speaker hiss

[jhsa]

you can port them yourself Just change the file names/numbers according to the ersky9x file list that is included with the ersky9x voice pack

@ Steven, I would go more with option/situation "b", noisy regulator. I don't think it is the speaker. I don't bet this time cos you always win

João

[andrewju]

a) I didn't use a high enough quality speaker;

Steven, may I add a comment?

I do not fully agree with your statement. I mean, I don't understand how a bad speaker could cause the "hiss" noise. A speaker converts an electric signal into sound. A bad quality speaker may do the conversion wrongly, which will result in bad sound quality, or even no sound at all. But it cannot "invent" a sound in the absence of the signal.

If a speaker makes a noise, it means something (some kind of an electric signal) is coming to it. By going for a higher quality speaker, you will likely just end up with a higher quality hiss...


Just thinking out loud...

[s_mack]

@ Steven, I would go more with option/situation "b", noisy regulator. I don't think it is the speaker. I don't bet this time cos you always win

I know, right? That's why I called it a "career path" and not "gambling addition" like the judge said.

Anyway. I'd be inclined to agree that the switching regulator (which are inherently "noisy") would be the likely culprit... except for one small detail: variance. Some people hear nothing (my own non-ancient, non-damaged ears, for example), others a lot (as evidenced in that video). The circuit boards are pretty precise. They should all be noisy or all not noisy if it is something on the board itself. That's my logic anyway. The speakers made more sense in that light, since they are a wound coil and hand soldered and perhaps not as tight of a tolerance.

What kind of made the most sense to me was that it had to be some other factor. For example, environmental influence, choice of wiring location, etc. But then people said it didn't matter where they were or how they moved the wires so that half went out the window.

Mine is *silent* except when playing a sound. Along with the voice, I hear some static in the "background" (and since these are digitally produced files from end-to-end, there really shouldn't be any). I've had my friends (both of them!) see if they could hear something I couldn't, and they say it is silent too.

Why the variance??? That's what's bugging me more than anything.

[s_mack]

a) I didn't use a high enough quality speaker;

Steven, may I add a comment?

I do not fully agree with your statement. I mean, I don't understand how a bad speaker could cause the "hiss" noise. A speaker converts an electric signal into sound. A bad quality speaker may do the conversion wrongly, which will result in bad sound quality, or even no sound at all. But it cannot "invent" a sound in the absence of the signal.

If a speaker makes a noise, it means something (some kind of an electric signal) is coming to it. By going for a higher quality speaker, you will likely just end up with a higher quality hiss...


Just thinking out loud...

As I've said before, I know very little about audio. But consider this: I had a c**p pair of PC speakers plugged into my computer and they would constantly make noise. I finally bought a more expensive set and now there's nothing. The computer stayed the same, the speakers changed. Both sets were unpowered basic speakers (not the ones with fancy surround sound and woofers). So... if not the speaker, then the cable? The soldered joint? I really have no idea. Two things are preventing me from diagnosing it: #1) mine doesn't hiss; #2) my [lack of] knowledge on the subject. That's why I issued the "bounty" and also why I'm not shying away from this. I completely believe that people are bothered by it and I want it resolved. I just don't know how yet.

[gb21914]

Please note that I took the video in a very quiet room. I'd like to add a couple of notes based on my poking around.

1. I don't believe it to be the LED panel. In my poking around I took the panel out and fired the unit up and it still crackled.

2. The speaker appears to make the noise regardless of where it is located or mounted.

3. The method of powering the board doesn't seem to have an impact (lipo or battery).

4. The sd card doesn't seem to be an issue as I can play the files just fine directly off the unit.

5. I have reseated my board about 4 times now...the only thing that I can think, is that a couple of the pins on my board were bent inwards a little bit. I did mention this to "Mr. M" when the board arrived, but as it fired up / is working...I haven't pursued that any further.

Perhaps any others that have "the hiss" should double-check their boards and confirm that the pins are straight up and down, or whether they are "leaning" at all?

[s_mack]

confirm that the pins are straight up and down, or whether they are "leaning" at all?

They will always be leaning to some degree. It is not possible to get them perfectly straight, nor do they need to be. They really don't even have to be all that close to do their job. I can't imagine they are causing this problem, but I guess I can't say I know for sure until I know for sure

[gb21914]

confirm that the pins are straight up and down, or whether they are "leaning" at all?

They will always be leaning to some degree. It is not possible to get them perfectly straight, nor do they need to be. They really don't even have to be all that close to do their job. I can't imagine they are causing this problem, but I guess I can't say I know for sure until I know for sure

I don't disagree with you. I'm just trying to come up with some logic that may determine the source of the noise. As I've said before...I'm pretty sure I can live with what I have. It's just "hissy"

[jhsa]

I also don't think that the pins cause the problem..
Steven, you are right about some having the noise and some don't, the same goes to all the other points you mentioned. All boards are equal. still worth it to scope the 5V rail. is it possible to turn the 9x mainboard off while still running the 9xtreme? It looks like the 9x itself is one of the factors that isn't common to all radios. So, could some 9x radios induce the noise while others don't??

João

[andrewju]

But consider this: I had a c**p pair of PC speakers plugged into my computer and they would constantly make noise. I finally bought a more expensive set and now there's nothing.

I'd say in this case it could have been some interference (electromagnetic radiation) that was caught by the old speakers - probably due to poor cable quality.
Another (though, much less likely) reason could be that your new speakers are simply unable to produce the sound in the "noise" spectrum coming from your computer. And that's why you don't hear it on the new speakers.


In case of 9Xtreme, having a closer look at the speaker signal may really help. Especially on a board with the higher noise level.
I hope Mike will find an answer when he gets time to look at it with his 'scope.

[s_mack]

is it possible to turn the 9x mainboard off while still running the 9xtreme? It looks like the 9x itself is one of the factors that isn't common to all radios. So, could some 9x radios induce the noise while others don't??

Man, sometimes you make the plainly obvious easier to see

I hadn't thought about the 9X itself. Of course that could be an influence. Good call. I've got two 9X's here, one being half trashed... but it at least gives me the ability to test in one then the other. And one is the "new" style and one the old, so that could be a clue.

Mike gave me a pre-production testing firmware that may be useful here. It would play a sound from the SD card but didn't require the board to be in the 9X at all. Time to go dig through my files and crack open a pair of 9x's!

[jhsa]

Just don't destroy your board and then blame it on me.. Unless you know the person who makes them, of course..

Looking forward to the results..

João

[MikeB]

Looking at the speaker signal itself probably won't help as we have a class 'D' amplifier that outputs PWM, which may, of course, be part of the problem. Maybe the PWM frequency is slightly different and causes the problem.

Mike.

[jhsa]

And what would be the fix?? Software?
João

[andrewju]

Looking at the speaker signal itself probably won't help as we have a class 'D' amplifier that outputs PWM, which may, of course, be part of the problem. Maybe the PWM frequency is slightly different and causes the problem.

Ahhh, right... Then locating the root cause is not that easy.
But it should be manageable, right?

It's really interesting to know what the issue is!

[PacketLoss]

The backlight is LED, which runs on DC. So no inverter.

It most likely comes down to either a) I didn't use a high enough quality speaker; b) the switching regulators are producing too much noise; c) I didn't get the ground path right (since i have zero experience with audio, I wouldn't be surprised); or d) we're using single-sided input instead of differential, which is recommended in the datasheet to reduce noise. Or some combination of the above.

The variance in how much it bothers me suggests it could be 'a'. Or it just comes down to peoples' ears and their differences. That video... mine doesn't sound like that at all, but is that a difference between his and mine? Or is the camera picking up the audio differently than if I heard it in person?

This is tricky.

I'm tempted to buy a Taranis, nothing to do with this but all to do with an irrational desire to try SBUS (and I'm maxing out 8 channels) and no point buying a replacement for my DJT given the price delta.

If it's any use, I'd be happy to send my 9x down to Mike to help isolate (even if it's just the the human element)?

(and if you say yes it somehow rationalizes the expense and I'm sure I'll be able to explain the expense better to the mrs)

[MikeB]

Finally managed to put the 'scope on things. I can't see any obvious noise source, although looking at the 5V supply I can see something repetitive at 500Hz.
I wonder if those with noise on the speaker have a problem with the 5V regulator. This is a switch mode regulator, so could have a problem.
Near the middle of the board is relatively large black component (the inductor) labelled "101".
Around this are some capacitors. If one, or more of these is not soldered properly, then the output of the 5V regulator will not be very good, and could be the cause of the noise.

A close inspection of the soldering of the components (or a good, high resolution image) around this inductor is therefore worth doing.

Mike.

[bob195558]

[gb21914]

Here is mine...I can get much closer if you really need me too

[duststorm]

I noticed that the hissing changed if I changed the external transmitter PPM from 300usec to some other value. On some settings the hissing even stopped completely.
Perhaps this can help to find the source?

[s_mack]

I noticed that the hissing changed if I changed the external transmitter PPM from 300usec to some other value. On some settings the hissing even stopped completely.
Perhaps this can help to find the source?

Interesting. Could you share the settings and and procedure so others with this hissing annoyance can test and see?

[duststorm]

Ok, trying it again these are my observations:
The sound I get from the speaker is a quiet high pitched wail from the speaker.
It appears directly related to the settings of the external transmitter. Turning the external port off stops the sound immediately.
Internal radio settings have no effect on the speaker whatsoever.

In PPM mode, the FrLength parameter has a very noticeable impact on the speaker. I have it at 22.5usec by default, lowering it towards 12.5usec makes the pitch higher, moving it towards 32.5usec makes the pitch a little lower, and includes inter-silences so the high pitched sounds are more spread out. I believe that the hiss I hear is actually a pulsation between sound and silence, and making it pulsate slower lets you clearly hear the silences in between.

Setting the protocol to PXX reduces the sound to almost nothing, Assan protocol completely eliminates the noise. DSM2 and Multi create a totally different sound, a low hum that pulsates between silence and hum at a slower frequency.

The sound appears to be generated entirely by my 9xtreme board, and not caused by the transmitter module. I can power my DJT module with the internal PPM on the expansion pads (something I had to do to circumvent an issue I currently have with the board), and it works without causing a buzz on my speaker.

But while writing this I am starting to doubt whether my findings are going to be useful... The issue I have with my (partially defective) 9xtreme board is that the external ppm signal is connected to ground, but this could be what is causing the hum on the speaker, as it is varying the ground.
So perhaps you can ignore my findings.

[s_mack]

Weird. At the beginning of this discussion on page one I thought we already ruled out the rf being the problem??

and the Drawing Board welcomes Steve back. He always knew he'd come crawling back.

[duststorm]

I just heavily edited my post, make sure you re-read it before I cause too much confusion

[gb21914]

I noticed that the hissing changed if I changed the external transmitter PPM from 300usec to some other value. On some settings the hissing even stopped completely.
Perhaps this can help to find the source?

We may be looking at two issues then.

I played around with my transmitter last evening, and went through all the settings (including turning both internal and external sources off).

I've even removed the DJT module completely...and the noise persists.

It is constant...I don't seem to be able to make it better or worse...so I'm still partially leaning towards either an interference issue, speaker issue, or the board itself somehow. I'm going to try and source another speaker and see if that makes a difference.

[ShowMaster]

The path from the CPU to the audio ic input does take a path from the top and bottom of the board. That may allow coupling of data onto the audio traces. All guesses that will be looked at as more tech types install their boards.
I honestly showed off my 9xtreme today at a large fun fly. After they played with it, I asked them to comment on any hiss, buzz, they heard when I triggered audio or not. They said they only head the announcement and it was ok! My point is, try it at the place you fly. You may not hear anything over the field noise, I don't. This is being looked into but again honestly, my Taranis buzzes also. I just ignore it in quiet places.
The biggest issue on my list it how does it fly my plane. So far it's great!

[SkyNorth]

Hi, here are some thoughts on the noise problem.
I found some good info from some TI data sheets ...They are much better written than the ISSI one which is a obvious copy of this chip from TI.
http://www.ti.com/lit/ds/symlink/tpa2005d1.pdf

Another note I found:
In applications which use a DAC to drive Class-D amplifiers, out-of-band noise energy present at the DAC's image frequencies fold back into the audio-band at the output of the Class-D amplifier. An external low-pass filter is often placed between the DAC and the Class-D amplifier in order to attenuate this noise.

I used something like this on the Sky9x design

The differential input on the amp is very sensitive to in-balance. The input resistors need to be 1% or better. With the source input resistor split into two resistors , a greater risk of part tolerance coming
into play . The input capacitors also effect the input resistance ..so they need to be 10% tolerance or better.
The input caps should be around 3.3nF to cause a Low frequency cutoff off around 320Hz or so..this would help with supply noise rejection.

Once the speaker wires go past 10cm in length then output filters are sometimes needed ..They are only filter less when they have a very short direct connection..

Because the speaker is fed DC square waves , the voice coil does have to dissipate a bit more power than if fed with sine waves...

Also , to scope a Class D output you need a low pass filter on the probes...such as

I hope some of this helps.....

-Brent

[s_mack]

Thank you Brent.

When I chose to use a Class D, I did so because it seemed to be better, particularly in power consumption. I thought I was providing a higher standard... raising the bar, if you will. But I now realize the tradeoff in sensitivity to noise. The wording of the ISSI datasheet gave me comfort that it was good at noise rejection, but I suppose that's relative. I still don't understand why the variance though. I did *specify* 1% resistors for each in the audio circuit. Were 1% supplied? I would hope so, considering the cost difference is so little. The caps, however, are 20% not 10%.

This one's for Mike... can we fix this? Remember how we didn't use both DACs? We pulled PA5 (the other DAC) out as a "spare i/o" so we have a nice big solder pad there. Is it feasible to make an external audio circuit that would then need this relatively simple soldered connection and the end-user could rid themselves of the hiss? Of course, it would mean firmware support because the other DAC would have to be used instead of the current one. Not ideal.

The other way is to "cut-out" the existing circuit. While possible, it is going to be beyond the skills of most. The best way would probably be to remove R32 and solder to the exposed pad, but man that's a tiny pad. I've re-uploaded a graphic showing the location of the PA5 (spare i/o) pad. You can see R32 (unmarked) south of PA5. Removing that, the right pad would be a direct connection to the DAC.

[ShowMaster]

Good to see Brent's getting involved. He did focus on noise reduction on his Skyboard design.
Being that my install audio is to my liking, there's another 9Xtreme board heading my way. I'll install it and listen to its audio in case I got lucky the first time.
I do own a SMD microscope and from the part sizes, I'll need it!
I'm willing to try any filter or precision resistor mods as soon as I have a noise floor level to improve. I'll use the board that is higher if there is one. A few new posted noise videos or audio files would help. Please indicate the device used to record it and maybe a 12" distance from the speaker?
There should be a established realistic distance to relate to. I'm not sure it'll ever be 100%?
The next test will be with high powered rf add on's. RF and audio doesn't mix well as we ham radio operators and our neighbors can verify.

[MikeB]

I do have 'other' plans for using PA5.
I'm just trying to get a feel for how big this problem is, while still trying to fix it. I feel that of the many 9Xtremes installed, only a handful are reporting significant noise on the speaker.
On my own 9Xtreme, there is some very low level noise, but I can only hear it if my ear is within about 6 inches (15cm) of the speaker. I didn't hear much on the prototype either.
It may be, as Brent says, the noise is caused by 'beating'/'aliasing' with the ampilfier pwm output frequency, and the amplitude of audible noise depends on the actual pwm frequency, which could vary between chips. Just a few cause the audible noise to be too loud.

Mike.

[MikeB]

Anyone with load speaker hiss/noise, please start the radio in bootloader mode (both horizontal trims held inwards, then power on), then let us know if you still have a lot of hiss/noise. The voice output, and much else, is not enabled in this mode.

Mike.