage of the plastic tubing.

With short runs of tubing you can get resonances. With long runs not so much. Measurements should be made with a 1 kHz tone (0.3 meters wavelength). Therefore 5 meters tubing is already 17 wavelengths long, even longer for harmonics. Compare this with a long coaxial cable (in wavelengths) at RF frequencies with say 10 dB loss. With this kind of loss there is no way I can get it to resonate.

[leinho]

Thanks to both. I also thought that the plastic tube might alter the sound wave a lot considering its length but might work as a way of telling the difference from being radiated and when its in normal state.
What's the drive impedance? The input impedance of the earbusds' speakers?

We would only check the real time audio, with the oscilloscope, for ESD, Immunity Testing and Magnetic Field Immunity. Radiated and Conducted Emissions would not involve any real time measurement except those specified in the standard.
All the others requires the device being in charging mode where we can not play sounds. So, we would test before and after on those.
How does this looks to you?

I am currently really concerned about what you said here:

Since the 301 489 series is currently not harmonized, be sure that the latest draft standards are used, because due to the RED transition, as soon as the EU decides to harmonize these standards, you need to apply them immediately, without the usual 1 year transition period. Also be aware of several changes the last couple of weeks in the blocking measurements, especially relating the new drafts of the EN 300 220 and the EN 300 440 (check whatever radio standard you are using). Finally I do not know what safety standard you are using for the RED, but be aware that the EN 60950 will be removed in 2018. To be prepared, use the EN 62368.

Fortunately, the Wi-Fi module we use has been tested to the new RED (June 2013). Would this be enough? According to the certificate, they used EN 300 328 V2.1.1: 2016 for the radio, EN 62479:2010 for Health, EN 60950-1:2006 + A11:2009 + A1:2010 + A12:2011 + A2:2013 for Safety and ETSI EN 301 489-17 V3.1.1:2017 ETSI EN 301 489-1 V2.1.1:2017 for EMC.
Does this mean that we should use another module?
I do not think there is not much we can do about this. Right?

[Kire Pûdsje]

The drive impedance being the source that drives the coaxial cable. If the source impedance is higher, you will start to notice the capacitance of the coaxial cable.

As to the standards, unlike the FCC, the RED does not know of a modular approval (although there are some discussions ongoing in the RED-CA).
Of the reports, only the EN 300 328 and the EN 62479 of the WiFi module are of importance. You can use these reports as part of your risk analysis to cover the essential requirements of the end product.
The rest (EMC + safety) needs to be redone on the end product anyway.
The versions seem OK. For the complete product you might also consider for the 301 489-1/-17 the v2.2.0 respectively v3.2.0, as these might become harmonized as v2.2.1/v3.2.1. Just let the test lab include a statement in the test report covering both the x.1.1 and the x.2.0 versions. This will ease your duty of keeping the technical documentation and risk analysis up to date.

[leinho]

Thanks for your help Kyle.
In the quote provided by the lab, they say they are going to provide a report for EN 301 489-17:V2.2.1. Do I need a v3.2.1? I did not understand what respectively means in your comment.

[Kire Pûdsje]

As for the EN 301 489-17 v2.2.1, I think this is a typo by the other lab, as this standard has immunity up to 2.7 GHz. For the RED, you have to assess immunity to 6 GHz.
Since none of the EMC standards are harmonized under the RED, the best is to take the latest versions.

Generic base standard:
Latest approved by ETSI: EN 301 489-1 v2.1.1
Draft underway: EN 301 489-1 v2.2.0 (Will probably be harmonized as v2.2.1)
Changes: Correction of circular and generic references and annex A according to the EC RE-D Desk officer
http://www.etsi.org/standards-search#page=1&search=%22301489-1%20V%22&title=0&etsiNumber=1&content=0&version=1&onApproval=1&published=1&historical=1&startDate=1988-01-15&endDate=2017-10-30&harmonized=0&keyword=&TB=&stdType=&frequency=&mandate=&collection=&sort=1

Broadband data transmission:
Latest approved by ETSI: EN 301 489-17 v3.1.1
Draft underway: EN 301 489-17 v3.2.0 (Will probably be harmonized as v3.2.1)
http://www.etsi.org/standards-search#page=1&search=301489-17&title=0&etsiNumber=1&content=0&version=1&onApproval=1&published=1&historical=1&startDate=1988-01-15&endDate=2017-10-30&harmonized=0&keyword=&TB=&stdType=&frequency=&mandate=&collection=&sort=1

As said, the test lab should be able to state compliance with both the latest ETSI approved version as well as the lastest draft.

[leinho]

Thanks Kire, I will make sure of that.
Regarding what you mentioned: "Another way is to place the oscilloscope directly underneath the turntable with an HD video camera directed at the screen"

Do you think this would be better in terms of affecting the results compared to have a 15 meters BNC Cable and leave the oscilloscope outside the chamber? Let's say being the BNC Cable both shielded.

The test lab has an HD Camera that they can use to monitor the oscillsocope. 

[leinho]

Regarding the plastic tube, we have tried a signal of 10dB re 1V and there is not enough signal for the test microphone to get a reading through the 12 Meters of tubing. What kind of devices are tested this way?