PDA

View Full Version : RF interference in SB2's headphone output



ephemere
2005-07-14, 21:27
I just got a wireless SB2, and there is RF interference in the headphone output causing clearly audible noise. I know it's RF because it happens whenever there is any wireless communication (clock ticking, display scrolling, loading the buffer with a new song, etc). Removing the antenna makes the interference go away, but isn't an option because I'll be using it far away from the router. I tried this with two different headphones (AKG K501 and Etymotic ER4P). These are not powered headphones, and the noise happens only when the headphones are plugged in. That would seem to suggest that the fault is with the SB and not my headphones.

Thinking it was a defective unit, Slim Devices sent me a replacement, but it had exactly the same behavior. I did not test the line out on either unit.

Does anyone else have this problem? Is this a known issue?

UPDATE: Slim Devices is rolling out a fix. Read to the end of this thread.

seanadams
2005-07-15, 01:06
Line outs and headphone outs use separate amplifiers internally. I will have a look at it - hasn't been a problem before but there may be some simple fix.

Dave D
2005-07-15, 04:54
Removing the antenna makes the interference go away

There are two antennas; only one is removeable. This is a good hint as to where the noise entry is.

I wonder if there's a bit of crosstalk on the PCB traces feeding the high-impedance input to the headphone amp?

I also wonder how many folks are using the SB2's headphone output and wireless simultaneously, because (unless you got unlucky with two bad units) others should see the problem.

seanadams
2005-07-15, 07:05
I wonder if there's a bit of crosstalk on the PCB traces feeding the high-impedance input to the headphone amp?


I suspect so - it's a high impedance through the CMOS switch that provides the geekport capability (serial and IR data coming soon, I promise!). Had to wire it this way in order to keep the RCA line outs clean. Probably not much we can do about it except to bypass the data capability.

MeSue
2005-07-15, 07:48
I also wonder how many folks are using the SB2's headphone output and wireless simultaneously, because (unless you got unlucky with two bad units) others should see the problem.
I am. I received my SB2 two days ago and I requested an exchange for a problem of buzzing while the headphone jack was in use. After reading this, I tried removing the external antenna but it didn't make any difference. I'm not sure if my problem is related to RF interference or not… how would I check?

More on my issue: If I twist the plug in the headphone jack the buzzing comes and goes, and if I turn the speaker volume way down (these are PC speakers) and get the cord situated just right, I can listen for a while without any buzzing. The same thing happens with different PC speakers. With headphones it's not as loud, but there is still a faint crackling.

I sure hope my replacement unit is OK. Other than this problem… I love this thing!

ephemere
2005-07-15, 10:27
First of all, thanks to Sean and folks for looking into this. I'll give you guys a call to discuss if you want me to hold on to my second unit to run any tests or experiments.

To answer MeSue, the RF interference I'm hearing is not a constant buzzing. The way to tell is to turn the unit off, so that it's displaying the clock. Then plug headphones in (not PC speakers). If you hear a <click> every time the seconds change on the clock display then that's the RF issue I'm having. Also, PC speakers are different from headphones because they're (often) powered and pick up RF interference on their own, not coming from the SB. One test you can do is to unplug the speakers and see if the buzzing is still there.

Also, for Sean and others, moving the antenna has an effect on the sound, as well. I've tried different positions -- vertical, horizontal to the left, horizontal to the right, and horizontal sticking straight out the back. Each position has a certain effect. The position with the least interference is sticking straight out the back.

seanadams
2005-07-15, 10:58
If looks like the headphone amp chip itself (not surrounding traces) is susceptible to picking up the 802.11 signal (it does not appear to be DC power related). There's probably not much that can be done short of shielding that whole section. I'm trying a few experiments to see if there's an easy fix. Anyway yes, antenna orientation definitely makes a difference - by pointing the external antenna either straight out or towards the power input plug, I can make it go away. YMMV though - the radio will switch dynamically between the external and the internal antenna as needed.

MeSue
2005-07-15, 11:36
[QUOTE=ephemere]One test you can do is to unplug the speakers and see if the buzzing is still there.[QUOTE]

No buzz with speakers unplugged. I guess my problem is something else. Thanks.

ephemere
2005-07-15, 14:54
Sean, can I ask how the headphone section is different from the line outs? They're both variable, but do they differ in max voltage, current delivery, impedance, etc.? I'm wondering if I can just use a mini-to-RCA adapter and use the line outs for my headphones? Note that I don't want to use a separate headphone amp because convenience is a factor -- I want to move this from room to room on a regular basis.

Dave D
2005-07-15, 17:06
If looks like the headphone amp chip itself (not surrounding traces) is susceptible to picking up the 802.11 signal (it does not appear to be DC power related).

So that's a 2.4GHz signal? If it was an RF circuit, I'd say maybe one of the traces is acting as a nice 2.4GHz antenna. (1/10 lambda would be about 1.25 cm.) However, for this audio amp, it seems you'd have to get that noise down into the audio range to hear it. (I'm not an RF guy; just speculating.) Maybe the noise is actually from modulation of the 802.11b/g signal?

Here's a thought: If the pop/buzz with each clock tick (once per second) is reproducable, you could trigger a (digital) scope on that and (maybe) get an idea of the frequency of the noise. Maybe the antenna is a secondary effect, e.g. it's acting as a receiver for a lower frequency signal (maybe noise from accessing a memory, for example, or the modulation mentioned above) and re-transmitting. Still, you need to be able to pick it up again in the audio amp section...

I also wonder if the Standby/Shutdown pin would be a convenient place to inject some noise.

Well, it looks like an interesting problem to debug.

Dang, I gotta stop reading threads like this; it's very hard to resist opening the box ;) (I've been going by the photos.)

- Dave

seanadams
2005-07-15, 17:21
What you're hearing is the "envelope" of the signal. Of course 2.4G is a ways outside our hearing range. :)

Dave D
2005-07-15, 17:37
What you're hearing is the "envelope" of the signal. Of course 2.4G is a ways outside our hearing range. :)

I bet some folks in the Audiophiles forum would disagree ;)

(now I'm in trouble)

Dave D
2005-07-15, 18:58
Anyway yes, antenna orientation definitely makes a difference - by pointing the external antenna either straight out or towards the power input plug, I can make it go away. YMMV though - the radio will switch dynamically between the external and the internal antenna as needed.

I just changed back to wireless mode to try this out and I also heard the buzz/clicking under certain conditions. My network is 802.11b, by the way.

A few more interesting pieces of info, which may help you debug:

1) I used the volume controls to recreate the buzz/clicking (nothing playing). While I was turning the volume up and down, I noticed that the volume of the buzz/clicking does not change. That's probably not all that useful, now that I think about it; volume control is probably not at the headphone amp anyway, since it affects the speaker outputs as well.

2) The antenna orientation (or squeezebox orientation--not the same thing) both affect the level of noise (but read (3) and (4)!). Noise level is not always equal in each ear. Sometimes it goes from the one ear over to the other. Sometimes louder in one ear, etc.

3) I can *remove* the external antenna and still hear the noise. I discovered the next part accidentally. The SB2 had gone over to the RSS screensaver. (When the screensaver is scrolling a fairly constant level buzzing can be heard, which is useful for debugging.) When I started to put the SB2 back up on top of my (metal) CD player, I heard a fairly loud buzzing (left ear only). [Read on; it's not the CD player.] I started playing with the power cord and found that I could cause the noise level to increase when a section of the power cord was brought close to the opposite (right) side of the SB2. But I don't think it's coming from the power cord, because I can put the headphone cord and power cord close together, but away from the SB2 body and hear no noise.

4) Another accident: with the external antenna removed, I turned the SB2 so that the screen was facing me and the left side was toward the floor (top of the SB2 facing to my left). The noise increased. Then I saw that my headphone cord was crossing straight underneath the SB2 (the bottom side of the SB2 is to my right). I then raised the headphone cord so that it was no longer near the left side of the SB2. The noise decreased. The loudest buzzing occurs when I cross the headphone cord almost across the center of the bottom of the SB2, but slightly diagonally toward the front (any orientation of the SB2). Oh, and it flips from left ear to right ear and I have no idea what causes that.

My stereo equipment was all powered off for these test, by the way. Only the SB2 was powered.

Forgot to say...when I reverted to wired mode, the noise went away completely (as implied by earlier posts, but not directly stated, I don't think.)

- Dave

ceejay
2005-07-15, 23:24
I was intrigued by this thread because I have been using headphones but had not noticed the problem, hence investigated. Hope this info may add to the pool...

I have two wireless SB2s. One on a B network, one on a G network. Both exhibit the clicking/interference problem described by ephemere. The level is low enough that you wouldn't hear it while the music was playing but I can imagine would be very annoying during the gaps / times when you're selecting the next piece etc. Or if you listen to very quiet stuff!

The reason I hadn't noticed it myself is that I don't normally use unpowered headphones - I just plugged some in for this test. Usually I use Sennheiser wireless phones (aside - they're great, BTW!! - and being able to wander round the house and garden with iPAQ remote in my pocket and headphones on is just great). For whatever reason (input impedance???), they don't seem to collect this interference at all - although they do pick up an amount of hum and other rubbish that has nothing to do with the SB2.

ceejay.

ephemere
2005-07-16, 09:52
The RF interference I am hearing is quite easily audible when music is playing. Maybe rock would drown it out, but try listening to a Schubert piano sonata. I'm sorry to report that I've never heard such a noisy audio output, even from the headphone jack in cheap old CD-ROM drives. This is why I sent the first one back, certain that it must have been defective.

I'd still like a specification of the technical differences between the line out and the headphone out (level, impedance, etc.). I'm wondering if it would be appropriate to drive a headphone from the line outs rather than the headphone jack. I'd prefer to find some solution, because I'd like to be able to listen to my music. I'd rather not use a separate headphone amp because I'll be moving this from room to room and don't want an extra box and outlet to move around.

seanadams
2005-07-16, 10:29
The line outs are meant to drive low power into a relatively high impedance, heaphone amp is vice versa (50mw into 16 ohm IIRC). You'll certainly get sound if you connect headphones direct to the line outs, but it won't be loud.

ephemere
2005-07-18, 10:16
More information....

I tried this again last night, and there is audible interference even if the external antenna is removed. It's at a much lower level, but still audible during quiet passages.

Also, with the external antenna installed, the interference gets louder as the signal gets weaker. It's already audible at 100% signal strength, but at 30% it's audible even during the loudest passages of music.

seanadams
2005-07-18, 12:17
I have a solution for this now (thanks to Andrew Weekes for the pointers).

It looks like the signal is actually being picked up by the headphone cable (not internally). It goes into the 1/8" jack and then gets amplified back out by the headphone amp. The behavior seems different depending on the type of phones - Sennheiser HD 600 phones only exhibit the problem if the cable is actually wrapped around the antenna, whereas Sony MDR-7506 picks up the signal easily. YMMV.

Anyway, the fix is pretty simple but it does require soldering: attach two 33pF caps to ground on the headphone jack - make sure to keep the leads as short as possible - surface mount caps are best. This eliminates the problem for my test case, irrespective of antenna arrangement etc.

If you're comfortable doing this yourself, I've posted some step-by-step instructions here:

http://www.seanadams.com/sb2_headphone/

Otherwise please contact support@slimdevices.com if you want to have the repair done by us.

sbjaerum
2005-07-18, 12:25
I have a solution for this now (thanks to Andrew Weekes for the pointers).

It looks like the signal is actually being picked up by the headphone cable (not internally). It goes into the 1/8" jack and then gets amplified back out by the headphone amp. The behavior seems different depending on the type of phones - Sennheiser HD 600 phones only exhibit the problem if the cable is actually wrapped around the antenna, whereas Sony MDR-7506 picks up the signal easily. YMMV.

Anyway, the fix is pretty simple but it does require soldering: attach two 33pF caps to ground on the headphone jack - make sure to keep the leads as short as possible - surface mount caps are best. This eliminates the problem for my test case, irrespective of antenna arrangement etc.

If you're comfortable doing this yourself, I've posted some step-by-step instructions here:

http://www.seanadams.com/sb2_headphone/

Otherwise please contact support@slimdevices.com if you want to have the repair done by us.

Will this fix be incorporated in units you ship in the future?

Steinar

takashi37
2005-07-18, 12:25
How much will it cost to have this done? And is it same to assume it won't invalidate the warranty?

Thanks.
TL


I have a solution for this now (thanks to Andrew Weekes for the pointers).

It looks like the signal is actually being picked up by the headphone cable (not internally). It goes into the 1/8" jack and then gets amplified back out by the headphone amp. The behavior seems different depending on the type of phones - Sennheiser HD 600 phones only exhibit the problem if the cable is actually wrapped around the antenna, whereas Sony MDR-7506 picks up the signal easily. YMMV.

Anyway, the fix is pretty simple but it does require soldering: attach two 33pF caps to ground on the headphone jack - make sure to keep the leads as short as possible - surface mount caps are best. This eliminates the problem for my test case, irrespective of antenna arrangement etc.

If you're comfortable doing this yourself, I've posted some step-by-step instructions here:

http://www.seanadams.com/sb2_headphone/

Otherwise please contact support@slimdevices.com if you want to have the repair done by us.

seanadams
2005-07-18, 12:37
Will this fix be incorporated in units you ship in the future?

Steinar

yes, soon. I'm going to test this a bit more and make sure it's really the ideal fix - but it seems to work perfectly for me. I just wanted to get something out quickly for the guys who are hearing this.

seanadams
2005-07-18, 12:44
How much will it cost to have this done? And is it same to assume it won't invalidate the warranty?

Thanks.
TL

We'll do it for you no charge under warranty (you pay shipping to us, we cover shipping back to you).

If you want to do it yourself the parts are about $0.02 each (we can mail them to you if you want). We'll honor the warranty in this case, but I should stress that we don't expect everyone to go taking apart their SB2s themselves - only if you're comfortable with electronics. That said, it's a pretty easy tweak.

pfarrell
2005-07-18, 12:57
On Mon, 2005-07-18 at 12:44 -0700, seanadams wrote:
> If you want to do it yourself the parts are about $0.02 each (we can
> mail them to you if you want). We'll honor the warranty in this case,

Just another example of why I love the folks at Slim Devices.

Do these extra parts and the labor to install them mean that the latest
price cuts will be rolled back?
:-)

--
Pat
http://www.pfarrell.com/music/slimserver/slimsoftware.html

Andrew L. Weekes
2005-07-18, 14:15
Just another example of why I love the folks at Slim Devices.

Agreed! I've said it before, but I'll say it again, the guys at Slim Devices are refreshingly open, it's great stuff.

For a techy aside to the problem, it's very common for electronics to demodulate RF signals. You cannot hear the RF (even as an 'audiophile' I wouldn't argue that!) but what you are hearing is the effect of the electronics turning into an RF detector (i.e. radio receiver).

It only really happens in junction devices, so bipolar transistors are worse in this respect than FET's, which have no actual rectifying (i.e. diode) junctions. For that reason FET-input amps are much better in RF environments than bipolar parts. Heavily degenerated bipolar parts perform better (in this respect) too.

With AM transmissions you can actually demodulate the audio content (or data content) quite readily, which is why AM CB radios, or amateur radio / PMR transmissions are more prone to obvious audio breakthrough. Because the method for decoding FM transmissions is different, you often just get the effect of the carrier affecting the item, since the data or audio content is harder to demodulate.

11g WiFi uses OFDM modulation, a complex spread-spectrum technique, to get more data through a given bandwidth link, which probably looks like a bit of a mix between AM and FM to the device picking it up, in reality.

Low level (low current) stages and high impedance areas of a circuit are most prone to this problem (e.g. phono stages, high impedance op-amp inputs etc.), and the problen can manifest itself as direct effects, or the effect of the stage being driven into non-linearity by the RF.

What I suspect, based on experience (but haven't analysed in detail yet) was happening here is the headphone amp cable was acting as a reasonable antenna and the RF was getting back into the headphone amp chip, which then gives it a direct path back to the input stage (via the chip's feedback loop).

Good to see it fixed though and I'm happy to be of assistance!

Andy.

References for those that fancy a bit of light reading: -

http://www.analog.com/UploadedFiles/Application_Notes/5731282210277AN671_0.pdf#xml=http://search.analog.com/search/pdfPainter.aspx?url=http://www.analog.com/UploadedFiles/Application_Notes/5731282210277AN671_0.pdf&fterm=input&fterm=stage&fterm=rfi&fterm=rectification&fterm=sensitivity&fterm=input stage rfi rectification sensitivity&la=en

Page 10.39 on of this one...

http://www.analog.com/UploadedFiles/Associated_Docs/5619348511532174844452507435855319349073202222160F FSect10.pdf#xml=http://search.analog.com/search/pdfPainter.aspx?url=http://www.analog.com/UploadedFiles/Associated_Docs/5619348511532174844452507435855319349073202222160F FSect10.pdf&fterm=input&fterm=stage&fterm=rfi&fterm=rectification&fterm=sensitivity&fterm=input stage rfi rectification sensitivity&la=en

ephemere
2005-07-18, 15:02
As the original poster, I would like to thank Sean and everyone at Slim Devices for their willingness to address this problem, not to mention their blazingly responsive turnaround in finding and committing to implement a solution. This attention to detail is well above and beyond the strict call of duty.

When I bought my first SB1 a while back, one of the reasons I picked it over the competing Roku unit was that I had a much better feeling about Slim Devices as a company. This experience validates that decision in no uncertain terms, and speaks very highly of Slim Devices as a company.

takashi37
2005-07-18, 21:14
We'll do it for you no charge under warranty (you pay shipping to us, we cover shipping back to you).

If you want to do it yourself the parts are about $0.02 each (we can mail them to you if you want). We'll honor the warranty in this case, but I should stress that we don't expect everyone to go taking apart their SB2s themselves - only if you're comfortable with electronics. That said, it's a pretty easy tweak.

So I'm not sure I'll even be taking you guys up on the offer but had to reply to commend your top-notch customer service.

It is remarkable, in the literal sense of the word. I remark about your customer service to people practically every day.

Constantly amazed by you guys,
TL

seanadams
2005-07-18, 21:43
Thanks guys for your kind words! But the real kudos go to Andrew for actually offering some direction. I was getting nowhere looking at the internal signals, and quite overlooked the idea that the 802.11 output was getting coupled back into the system via the headphone cable.

Dave D
2005-07-24, 13:03
I just put in this fix this afternoon. I used 33pF, 0805-size caps (forgot to check the dielectric type; they may have been X7R. I can check and get back to you if you care about that). Anyway, it does dramatically reduce the noise, but in my case, does not eliminate it. With the headphone cord in the position I described in my 7/15 post, i.e. across the bottom center of the SB2, there is still audible buzzing or clicking in wireless mode.

Just curious: how did you determine the 33pF value?

By the way, I really don't care about the little bit of noise; I don't expect I'll use the headphones much at all, and when I do, I'll be careful where the headphone cord goes. This is just a data point for you and others.


Anyway, the fix is pretty simple but it does require soldering: attach two 33pF caps to ground on the headphone jack - make sure to keep the leads as short as possible - surface mount caps are best. This eliminates the problem for my test case, irrespective of antenna arrangement etc.

ephemere
2005-08-10, 13:03
yes, soon. I'm going to test this a bit more and make sure it's really the ideal fix - but it seems to work perfectly for me. I just wanted to get something out quickly for the guys who are hearing this.

Sean, I'm wondering if you could provide an update on this. Did the fix work out to your satisfaction? If so, when will it be incorporated into shipping units?

seanadams
2005-08-10, 18:25
Sean, I'm wondering if you could provide an update on this. Did the fix work out to your satisfaction? If so, when will it be incorporated into shipping units?

We're already shipping this fix as of a couple weeks ago. If you want to send your unit in please contact support@slimdevices.com