2
\$\begingroup\$

I've created a custom, bluetooth audio harness for my Jeep WJ using off the shelf parts.

  1. Amp - Clarion XC1410
  2. Converter - 12v to 5v buck converter
  3. Bluetooth Audio - Logitech Wireless Speaker Adapter

12v is supplied by the vehicle stereo harness in the dash. The amp and converter both connect to the 12v here. The converter supplies power to the BT adapter and the audio output from the adapter is connected to the amp. I have a constant alternator whine (vehicle on) in varying degrees when configured this way. Today I isolated some components to identify the source of the noise. While the converter is connected to an isolated power source the noise disappears. I would prefer to have the converter connected to a switched 12v power supply in the vehicle. Can anyone help me figure out why this is happening and how to address it?

\$\endgroup\$
1
\$\begingroup\$

From your description it's probably a ground loop.

You're presumably supplying ground to your DC-DC and BT at both power in and signal out (to the amp).

I'd break the loop, probably by having the connection to the amp grounded at only one end of the cable. I wouldn't break the loop at the power side because if you were to do that then DC-DC would have to use the amp's ground as a return path, which would be less than ideal.

An audio isolation transformer, like this one from ebay, should do what you need with minimal fuss.

\$\endgroup\$
  • \$\begingroup\$ How do I break the loop? \$\endgroup\$ – jcconnell Jan 13 '15 at 5:43
  • \$\begingroup\$ Are you referring to the RCA cables or the speaker outs or both? How would I break the loop in any of those? \$\endgroup\$ – jcconnell Jan 13 '15 at 6:00
  • \$\begingroup\$ I'm referring to the RCA cable/s between the BT module and the amp. They are (should be) a co-axial cable, with a centre conductor carrying the signal and a ground shield wrapped around (but insulated from) the centre conductor. That ground shield is joined to the connectors at both ends of the cable. You want it to be joined to the connector only at the amp end. RCA cables are cheap, try hacking one up. \$\endgroup\$ – markt Jan 13 '15 at 6:04
  • \$\begingroup\$ Actually, another option is to find an audio isolation module. I'll have a look around on google for you... yeah, search for "audio isolation transformer", I got immediate hits. Several references to "hum killer". Here's one from ebay, $10, built-in leads: ebay.com/itm/… \$\endgroup\$ – markt Jan 13 '15 at 6:05
  • \$\begingroup\$ Wow I can't say thanks enough for the details! Should I go with the RCA cables or transformer first? Additionally, I'm using higher-end RCA/3.5MM cables. Would using the low quality cables to test this be alright? \$\endgroup\$ – jcconnell Jan 13 '15 at 6:11
1
\$\begingroup\$

There is a likelihood that your power supply is contributing to modulation of the signal.

If you go to web sites obsessed with hand-crafting high fidelity high power audio amplifiers, you would realise our obsession with attaining extremely low noise/signal ratio.

There will be a point of improvement where the s/n ratio would be at a standstill due to signal modulation contributed by the power supply. So much so that a complex power supply design is inescapably an essential part of a hand-crafted ultra high fidelity audio amp.

But my preamble simply serves to illustrate that a power supply fluctuation does modulate the audio signal.

These are the reasons of power supply fluctuation

  1. Ground loops, the usual suspect
  2. Voltage dips due to interaction between variation in current drawn and internal impedance of power supply. Where voltage dips are directly proportional to the current drawn and internal impedance.
  3. The power supply being too small a hammer to hammer a big nail.

There must be more than sufficient materials on the internet discussing how to mitigate ground loops. A ground loop forms an antenna broadcasting and receiving signals whose wave-lengths are within the vicinities of the main harmonics of the effective diameter of the loop.

Therefore a ground loop also forms a resonant impedance to certain frequencies. Where modulation noise due to the power supply is more pronounced at those frequencies. With impedance, comes fluctuations.

An obvious but oft ignored cause of signal modulation due to power supply dipping is the power supply being too small a hammer for too big a nail. For example, if you are driving a 12V output, your input supply cannot be 12V, which is not your case.

Also, a 15 W peak supply can at most drive a 10W mean output. So, if your bluetooth is consuming 2 A, 15W is insufficient juice. I don't really know how much power bluetooth consumes. When I speak on my LG phone unplugged for 5 minutes, it stayed at 100%. When I have it on bluetooth (unplugged) for 15 mins listening to amazon music on my car amplifier, its charge dipped to 92%. I conclude that bluetooth must be pretty intense consumer of energy.

I am thinking these are the possible reasons for your hum

  • 15W is not enough juice. Therefore your bluetooth preamp is noise-modulated by the 15W step-down.
  • Your car has a lousy 12V supply. I am not well-informed in automotive electrical systems. Is it normal for the alternator to noise-modulate the battery's 12 V output? I can't even parallel park a car properly, so someone else needs to answer this question. Maybe you should try borrowing a strong big car battery and see what happens. As far as I have heard, the car battery should be a very stable supply of DC.
  • Do your car lights dim or fluctuate when you turn up your audio? I am not an expert in this either. Does the hum frequency vary with the engine RPM?
  • Maybe 15W is enough juice, but it is noise-modulated by your unclean battery supply. Such that even if you had a 25W step-down, the problem would remain unsolved.

You could experiment with a low pass filter (parallel-capacitor + series-inductance pair) at the 12V input of the step-down, using appropriately power-rated components. I suppose we could calculate the henries and farads using http://en.wikipedia.org/wiki/Capacitor-input_filter. Which I don't think is viable due to the resultant sizes of the components.

After all these troubles, you might as well spend $60-$80 on a well-meaning car audio that comes with bluetooth, which also would come with its own respectable switching supply, and you would use it as your preamp.

\$\endgroup\$
  • \$\begingroup\$ Apparently it's not possible to add all my responses in one comment so I'll split them below: \$\endgroup\$ – jcconnell Jan 14 '15 at 15:52
  • \$\begingroup\$ I tested the BT power supply with both a large, 12v marine deep cycle battery and a very small 9v battery. The 9v connects to the converter, and the converter to the BT audio adapter. Admittedly, I have been using the 9v battery since Monday to listen to clean music. I made a temporary harness that I plug in when I get in and disconnect when I get out. I'm actually surprised it has lasted this long but that leads me to believe that it's power draw is not that significant. \$\endgroup\$ – jcconnell Jan 14 '15 at 15:53
  • \$\begingroup\$ The alternator in my Jeep is new and outputs a consistent voltage of about 13.8v (Measured using the On-Board Diagnostic port). I've also installed and aftermarket, oversized AGM battery (Sears Platinum Diehard). In the Jeep and offroad community these are very highly regarded. The battery is nearly brand new. This leads me to believe the power system is working well. \$\endgroup\$ – jcconnell Jan 14 '15 at 15:53
  • \$\begingroup\$ The vehicle makes no indication of excessive draw through lights or otherwise. I don't play my audio that loud though. I've taken every opportunity to replace lights with LEDs and I've retrofitted the headlights with HIDs. The hum does fluctuate with engine RPM. Does this extra information help in your diagnosis? \$\endgroup\$ – jcconnell Jan 14 '15 at 15:54
  • \$\begingroup\$ Finally, I chose to create my own harness rather than purchase a head unit because I've installed an Android tablet into my dash. The tablet has replaced the headunit altogether. I've tried several other routes including an Alpine headunit and a Pioneer AppRadio but ultimately I prefer the tablet. The most convenient method of interfacing the tablet with the factory audio was using bluetooth (So the tablet may be removed). While I agree that purchasing a headunit is a much easier route, the tablet experiment is more fun and rewarding. I also have no place to put the headunit. \$\endgroup\$ – jcconnell Jan 14 '15 at 15:59
0
\$\begingroup\$

How much current does your Bluetooth module consume?

If it's only a few mA, there is a simple trick that might help.

First, you need to do a simple experiment. Power the Bluetooth module with your isolated power supply. If the whine is NOT present, then connect only the ground wire from the Bluetooth module to where you had it connected when you were getting the noise. Note that your isolated power supply is still powering the module.

What we need to find out is where the whine is coming in: power or ground.

If the noise is still present while the Bluetooth module is being powered by an external power supply but the grounds are connected, this trick probably won't help. BUT: if the noise is gone, this trick should work.

What you proved was that the noise was coming from the 12V power supply rail. The trick is to remove that noise.

It turns out that it is very easy to do this. A transistor configured as an emitter-follower has a very high impedance at the collector. That high impedance is what is going to block the noise.

schematic

simulate this circuit – Schematic created using CircuitLab

Zener diode D1 is 5.6V. I use the 1N5200 series in situations like this - they require less current for regulation than the 1N4700 series.

Transistor Q1 drops about 0.65 Vdc across the E-B junction. Pick the Zener voltage to be about that amount higher than the desired output voltage.

Note that there is NOT an input bypass capacitor. This circuit configuration doesn't need it and you do NOT want to couple that input noise to your signal ground. That is one of the main reasons for using this circuit topology.

You want to do a star-point ground at the power amp input. This regulator and the Bluetooth module both connect to that point. The amp is then grounded as normal to the vehicle battery.

Grounding is just as important as having a clean power supply. Doing a star-point ground at the amplifier input is the best way to keep noise out of that ground.

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

This is definitely a ground loop caused by the bluetooth module. In this video (see link below) on YouTube, a guy shows you the solution. You need an isolated DC/DC converter between the powersource and the bluetooth module.

Probably it will NOT help to isolate the audio or both amplifier+bluetooth at the same time because mostly ground of the modules are connected together. So if you isolate the input alone or just the powerline for both modules at the same time, the amplifier and blutooth module still share the same ground and the ground-loop still exists.

Module i'm talking about (isolated DC/DC converter, 1W is mostly enough): isolated DC/DC converter ($4.48 for 5 pieces on Aliexpress)

(datasheet: http://www.mornsun-power.com/uploads/pdf/B_S-1W.pdf )

Video with demonstration: https://www.youtube.com/watch?v=b2Pb1KuVDS0

\$\endgroup\$

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.