Bluetooth project - Noise, Line input

[fgomes]

I've been using the KRC-86B V3.3 Bluetooth module (Bluecore 5 based) in a wireless speaker and have some issues, since there are some other posts referring this module perhaps someone has already faced them and found some solution.

One issue is some noise appearing when I pair a device with the module. Until a device is paired with the bluetooth module there is no noise in the module output, but after pairing there is some noise, not very loud but it can be listened when the music is more silent. My setup is not very clean since I've not designed yet a PCB for it, my doubt is if this can be cleaned up with a decent PCB design and ground separation between digital and analog part, or if this is a problem with this type of module and impossible to be solved.

Other thing I noticed is that the module has two audio inputs (left and right), but my devices detect it as a Multimedia audio device without any input, so I wasn't able to send audio from this module to the device paired to it (to use it as a hands free, for example). This is a firmware limitation?

Best regards

Fernando

[Richard Crowley]

The reason you hear no noise before pairing is because the output is MUTED. Bluetooth is NOT a high-quality audio transmission media. It is a convenient shortcut for casual use.  It is unreasonable to expect high-quality, noise-free performance from any Bluetooth scheme, especially one as remarkably inexpensive as that KRC-86B.  It is what it is.

Do you have drivers that support input mode for your module?  Your symptoms would seem to indicate that you do not.

[fgomes]

Hi Richard, thanks for your help. I was expecting some noise and some distortion, but not so much 'digital' noise, like we find in some 'not so good' sound cards, where we can listen the computer activity in the speakers or headphones. What I hear is a 'digital' noise, perhaps generated by the module CPU with a power supply not very well decoupled used also for the analog part. Since my prototype is far from perfect (perf board without any ground plane) my question was to find if this could be at least partially solved with a good design of my PCB, or if it is just impossible with this type of module.

Regarding the use of the analog inputs on the board, I didn't use any particular driver, just paired the module with my PC (Toshiba R630 with Win 7 64bits) and with my mobile phone (Samsung Galaxy S2). In the PC it was considered a headset device, but only added an output device, not an input. In the mobile phone, it was considered a multimedia device, only output also. This is a driver issue in the PC/mobile phone, or a profile issue in the module?

Best regards

Fernando

[Richard Crowley]

You could apply extraordinary methods to try to quiet down your Bluetooth module. Adding power supply bypass/filter capacitors, etc. etc.  But Bluetooth was never designed to support "high-fidelity" audio transmission, so it seems like a losing battle IMHO.

There are literally dozens of Bluetooth protocols and drivers must be written to implement each protocol.   IMHO, it seems quite normal to find that not all the protocols are implemented on any given device.

[andybren]

I know this was posted over a year ago, but for posterity's sake I wanted to add a bit of info as I didn't find any other topics referring to this issue on Google.

I'm using the  KRC-86B V4.0 module hooked up to a single board IC power amp and experienced the same problem as the OP. i.e. when the BT module is connected and active you can hear a high pitched 'digital noise'.

Just wanted to clarify on this post that these noise issues are NOT coming from the BT's lack of sound quality or bad drivers as per the previous response, but rather are coming from the power supply.

My original configuration was using a perfboard circuit with a l7805 IC (with two filter caps 10/100µF) dropping down the 12V input from a standard (cheap) 2A wall-socket transformer for the BT, which was being shared with the power amp.

On a hunch I reconfigured the circuit to power the BT module independently using a 5V phone charger (only a bit cheap) and that resolved the noise problem completely. Just to be thorough I also tried powering the BT independenty using a second (cheap) 2A wall-socket transformer through a different step-down circuit and that had the noise again.

So.. For posterity's sake, if you experience this error, check your power.

I'm still fairly new to this stuff, so I'd be very grateful if anyone could take the time to explain what this noise might be and if there's a way for me to filter it. I live in Thailand and the wiring in buildings here is notorious for being problematic. On a different project, I came across the idea of ground loop noise, but never really found any conclusive info on what to do about it... Could this be the problem here?

[Richard Crowley]

Thank you @andybren for your additional information.  Of course there is more than one kind (and source) of "noise". And other audio artifacts (such as intermodulation-distortion) are perceived by many people as "noise".  I did suggest additional filtering and bypass capacitors, and @fgomes admitted that his project was "far from perfect (perf board without any ground plane)".

The term "noise" is so loosely defined, it is almost useless as a debugging symptom.  Ground loop problems typically produce symptoms of power-mains-frequency hum.  That would be 50Hz (or 60Hz in some places) and the first harmonic, 100Hz or 120Hz.  And if it is bad enough, even more harmonics up the scale.  Many audio processing software applications include "hum filters" which have notches at 50, 100, 150, 200, 250, etc. Hz.

And the "quality" of "wall-wart" (and "line-lump") external power supplies varies all over the map. Some are really well filtered and quiet, while others are "raw" and depend on additional filtering and regulation internally to the device they are powering.  So using random power supplies will be expected to give varying results such as you are describing.

[andybren]

Thanks for clarification Richard.

Could you point me in the direction of what I can do to identify and mitigate problems such as this? I know I can just get a better power supply, but the point is that I'd like to learn to troubleshoot problems like this.

I have an interest in audio electronics and I've put together a 35 watt A/B power amp from a kit before and am now experimenting with IC amps in conjunction with Bluetooth, Arduino and the Raspberry Pi.

I think I'm familiar with the typical ground loop 60/120Hz hum, and this is a very different sound (imagine "digital noise" in the context of a sci fi movie and you'll be pretty much spot on). However, it's distinctly caused by the power supply so I would think that maybe it's some kind of artefact of the signal processing done by the bluetooth and still a result of a ground loop problem.

So, could I kindly ask you for a few bits of knowledge/advice?

1) Is the difference between the crappy wall-wart and the 5V phone charger a question of quality (e.g. better filtering) or does the charger work in a different manner that solves the problem?
2) Would it make sense for me to use an oscilloscope to check the quality of the DC coming in from a power supply? Would this be a viable/conclusive approach to identifying problems like this?
3) Can you, in general terms, point me in the direction of what I need to do to solve this problem? "Additional filtering and bypass capacitors" I gather would be the first thing, so I'm going to read up more on this topic now (but any advice would be welcome).

Many thanks in advance.

[Richard Crowley]

I think I'm familiar with the typical ground loop 60/120Hz hum,

If you are in Thailand (as your flag indicates), then you have 50Hz mains, don't you?
60Hz is mostly North America and a few scattered non-conformists around the rest of the planet.

and this is a very different sound (imagine "digital noise" in the context of a sci fi movie and you'll be pretty much spot on). However, it's distinctly caused by the power supply so I would think that maybe it's some kind of artefact of the signal processing done by the bluetooth and still a result of a ground loop problem.

Without any explanation of the larger picture (what gear is involved, how is it interconnected, how is each piece powered, etc. etc.) it is almost impossible to make any specific comments.  A few seconds recording is worth 10000 words of description.

1) Is the difference between the crappy wall-wart and the 5V phone charger a question of quality (e.g. better filtering) or does the charger work in a different manner that solves the problem?

No, it essentially comes down to how well filtered the supply is.  Some tasks (like charging batteries) don't care how much noise there is, so it is cheaper to make "noisy" supplies whose only (or major) task is simply to re-charge batteries.

2) Would it make sense for me to use an oscilloscope to check the quality of the DC coming in from a power supply? Would this be a viable/conclusive approach to identifying problems like this?

No, you don't need an oscilloscope to work on this kind of noise.  It might be handy to have one, but it is not essential. You have already demonstated that you can tell just by listening the difference between a better filtered supply vs. a noisy one.

3) Can you, in general terms, point me in the direction of what I need to do to solve this problem? "Additional filtering and bypass capacitors" I gather would be the first thing, so I'm going to read up more on this topic now (but any advice would be welcome).

Shielding and grounding (and in some ways, filtering and bypassing) are almost a "black art". Much (most?) effective measures to reduce noise come from experimenting with the specific circuit and physical configuration of the gadget.  Having a handful of capacitors of various values (from 0.1uF up to 100uF or even 500uF) is valuable for experementing to see where the filtering or bypassing provides the most benefit for the specific circuit configuration.

[NiHaoMike]

Could it be ground bounce? Try adding 470uF or so of capacitance right at the module's power rails and put a 1 ohm or so resistor in line with the connection from the power supply to the power rail, so that the capacitor supplies the transients.

[andybren]

Richard, thanks for taking the time.. 

Yes, you are completely right, it is 50hz. Momentary brain fart from reading too much US-oriented material.

Shielding and grounding (and in some ways, filtering and bypassing) are almost a "black art". Much (most?) effective measures to reduce noise come from experimenting with the specific circuit and physical configuration of the gadget.  Having a handful of capacitors of various values (from 0.1uF up to 100uF or even 500uF) is valuable for experementing to see where the filtering or bypassing provides the most benefit for the specific circuit configuration.

That would explain why it's so hard to find concrete information about the subject. I've come across a lot of people giving specific suggestions but no info about consistent and repeatable procedures to solving the problem. (I'm a programmer by trade, in case your wondering  )

I guess I will just have to experiment and get a feel for this kind of thing.

Just one last question then.. Would it be viable for me to put a breadboard in between my power supply and the rest of my components so that I can experiment with various forms of filtering, or is a breadboard too noisy in itself? This would be on the 12V DC coming in from the wall-wart, which is then used by the amp (12V) and dropped down to 5V on two separate 7805 regulators for the BT module and an arduino nano.

If so, then whatever optimal filtering solution I find for this... Will that be specific to the wall-wart (i.e. filtering the noise generated by that) or specific to my components' requirements (i.e. filtering to remove the problems that my components are most sensitive to). In other words, if I put in a circuit that solves my problem here, will it work with a different power supply?

Again Richard, thank you very much for your help.

Could it be ground bounce? Try adding 470uF or so of capacitance right at the module's power rails and put a 1 ohm or so resistor in line with the connection from the power supply to the power rail, so that the capacitor supplies the transients.

I'll admit, Googling "ground bounce" turned out to be a bit of rabbit hole.. I'll have to spend some more time getting my head around that, but thanks for the tip.

In the meantime, I'm using a 7805 with three caps (100uF, 10uF and 0.1uF) as per the second schematic on this page: http://electronics.stackexchange.com/questions/8121/capacitor-sizes-for-7805-regulator with the BT module being the only thing connected to that. The 12V DC supply coming in is shared between multiple components. Would your suggestion be applicable in this case? If so, would I place this in between the 7805 and the BT module (i.e. directly on the BT's power rail)?

[Richard Crowley]

Would it be viable for me to put a breadboard in between my power supply and the rest of my components so that I can experiment with various forms of filtering,

NO!  Much of the "black art" consists of placement of components (mostly capacitors, but sometimes resistors, inductors, or even active regulators) at specific (and CRITICAL) physical locations. Especially for high-frequency noise/interference, a few mm can make the difference between complete silence and no relief at all.  If you read the data sheets and application notes for many high-frequency (digital and/or RF) you will see that the manufacturers VERY STRONGLY recommend bypassing as close to the chip power terminals as you can possibly get. Sometimes they even "hide" small SMD capacitors UNDER the chip to get the capacitor right at the power pin, etc.

or is a breadboard too noisy in itself?

EXACTLY!  Filtering lower-frequency (i.e. mains) noise can be done "in bulk" farther away inside the power supply, or perhaps in the case of an external power supply, where the power enters the enclosure.  But higher-frequency noise generally requires bypassing/filtering closer to the critical part of the circuit.  In part because even the short distance between the bulk filter and the load is, itself, a receiving antenna to collect noise, both from the outside environment, and often from the circuit itself.

This would be on the 12V DC coming in from the wall-wart, which is then used by the amp (12V) and dropped down to 5V on two separate 7805 regulators for the BT module and an arduino nano.

What does your circuit sound like with the digital devices (BT & Arduino) temporarily disconnected from their regulators?
What does your circuit sound like with the digital devices powered separately.
That sharing the raw supply between the analog parts of the gadget and the digital parts is a PRIME SUSPECT of "digital hash" noise leaking into the analog amplifier.  Certainly, it can be done cleanly, and maybe that is not the source of your noise, but it is very often a critical point.

If so, then whatever optimal filtering solution I find for this... Will that be specific to the wall-wart (i.e. filtering the noise generated by that) or specific to my components' requirements (i.e. filtering to remove the problems that my components are most sensitive to). In other words, if I put in a circuit that solves my problem here, will it work with a different power supply?

If you were designing something to be manufactured inexpensively for mass-market, that is one thing. But since you are a hobbyist making a one-off gadget, you can use a technique that is more expensive, but also more universal and forgiving of the quality of various external power sources.

You can just assume that ANY external source of power is very noisy, and design your circuit to work well with a low-performance source. Then it will work with ANY quality supply.

In the meantime, I'm using a 7805 with three caps (100uF, 10uF and 0.1uF) as per the second schematic on this page: http://electronics.stackexchange.com/questions/8121/capacitor-sizes-for-7805-regulator with the BT module being the only thing connected to that. The 12V DC supply coming in is shared between multiple components. Would your suggestion be applicable in this case? If so, would I place this in between the 7805 and the BT module (i.e. directly on the BT's power rail)?

I would at least throw a significant bulk filter right at the 12V node coming in from the external supply. The point where the power splits off to the analog amp and the regulators for the digital parts.  It is not clear why you need two separate regulators for the digital loads?

The two most common causes of digital hash noise getting into the analog circuit are through the power supply, and "transmitted" from the digital parts and "received" by the analog circuit.  In some cases, you see that parts of a circuit are physically SHIELDED to prevent "reception" (or "transmission") of noise/interference both externally AND internally.

I don't know what "ground bounce" means, either. But @NiHaoMike has a good suggestion for further filtering the 12V to the analog amp part of the circuit. Essentially a "pi filter" with a large input and output capacitance, and the 1 ohm resistor between as a form of isolation.

[andybren]

That's really valuable feedback. Thanks Richard.

It is not clear why you need two separate regulators for the digital loads?

The Bluetooth and the Arduino are currently two separate modules. For now, I'm just experimenting with integrating misc components together before I come up with an integrated design.

What does your circuit sound like with the digital devices (BT & Arduino) temporarily disconnected from their regulators?
What does your circuit sound like with the digital devices powered separately.
That sharing the raw supply between the analog parts of the gadget and the digital parts is a PRIME SUSPECT of "digital hash" noise leaking into the analog amplifier.  Certainly, it can be done cleanly, and maybe that is not the source of your noise, but it is very often a critical point.

I've done various such experiments to seperate and isolate potential issues and it's clear that the shared power is not the problem. The problem either stems from the wall-wart or the circuit around the 7805.

Good tip about sharing digital/analog power though. Will definitely keep that in mind.

Anyway, thanks again for the advice. I'll need to read a bit more about some of the principles behind this stuff, but in any case I have a much better idea of where to go with this now and what to try. I'll be sure to post a follow-up for the benefit of others if I find a solution.

[hansV]

Hi all,

I also recently bought a KRC-86B V4.0 and wired it up to an amplifier to build a wireless speaker. Just as described in this thread I have the same cracking digital noise as soon the module gets paired. The amount of noise is not influenced by the volume setting on the bluetooth module. If you put volume on 0, you still have the same noise.

What I found out is that if I cover the antenna on the board with my fingers, the noise goes away. What could this mean? (addition: Touching the backside of the board in general makes the noise go disappear)

I have a 12V battery powering the amplifier. To power the bluetooth module, a voltage stepdown module is used. Could this module be a source of the problem?

I have to say, this is the only in depth thread I found concerning this problem. So thank you for helping to dig this out.

Best,
Hans

[Richard Crowley]

The amount of noise is not influenced by the volume setting on the bluetooth module. If you put volume on 0, you still have the same noise.

That is an EXCELLENT debugging test and observation.  Very good! 
By turning the volume down to zero, you effectively eliminate/isolate the part of the circuit "upstream" of the volume control.
That would seem to indicate that the noise is getting in to the circuit AFTER  the volume control.

As I said before, the two primary ways of noise like this getting into the signal stream are:
1) Through the power supply.  If the supply bus is not sufficiently filtered/bypassed, then it could have noise superimposed onto the power, and that will get into the signal stream.
2) Radiated into the wiring.  Wires from one place to another can be TRANSMITTING ANTENNAS and other wires can be RECEIVING ANTENNAS.  There are techniques for filtering both on the "transmitting side" and on the "receiving side". 

Sometimes we can filter the source of the interference to reduce its ability to transmit noise.  For example, a common source of noise is a "Phase-fired Control" (PFC) like a common SCR or Triac lamp dimmer.  We add largish inductors to the output of these dimmers that reduce the very steep rise-time of power going to the lamp. That very steep rise-time (or very steep fall-time) of a signal is a prime source of high-frequency harmonics that spread far up into the spectrum and find their way into other sensitive circuits (like an audio amplifier, etc.)
Ref: https://en.wikipedia.org/wiki/Phase-fired_controllers

And sometimes we can filter the "receiving side" by bandwidth-limiting the signal path to only the audio frequencies of interest. i.e. filtering out the frequencies above 15~20kHz to minimize the amount of that digital hash/noise that leaks into the audio path.

And you will note in some equipment parts of the circuits are enclosed in their own little metal shielded box to either protect them from interference getting into the circuit, or, in other cases, to prevent the circuit from "transmitting" unintended noise out into the world.

I also recently bought a KRC-86B V4.0 and wired it up to an amplifier to build a wireless speaker. Just as described in this thread I have the same cracking digital noise as soon the module gets paired. What I found out is that if you cover the antenna on the board with your fingers, the noise goes away. What could this mean?

I can think of a couple of likely suspects, but there almost certainly other possibilities as well...
1) You could be creating a makeshift "shield" preventing the RF signals from leaking back into the audio signal.
2) You could be actually preventing the RF (Bluetooth) from communicating so that it temporarily stops trying to pair.

[hansV]

Good news.

I wrapped the module in a layer of electric tape an after that I wrapped some aluminium foil around it, not covering the antenna area of the board. Guess what, all the distortion is gone .

Hopefully this is also the solution for you Andy.

Thanks Richard for all the insight!

Have a nice one.

Hans

[hemidenis]

Thanks a lot guys to discuss this topic, it is no much info on the web really.

I love the discussion about power sources, but powering it  it with a 12v battery would eliminate any power source noise. And someone pointed it already, the noise persist even doing that.

Another thing I tried was powering the Bluetooth with a switching voltage regulator, and actually I tried 3 in series with the same results.

So my conclusion is that the Bluetooth is generating a noise of some kind and transported to the amplifier using the ground.

Separating the BT and amplifier to different power sources actually cure the problem 100%, but this is no a solution for a Bluetooth installed in a car for example.

The aluminum paper didn't work for me.

But here is the bomb shell, I have a fm/usb/bluetooth device that has no noise perfect sound and I have another device that has the same Bluetooth module and it is noisy, so it is a way to filter this and we have to find it.

[hemidenis]

And just in case, in another forum an Einstein told me:
   
"The problem here is pretty clearly insufficient bypassing of the Bluetooth module. – Matt Young"

Of course This Albert didn't say how to fix it..... But someone in here can probably  make something out of it

[Alex_MGM]

I do not know it this is still actual as I found this forum only today, but this topick is interesting for me.
I use BT modules in a small project and met the same problem of noise.
Looking through different resourses I discovered interesting thing.
Normally audio BT modules have 2 separated groubds: analog and digital. If these 2 grounds are really separated - everything is more of less ok, no nouse. But sometimes (maybe often) manufacturers specially place jumper to short digital and analog grounds.
It is not critical when headphones are connected directly to BT module and no more sircuits are used. But if you want BT module to be operated in more complicated environments - many different noises (very unpleasant) arises.

[imk]

Hello,
I have same board and same noise issue (junk board).
However i have another BT board off ebay that also has SD/MP3 player and this board is noiseless :-) in either mode, paired or not.
similar to this but PCB is white
https://www.ebay.co.uk/itm/Mini-Bluetooth-4-1-MP3-Player-TF-SD-Card-Lossless-Decoder-Micro-USB-3-7-5V-Board/292599774644?hash=item44204e3db4:m:mEZKS465ErVnM7htJRCcOoA:rk:6:pf:0


I must say when i received this board and found the big bypass cap in the bag i had my suspicions, but was surprised how bad this board is.
Hope that help Ian

[Jon D]

Hi all
I came across this post searching for some other (non-noise related) info on the krc-86b and thought I had better post some info that may help since it seems lots of people have landed here looking for the same thing.

As some have already pointed out, the noise is coming from the bluetooth's interference with the power supply and running the bluetooth and amplifier modules on separate supplies solves the issue. I came across a far simpler solution when making my first bluetooth speaker. You can put a 4 pin dc-dc isolator between the bluetooth and amp supplies eg a B0505S-1W. That one is 5v. I soldered it in line with some shrink wrap for strain relief so it looks like a neat power cable with a bulge in the middle running from the bt module to the amp. Speaker is dead quiet!

Hope this is of some help to someone!