Fix for internal speaker howling

Post Reply
pe1nnz
 

Posts: 50
Joined: 30 Dec 2021, 19:29

Fix for internal speaker howling

Post by pe1nnz »

A speaker howling effect may appear when using the internal speaker and volume levels above 12. It is caused by electrical and mechanical feedback from the speaker to the very sensitive receiver circuitry.

The recommended solution is to use an external active speaker as proposed in this video https://m.youtube.com/watch?v=ijQKpjToHFw

Alternatively the howling sound on the internal speaker can be reduced by applying the following fix. Before considering to apply this fix, be aware that this modification is not easy at all to apply, as it requires fine soldering and PCB modification skills, and may lead to failure if not handled properly.

Fix:
1. Disconnect U9 from VCC by cutting trace of JP7. Reconnect U9 to 5V output of U2, e.g. by soldering a wire from middle contact of JP7 to top-side of L1.

2. Disconnect U6 from VCC by cutting trace of pin 16/U6 and C20. Reconnect U6 to 5V output of U2, e.g. by soldering a wire from side of pin 16/U6 to top-side of L1.

3. Disconnect anode D1 from VCC by desoldering it and pulling it upwards. Reconnect D1 to 5V output of U2, e.g. by soldering a wire from anode D1 to top-side of L1.

4. Disconnect R10 from VCC by removing it. Reconnect R10 to the pad connecting to C19 and to 5V output of U2, e.g. by soldering a wire from R10 to the top-side of L1.

After applying the fix it looks like this:
Internal speaker howling fix
Internal speaker howling fix
7CC723A2-E4A3-4767-B9B0-16A580E5F531.jpeg (171.87 KiB) Viewed 11292 times

Above fix improves isolation between micro-controller VCC and receiver VCC. This is important to break an electrical feedback loop, preventing tiny audio PWM current spikes leaking into the receiver section with 50+dB gain stages.

These changes already results in a significant reduction of the howling feedback, allowing the use of volume level 15. But there is still some microphony on the higher-bands (20m and higher), which can be heard by mechanical vibrations on the enclosure and sound feedback of the speaker.

To further reduce mechanical feedback effects between speaker and piezo-electric effects of the QSD capacitors (it the high-K dielectrics in MLCC (Multi-Layer-Ceramic-Capacitors) that is piezo-electric):

5. remove C25, C26, C27, C28 and replace them by 0.47uF through-hole capacitors (not shown in pic).


EDIT by DL2MAN: Alternatively to Guido´s Step 5 you can use Tantalum Caps in 0603. Very expensive, but a drop in fit.
https://www.mouser.de/ProductDetail/KYO ... t2xQ%3D%3D
See Pic:
0603_TAN.JPG
0603_TAN.JPG (61.15 KiB) Viewed 11241 times
EDIT2 by DL2MAN:
This Mod is only Valid for Mainboard PCB Revision 1.0.
KB8FJO
 

Posts: 7
Joined: 14 Nov 2022, 02:18

Re: Fix for internal speaker howling

Post by KB8FJO »

Dl2MAN’s edit suggests all you need to do is replace those components and not do the cut traces and wire up to 5V @ L1? Am I reading that correct?

Or is that just an alternative for step 5?
N6WLM
 

Posts: 20
Joined: 12 Jul 2022, 03:15

Re: Fix for internal speaker howling

Post by N6WLM »

That's great there is a fix! I might give this a try. Could you provide some close up pictures of cutting the trace for U6/19 and C20?

Do we need to cut only the red highlighted trace and leave the others intact?
traces.png
traces.png (42.53 KiB) Viewed 11008 times
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: Fix for internal speaker howling

Post by DL2MAN »

KB8FJO wrote: 08 Mar 2023, 02:14 Dl2MAN’s edit suggests all you need to do is replace those components and not do the cut traces and wire up to 5V @ L1? Am I reading that correct?

Or is that just an alternative for step 5?
My EDIT is just an alternative for optional Step 5. I still cut the traces. The hole mod looks like this in my case:
mod.JPG
mod.JPG (215.46 KiB) Viewed 10982 times
As you can see: way uglier than Guido´s work, but it works.

@N6WLM:
You´re right. The red trace is one of the 2 required cuts.
The other cut needs to be made at the voltage jumper pad (JP7) of the opamp, to disconnect opamp from 5V line.

@All:
There will be a Revision 1.1 of the Mainboard, that already has this mod done. So all the above is only true for Mainboard Rev 1.0

73 Manuel, DL2MAN
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: Fix for internal speaker howling

Post by Ohwenzelph »

If you can figure out how to do it, add some ESD protection
msax
 

Posts: 25
Joined: 15 Apr 2022, 19:12

Re: Fix for internal speaker howling

Post by msax »

Surely another fantastic step forward in the project. This weekend, if I find the time to mess about with it, I'll have a go to make that mod.
I'll keep an eye out for information on the new pcb revision so I can get fresh ones for the few of us in 9A who might be interested in the mod when and if they become available.

73
73 de Marko, 9A6PAF
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: Fix for internal speaker howling

Post by Ohwenzelph »

So are the filtering/conditioning of the power supply provided by the 100uH inductor and 470uF cap and 100 nF caps not needed for these components?
Not needed after all?
I could rewire and add duplicates of those components in the new path if there was a benefit. My impression is that no one else has. So are they not necessary? Or are they good but not important enough to go thru the bother and expense?
Last edited by Ohwenzelph on 13 Apr 2023, 01:35, edited 1 time in total.
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: Fix for internal speaker howling

Post by Ohwenzelph »

So I did the anti-howl mod but included in the alternate path a duplicate 100uH inductor “H”, a 470uF cap “C”, and a 100nF cap. Don’t know if it’s better with or without them cuz I have only tried it this one way.
F778B383-EF06-48A9-881F-2FE8EF44E73A.jpeg
F778B383-EF06-48A9-881F-2FE8EF44E73A.jpeg (142.46 KiB) Viewed 9992 times
Note C22 & C29 are NPO, so not microphonic, caps.
Other stuff:
“S” Schottky Diode RBS1MM40ATR, yes Anderson power poles would be better
“V” 14V Varistor VC120614D300DP
“T” 14V bidirectional TVS Diode/ESD Suppressor SMBJ14CA-13-F
Just sharing fwiw
73 Jerry aa1of
Follow up:
Howling much better but it still will sometimes if volume turned all the way up. Shorting past the duplicate inductor and caps does not seem significantly different, so probably not needed and also probably not bad.
Had very nice audio into head phones, tuning down to CW portion of 40 and suddenly it stopped. Birdies still there but the reception gone. Checked antenna and changed cable and such then.. Opened the rig and poked around and at some point, not clear if it was already that way, the wire to the floating end of D1 had come loose. Resoldered it but rig is still not good, did ‘factory reset’ but still no good. Tuning nob not tuning and very odd behaviour. I’ll try to do a boot loader and software reinstall next.
SO: The wire to the floating end of that tiny diode is precarious. Beware.
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: Fix for internal speaker howling

Post by Ohwenzelph »

Re did software, tuning knob and interface behavior are now better again but, it’s still Deaf to RF! I reinstalled the boot loader, with reburning the fuses, and reinstalled the firmware. Pretty sure the wire to D1 was intermittent connecting at least for a little bit. Culprit??? Atmega vs si5351 vs maybe Q1 got zapped vs maybe something else?
Suggestions please please please???
update-
replaced Q1, no help, and the removed one checked out as still ok; tested transmit and it still puts out rf at about the same strength and on frequency, so at least some of the si5352 is still working correctly.
Will have to drag out the o’scope…. Any insights out there? 73
Last edited by Ohwenzelph on 13 Apr 2023, 01:36, edited 3 times in total.
PabloW
 

Posts: 2
Joined: 10 Feb 2023, 11:29

Re: Fix for internal speaker howling

Post by PabloW »

@ Ohwenzelph : In your photo I see a mistake in the conecction to U6,
Wire should be conected to U6 pin and not to C20,
am I right?

73, DO5CTW
Attachments
F778B383-EF06-48A9-881F-2FE8EF44E73A-1.jpeg
F778B383-EF06-48A9-881F-2FE8EF44E73A-1.jpeg (142.46 KiB) Viewed 9781 times
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: Fix for internal speaker howling

Post by Ohwenzelph »

Yes, it looks like that because it’s hard to see pin 16 but no, that connection is ok. C20 is a 0.1uF cap to ground right at at the power supply pin, pin 16, for U6. Originally one end of C20 is connected to ground and the other end to a pad that supplies the 5 volt VCC line (i.e. 5v ‘after’ L1) from a path deeper in the board. That pad is also connected via a short trace to pin 16 of U6. Guido has you cut that trace and supply the the 5 volts with a wire connected to ‘before’ L1, but then you lose the 0.1uF bypass to ground right at at pin 16. I have the wire connected to pin 16 but also moved one end of C20 over to pin 16 so it is still bypassed to ground. Only the VCC connection from deeper in the board is disconnected and you can still see it as a tiny yellow dot in my picture in the black area where the pad was obliterated.
With mine you do lose the nearby bypassing of VCC to ground and that might matter in reducing the howl but I do not know.
Guido’s way is simpler and easier and works, so probably overall better but I couldn’t resist. : )
Ohwenzelph
 

Posts: 207
Joined: 01 Jan 2022, 03:47

Re: Fix for internal speaker howling

Post by Ohwenzelph »

Followed Manuel's instructions (viewtopic.php?p=3244#p3244), replaced U6 and reception is back! Q: why did U6 get zapped? My impression was the connection to the free floating diode end broke free and might have been intermittently connecting. That path leads to suppling 3.3 volts. How could that lead to U6 failure? It was working great and suddenly reception was lost. We will see how long it lasts..
New U6
New U6
IMG_0634.jpeg (217.66 KiB) Viewed 9679 times
replaced C20 with a 0402 size
epfly.yctseng
 

Posts: 4
Joined: 01 Nov 2022, 12:39

Re: Fix for internal speaker howling

Post by epfly.yctseng »

I'd follow the instruction to change the wiring. The result seem have solved the howling problem.
https://www.youtube.com/watch?v=_FLsuF1c2bc
Attachments
IMG_2619.jpg
IMG_2619.jpg (134.41 KiB) Viewed 9210 times
dogbit
 

Posts: 2
Joined: 23 Jun 2023, 15:50

Re: Fix for internal speaker howling

Post by dogbit »

Sorry for dumb question, just wondering if this upgrade will improve the receiver signal-to-noise ratio (sensitivity), or it solves the howling problem only? Thanks in advance.
DL2MAN
 

Posts: 706
Joined: 30 Dec 2021, 19:18
Contact:

Re: Fix for internal speaker howling

Post by DL2MAN »

dogbit wrote: 23 Jun 2023, 16:13 Sorry for dumb question, just wondering if this upgrade will improve the receiver signal-to-noise ratio (sensitivity), or it solves the howling problem only? Thanks in advance.
It does lower the noise floor a bit, so yes, you might say S/N Ratio is improved.
Roger Volver
 

Posts: 1
Joined: 27 Feb 2024, 06:21

Re: Fix for internal speaker howling

Post by Roger Volver »

Can not understand anything. Raise leg, cut off the path, throw a wire... I don't understand. And I’m afraid of doing something wrong, not being able to check using the new scheme. Please post a normal electrical diagram of the modification. Thanks in advance
Post Reply

Users browsing this forum: No registered users and 1 guest