speaker hiss

[s_mack]

I'm glad it was brought up. Nobody's expecting perfection. Im personally pretty elated that the biggest problems with this project is an incorrect screw size and a humming speaker that wasn't even expected to make it in the first batch.

[PacketLoss]

I'm glad it was brought up. Nobody's expecting perfection. Im personally pretty elated that the biggest problems with this project is an incorrect screw size and a humming speaker that wasn't even expected to make it in the first batch.

Here here, you should be taking a well deserved bow right now

[s_mack]

But then my enlarged head would hit the keyboard.

[PacketLoss]

But then my enlarged head would hit the keyboard.

LMAO and a sense a humor, stop showing off :p

Just ordered a couple of different speakers for testing, worth me updating the original post with suggestions/things-tried/bounty?

[don9f]

I was just playing around with my 9x and have a couple of observations that may assist diagnosis by those more capable than me :-

1) The "noise" seems to be independent of volume control (setup on P2).

2) It doesn't really become apparent until after all alarms have been cleared after switch on.

I suffer Tinitus anyway, so have ringing in my ears all the time

Cheers Don

[ShowMaster]

I know many that have seen combat or 50+ 70 group can't hear, and now add the 20-40's thanks to rock concerts and iPods. So that means a lot of 14 year olds or women are taking up RC.
I hope these women come to fly at my field, all we get are the grey haired men. Wait, looking in the mirror I just realized why I don't hear the hiss much! Eureka, a break through!

[ShowMaster]

I swapped my throttle and trainer switches. While I had my case open I powered it up and listened with my good ear, the one I use to hear " dinners ready" not the one I use that doesn't hear "take out the trash"!
In any case, I of course heard what we call hiss. The audio ic is just below the speaker connector, wow are those tiny ic's.
The numbers on the ic are p5204 and then 2005. I couldn't find it on google so I'm sure I'm missing a #?
In any case using the old trick of laying a finger on it, I was able to force a louder buzz condition along with other sounds. It and the the resistors below it had the same effect. The original hum/buzz didn't decrease.
I did scope the speaker connector referenced to ground (-) and got a square wave on either speaker lead. I suspect this is some form of a balanced output not referenced to ground. Making it speak, the modulation was on the sides of the the square wave type display, not amplitude.
Until I know more about the speaker ic and circuit, I'm holding off. I did probe the ic leads but was afraid of a slip!
I did scope the power supply busses showing up on the pads. Looked clean this first go around.
My thinking is maybe a gain structure being too hot?
I'm flying blind but it's a start. I don't want to blow up my board this early chasing something I'm not really bothered by. From a technical view point, it would be great to track it down.
Feel free to join in on the search or that IC data sheet?
SM

[s_mack]

This is the IC: http://www.issi.com/WW/pdf/31AP2005.pdf

[ShowMaster]

I'm assuming it's input is being used single ended? If I get the time and courage, I'll short out the input pin to ground after the cap, before the resistor. That should establish if it's inherent ic noise or earlier. I see it has a pin for shutdown. Maybe it could be turned off in fw when at idle?
And yes, the speaker output's not referenced to ground as I suspected, so my easier measurements I'll ignore.

[jhsa]

That is a class D amp. The output is PWM I think.

Steven, did you ground the unsused input pin?
Also what is the power rating from the speaker you're supplying? Nothing under 2W I hope
That thing is rated at 1.70W at 5V?

João

[s_mack]

What it is rated at and what it is outputting are two different things. I think. Again, you're speaking to an audio idiot.

I believe I asked around for what people were using and 0.5W seemed to be the norm. I got the speaker one of you Voice folks told me to get. It is a 0.5W. I considered getting 2W for the very reason I think you're suggesting, but was told it wasn't necessary. Someone said it would be too quiet. I just shut up and did what I was told.

By all means, test with a 2W so I know for next time!


Remember folks, the speaker wasn't even going to make it into this batch. It was not tested in the slightest.

[s_mack]

Here's how the amp is connected.

[andrewju]

Is IN+ connected to GND?!

[s_mack]

By way of r/c, yes.

[andrewju]

Oh, right. We have a single input here. Ok.

[andrewju]

Hmm...
I'm not saying a single-ended scheme is the root cause of the issue discussed in this topic, but here's a note from the datasheet:

The fully differential IS31AP2005 can still be used with a single-ended input; however, the IS31AP2005 should be used with differential inputs when in a noisy environment, like a wireless handset, to ensure maximum noise rejection.

I understand the differential input mode should be more "interference-proof".

[ShowMaster]

I think that statement is saying that any noise coming in through the balanced input will be cancelled out if on both inputs. I'm thinking since this is used single ended, shorting the input should stop anything on the input. If the noise/hiss is still there, it's the IC amps gain structure or design. Once isolated to a stage or circuit, more options can be thought about.

[jhsa]

By all means, test with a 2W so I know for next time!


Remember folks, the speaker wasn't even going to make it into this batch. It was not tested in the slightest.

Steven I'm not complaining. But it's just that if you force 1W into a 0.5W coil something could burn in this case the coil itself. If people never use full volume all the time that might never happen.
It already happened to me twice with some 0.5W speakers I had installed on my 9x radios. And the power output of the LM386 is much less than the 1.7W from the chip you use on the 9Xtreme.
This is just my 2c

João

[MikeB]

I've just done a quick experiment where I change the DAC output while listening to the 'noise' I can hear on my speaker. I tried "normal", disabling the DAC output (floating) and driving the DAC output to 0V.
The 'noise' got louder when I had the output floating, but it did get quite a bit less loud when I made it 0V.

This suggests to me that there is some signal pickup on the board. Whether this is all the hiss is, is another question.

Mike.

[jhsa]

I need to get my board so i can hear it.. ; (

João

[ShowMaster]

I too wish you would get his board Jsha . You and Mike have been scratch building so many Rc systems since I've know you!
I actually think they are responsibly for all the talking DIY systems we use everyone.

Mike, are you able to power up the board enough when not installed, to listen to it? I too think it's a coupling issue and not a simple power buss noise. I'm limited to what I want to try. If a trash the board, I can't test the many things it has to offer.

I know Brent worked very hard on the Skyboard design and the Pro to keep out rf and data from his board designs. I'll have to listen to the ones I've never been bothered when flying, so I'll have to try a side by side test to listen to them just for a reference. The Taranis we know came with a buzz that took 3-4 generations of upgrades to reduce.

Hopefully it'll be a data trace running too close to an audio path. That or cut a audio trace and insert a micro toroid type fix? I'm thinking a simple cap fix isn't going to happen, but let's be positive it will.

Eraky9x seems stable as is the board so far in my flying tests. The haptic I installed is also working as expected.
I put 2 pancake motors in parallel and used the 5v buss. They're rated 80ma ea @4.5v , 160 ma. total.
I used a 2N7000 FET for the switch.
The premade Frsky haptic board should work fine for those not wanting to DIY build one. I don't hear and added audio buss noise when my motors run, but they do make their own noise of course.
SM

[jhsa]

I too wish you would get his board Jsha . You and Mike have been scratch building so many Rc systems since I've know you!
I actually think they are responsibly for all the talking DIY systems we use everyone.

Guess what, I have just received a certain package today
Mike is responsible for most (if not all) DIY open source talking systems. I have been building and testing them of course but that's it So, I take no credit for anything whatsoever..

I know Brent worked very hard on the Skyboard design and the Pro to keep out rf and data from his board designs. I'll have to listen to the ones I've never been bothered when flying, so I'll have to try a side by side test to listen to them just for a reference. The Taranis we know came with a buzz that took 3-4 generations of upgrades to reduce.

I have just rest my ear against the skyboard REV 1B speaker.. Could hear a very faint noise. started the radio in trainer mode (frsky module turned OFF) and silence..

My 9XR-PRO has a little humming, very faint also.. I will compare them to the 9Xtreme as soon as I install it. But that might have to wait a little while and I explain why..
The radio I will install this board is the one that I'm currently using to fly my micro quads using my DIY Multiprotocol module..
The problem is that this radio is the only radio I have with er9x and that can use the external module bay. The "Multi" protocol fully works in er9x but not in ersky9x as far as I know. At least the protocol I need doesn't. I have tried it. So, I'm stuck with not upgrading the radio until the protocol is fully supported in ersky9x. Then I could use the 9Xtreme or the PRO to fly the quads. My son loves them and likes to fly them together with me, so I can't do anything about it.. It must wait.

João

[s_mack]

Steven I'm not complaining. But it's just that if you force 1W into a 0.5W coil something could burn in this case the coil itself. If people never use full volume all the time that might never happen.
It already happened to me twice with some 0.5W speakers I had installed on my 9x radios. And the power output of the LM386 is much less than the 1.7W from the chip you use on the 9Xtreme.
This is just my 2c

João

No, I didn't think you were complaining. I had the same concern, but was impressed upon that while it *can* push 1.7W, that's within our control so it just doesn't. I don't think burning coils is a concern, but I do wonder if a higher RMS speaker reduces the noise?


ps. which package showed up? The DHL one wasn't expected until tomorrow I believe, so is it early or did the other one finally make it to you?

[jhsa]

I sent you an email as soon as it arrived it was the first one that arrived.. DHL tomorrow according to the tracking info..
I still didn't open it though..

João

[ShowMaster]

João, it was your persistent but caring nagging, that got this voice this started years ago[emoji3]
I have to give you credit for that. Without those may request posts, we would still be listening to beeps and bops today.
I'm glad your board arrived and you're on board now. Anyone near João have a spare 9x, working or not? He's good at fixing. I may have enough parts but the shipping may cost as much as an eBay 9x?
That or we all donate $1 to get him one to help solve and test with.
SM

[MikeB]

Multi protocol is enabled on ersky9x for the SKY board and the 'PRO. I think I need to just sort a couple of compile options for the 9Xtreme for it to be working there. Since I don't have and hardware multi protocol module, I can't test it.

Mike.

[jhsa]

Showmaster, thanks, but I will mod it soon.. It might just be a little bug in ersky9x's "Multi" protocol as er9x works. I'm sure Mike will fix it soon

I don't think someone can be given credits for "Nagging" ??

Anyway, thanks for your idea to get me another 9x but at the moment is not needed.
Here, as far as I know, there are very few using the 9x. There are some using the taranis.
I visited another club's field near where I live today, while going for a bike ride with my Son, and I only saw very expensive Futaba and Multiplex radios.. I think many buy the radios not for what they can do, but because they are expensive unfortunately..
Also, I believe most don't even know about the open source projects nor the 9x radio.. And the ones that know don't think or say very nice things about it. Actually they make fun of it.. That's what I get at my club's field. And that is what I call pure ignorance. But I don't care because I know I have a very good system. But sometimes I just use some nice features, like my automatic throttle control with altitude on my glider, or the trainer over bluetooth, and I can see the jealousy on their faces. But not to show it, they just start trying to find problems and making fun of it.. But I don't care. I have learned a lot from all this, and I really appreciate what people like Mike, Steven, Brent and many other do.. So, a big THANK YOU to all of them..

João

[jhsa]

Multi protocol is enabled on ersky9x for the SKY board and the 'PRO. I think I need to just sort a couple of compile options for the 9Xtreme for it to be working there. Since I don't have and hardware multi protocol module, I can't test it.

Mike.

I have just tried it again on both the pro and the 9x with the skyboard B1 as maybe the problem could be the pins not making contact.. But no.. It is really the protocol that is not working.. I inserted the module in my er9x radio, and the little quad bound immediately..

João

[gb21914]

I have installed the audio files and get a 'warble' when speaking as well as a hiss (when the voice portion is not in action). The attached video demonstrates the hiss pretty clearly I think.

https://www.youtube.com/watch?v=yIknaq6MFnI

[jhsa]

That sounds like it might be caused by the RF module. What module do you have there connected?

Please remove the module and turn the radio ON.. See if the noise is still there...

João