Loud click when changing band.

[Christian3x]

Isn't possible to mute or turn down volume when changing band to avoid loud click ?
Version 2.21

[max-arnold]

• Which bands produce the click (AM, SSB, FM)?
• Does it happen in the latest test build? https://github.com/esp32-si4732/ats-mini/actions/runs/15204365584
• Whether the click is audible on the built-in speaker, headphones or both?
• What hardware version of the receiver you have - V1, V2 (with headphone amplifier), or V3 (amp + Hi-Z)?

[mengzi0727]

v2 version, 10M-CB-VHF-ALL, MW1-160M, and all memory

5.31.1.mp4

[Christian3x]

Yes they have to put mute in when switching band.
void selectBand(uint8_t idx, bool drawLoadingSSB)
{
muteOn(true);

muteOn(false);

}
This works on any band switch. I just merge my changes and then I get as I want :).

[mengzi0727]

I don't know anything about coding (ToT)

[Christian3x]

Then we got to get them to put it in the code for these new versions, it is normal that you mute a radio when setting up the frequency synthesizer otherwise it will be noisy.

[Christian3x]

All band switch, loud click.
Version 2.21, I can see there is a new version but if there isn't done anything for the click it should still bee there.
Click is in the build-in speaker, have not tried headphones.
My ast is a version 2 with headphone amp.
You can't avoid a loud click if the sound is NOT muted when changing the frequency synthesizer.

[Christian3x]

Precisely, that’s why I have downloaded the source and done a few changes my self. I don’t have the clicks and I have changed a few other things.

[max-arnold]

@mengzi0727

The popping sound is amplified, which is a software problem.

It is not a software problem, it is a hardware problem. Just confirmed with another owner of V2(B) PCB (where the headphone amplifier is located under the display). This PCB version, as noted by Dave, lacks the audio mute circuit (present in the original schematics) that is driven by the firmware.

Because it is not an isolated hardware problem, I'll try to push a workaround next week, if it doesn't cause new problems on other PCB versions.

[Christian3x]

Yes I think maybe it's good to make a small work around in SW because I guess there will be sold more of the v2.
Putting mute in "selectBand(uint8_t idx, bool drawLoadingSSB)" works also in SSB mode.

[max-arnold]

Does it click for you when you switch modes (am/lsb/usb)?

[Christian3x]

The 2.27 does not but I had the issue in 2.23 ? and I put in a small delay and that worked. esp_rom_delay_us(1000); can be shorter but it doesn't matter it's so short anyway so you don't notice it.
void selectBand(uint8_t idx, bool drawLoadingSSB)
{
muteOn(true);

esp_rom_delay_us(1000);
muteOn(false);

}

[Christian3x]

OK, installed "arduino-cli" and fetch v2.22, I wrote "arduino-cli compile --clean -e -p COM_PORT -u ats-mini" it build but of course no COM.

Checked the bin file and I can se my path is in the file so it will never be the same.

But do you set any of these options
DISABLE_REMOTE - disable remote control over the USB-serial port
ENABLE_HOLDOFF - enable delayed screen update while tuning
HALF_STEP - enable encoder half-steps (useful for EC11E encoder)

I made a build with change and it didn't work still click. I will try to find the problem.

[Christian3x]

Hi
OK this works NO clicks when switching band.
void useBand(const Band *band)
{
muteOn(true);
// Set current frequency and mode, reset BFO

muteOn(false);
}

Don't know what's wrong with "tempMuteOn(true);" is doesn't work but muteOn(true); works fine, so maybe you could add it in future versions, I can't be the only one with this problem, I just bourght my ets-mini so maybe I'm one of the first in this version v2 from a new batch..

[cdtsuipo]

Have you opened the back cover of the radio?Early on I noticed that the speaker wire, because of its length, would produce a CLICK sound as it folded over and passed near the antenna holder.You might try organizing the speaker wire.

[Christian3x]

Hi
Thanks for the tip, I tried to twist the wires and move them, remove the mute but then the click was back, normally RX doesn't interfere like that, TX does but again it really sound like frequency synthesizer noise when changing and that is quite normal if you don't mute. I have reinserted the mute and now it works nice and smooth so I'll keep it like that.

[Christian3x]

After playing around with it I found that there was still a click when it switch to SSB so I moved the "muteOn(true);" to:
void selectBand(uint8_t idx, bool drawLoadingSSB)
{
muteOn(true);

muteOn(false);
}
and now it works on every band switch.

[Christian3x]

Is there a way to Debug can I send debug over serial port ??
or do I have to make a debug routine.
I'm not quite satisfied with the signal indicator.

[max-arnold]

Use Serial.println()

[Christian3x]

Thanks, I have been trying to do a calculation to get the right S0-S9+60 but that's not possible on this chip "Si4734/35" because the RSSI is not true dBm it's just a level indicator so I have just altered getStrength() and changed the noise floor to rssi=20 and that makes a more realistic readout. Don't know if you are interested in that.

Key RSSI Details for Si4734/35:
Default RSSI Range:

The RSSI value is not in dBm but is a relative signal strength indicator.
The datasheet suggests that higher values = stronger signals.
Typical observed maximum RSSI in real-world conditions is around 100–120 for very strong signals.

Empirical Observations:
S0 (weak signal): RSSI < 20
S9 (strong signal): RSSI ~ 90–100
S9+60 (overload): RSSI > 150 (but may not be reliable beyond ~120–130)

Register Details:
RSSI is read from 0x26 (RESP0) in the STATUS response after tuning.
The value is not calibrated to dBm, so exact conversion depends on your setup.

[Christian3x]

I have changed some of the bands to what they are in real life.
{"FM", FM_BAND_TYPE, FM, 8750, 10800, 10400, 2, 0, 0},
{"FM+", FM_BAND_TYPE, FM, 6400, 10800, 10280, 2, 0, 0},
{"LW", LW_BAND_TYPE, AM, 150, 531, 210, 0, 4, 0},
{"MW", MW_BAND_TYPE, AM, 531, 1602, 855, 0, 4, 0},
{"MW+", MW_BAND_TYPE, AM, 1602, 3001, 1602, 0, 4, 0},

[G8PTN]

Hi,
The audio click/pop issue is documented in the PU2CLR GitHub area and is due to the audio outputs from the SI4732/5 changing DC level when the device is reconfigured (e.g. mode change). The suggested audio clamp circuit is shown below.

On the original ATS-MINI hardware V1, the audio clamp circuit from the schematic is as follows.

On subsequent hardware versions with the added headphone amplifier, there still appears to be some sort of transistor clamping circuit, but the schematics have not been published. It is not known if the “MUTE” or “AMP_EN” signal is connected to the headphone amplifier.

With respect to the reported issue in this thread, for whatever reason the audio clamping is not working as expected on hardware V2(B). This may be due to an error in the design or a result of how the audio clamp circuit is working in conjunction with the headphone amplifier.

The workaround will disable the speaker amplifier via the AMP_EN signal on a mode/band change, but the headphone output should also be checked.

It is unclear what other hardware differences may exist, and therefore maintaining compatibility with a single firmware build could become a problem. For example, the version with the Hi-Z buffer would not work as expected if the auto tuning capacitor feature is enabled on the SI4732/5 front end. The auto capacitor function was designed to optimise the reception for LW and MW bands.

There are also component differences such as the ESP32-S3 variant and display (driver) type. At the moment, there are two known display types and the ats-mini firmware automatically detects the type and adjusts the parameters accordingly.

73, Dave

PCB images of the versions that have be reported so far.

Another version has been identified without the PCB tracking for the BOOT and RESET buttons.

[Christian3x]

Why all that when you can solve it by adding mute when selecting band I have done it and it works perfect so all that is not necessary.
void selectBand(uint8_t idx, bool drawLoadingSSB)
{
muteOn(true);

muteOn(false);
}
Mine is a V2..

[G8PTN]

Hi,
On the original design, there was no headphone amplifier, so if you connected the 3.5mm output to an external amplifier, you would still get the clicks. In addition if the audio amplifiers did not have an enable pin, you would still need to add the discrete clamps.

The original firmware does drive the "MUTE" signal (part of the PU2CLR library), but your hardware does not appear to work as expected.

73, Dave

[Christian3x]

Hi
I just got mine a week ago and it's a v2 with amp, I have been messing around in the sw and tried to put mute in to useBand() that worked for everything but SSB so I moved it to selectBand() and that works on all band switch and I will say that it is normal when your making software for receivers, when you mess around with the frequency synthesizer you mute otherwise you get problems with noise.
I have NO clicks on my v2 now.

[G8PTN]

Hi,
This problem is not related to the frequency synthesiser in this case. It is because the device has to effectively go through a reset sequence to start the different modulation modes in DSP. The initialisation causes the DC shift on the audio pins. A jump between frequencies should not cause this particular problem.

73, Dave

[Christian3x]

Hi
No of course not then it would click every time you role the dail it's normally when you setup the synthesizer but again it works fine putting in the mute the right place and of course if there is a DC problem it would click but it doesn't sound like that.
73, Chris OZ1CD/EA5JPZ

[max-arnold]

AMP_EN can also switch NS4160 into different modes https://www.lcsc.com/datasheet/lcsc_datasheet_2304140030_Shenzhen-Nsiway-Tech-NS4160_C219017.pdf

[G8PTN]

Hi Max,
Hopefully, the firmware would not cause the situation for Class D selection, which may generate EMI. If it was a problem, some small delays could be added in the mute function.
73, Dave

[Christian3x]

Do you Know what pin the ctrl is connected to ? Just so when the radio is powered up 2 pulses are generated. For class A/B ctrl is just high.

[Christian3x]

OK found it pinMode(PIN_AMP_EN, OUTPUT); should make a function so it can be switched between AB and D so D can be used in VHF mode to save battery.

[G8PTN]

Hi Chris,
Although Class D is more efficient for power, I would have thought it is undesirable from a noise point of view, particularly with the speaker wires being right next to the receiver chip.
73, Dave

[max-arnold]

Does tempMuteOn() work if you put it in selectBand?

[Christian3x]

I tried that but it didn't work for me.

[Christian3x]

Vid.20250601.010523.mp4

This is with the mute in selectBand.

[Christian3x]

This is my version. I tried to plug it into my stereo and I'm sure it played in stereo.

[mengzi0727]

这是我的版本。我试着把它插到我的立体声音响里，我确定它在立体声中播放。

看起来，我的机器和你的一样

[G8PTN]

Hi Chris,
This looks like hardware V2(B), so the headphone amplifier will probably be on the other side of the PCB.
Unfortunately, there is no schematic for this version, so you would need to trace the signals to work out what is actually happening.
The original schematic for hardware V1 is in the link below.
https://esp32-si4732.github.io/ats-mini/hardware.html

I guess as long as the f/w patch is working for you, it is best to leave it alone.
It would be nice to understand how it has been connected. From the photo, it looks like both “MUTE” and “AMP_EN” are connected on the ESP32-S3 module.

PS: The 3.5mm output should be stereo.

73, Dave

[jumbo5566]

the AMP_EN signal both connect NS4160 _1 _CTRL and PAM8908 _13_EN ,just change v2.23 -> Menu.cpp ->line 874 add
digitalWrite(PIN_AMP_EN, LOW);
line 898 add digitalWrite(PIN_AMP_EN, HIGH);
tested ok

Menu.zip

[jumbo5566]

void selectBand(uint8_t idx, bool drawLoadingSSB)
{
// Set band and mode
bandIdx = min(idx, LAST_ITEM(bands));
currentMode = bands[bandIdx].bandMode;
digitalWrite(PIN_AMP_EN, LOW);//fix Loud click when changing band
// Set tuning step
stepIdx[currentMode] = bands[bandIdx].currentStepIdx;

// Load SSB patch as needed
if(isSSB())
loadSSB(getCurrentBandwidth()->idx, drawLoadingSSB);
else
unloadSSB();

// Set bandwidth for the current mode
setBandwidth();

// Switch radio to the selected band
useBand(&bands[bandIdx]);

// Clear current station info (RDS/CB)
clearStationInfo();

// Check for named frequencies
identifyFrequency(currentFrequency + currentBFO / 1000);

// Set default digit position based on the current step
resetFreqInputPos();
digitalWrite(PIN_AMP_EN, HIGH);//fix Loud click when changing band
}

[Christian3x]

Works fine on my v2 but OK I se that muteOn(true); does the same.

[jumbo5566]

i add the code change AB to D mode
digitalWrite(PIN_AMP_EN, HIGH);
delayMicroseconds(6);
digitalWrite(PIN_AMP_EN, LOW);
delayMicroseconds(6);
digitalWrite(PIN_AMP_EN, HIGH);
sound 55 VHF Current form 2xx mah MAX to 1xx mah MAX
it works!

[Christian3x]

Sounds great I didn't have a schematic so if you have one of the v2 I would really like one, thanks..

[jumbo5566]

i just add PAM8908 Chip ,si4732 to PAM8908 then to Phone ,then to NS4160, so simple.

[G8PTN]

Although Class D is more efficient for power, I would have thought it is undesirable from a noise point of view, particularly with the speaker wires being right next to the receiver chip.

Hi,
Is there any hash in Class D mode from the switching frequency or its harmonics?

73, Dave

[jumbo5566]

i tested harmonics on MW LW SW noise anywhere.

[jumbo5566]

on 2.27->Menu.cpp-> line 934 i add:
// Set default digit position based on the current step
resetFreqInputPos();
if(bandIdx==0){
digitalWrite(PIN_AMP_EN, HIGH);
delayMicroseconds(6);
digitalWrite(PIN_AMP_EN, LOW);
delayMicroseconds(6);
digitalWrite(PIN_AMP_EN, HIGH);
}else{
digitalWrite(PIN_AMP_EN, HIGH);
}
Class D mode is only For VHF ,it works great,VHF Current is 120mah MAX,Others is Class AB type ,2XX Mah MAX

[jumbo5566]

when sleep mode on,just 50mah Current on VHF(FM) Speaker Loud 55 ,
so great,750mah battery is 15hours listen!

[max-arnold]

Could you please test the following build to see whether it fixes the click?
ats-mini-test-build-v2.27-mute-click.zip

[max-arnold]

I still do not understand the testing results... Let me repeat the question... Below are two test builds:

1. ats-mini-test-build-v2.27-mute-click.zip
2. ats-mini-test-build-v2.27-mute-click-2.zip

Which of them (if any) fixes the "Loud click when changing band" issue?

[Christian3x]

ats-mini-test-build-v2.27-mute-click-2 works perfect NO clicks with band switch or squelch

[Christian3x]

max-arnold
Can I ask why there are 2 mute functions muteOn() and tempMuteOn() ?
Normaly you would manage with just one and maybe some parameter.
Also there are missing a few libs for v2.28 when you try to build, i switched to the old ones.
Missing:

• Async TCP (3.4.4)
• ESP Async WebServer (3.7.8)

[max-arnold]

muteOn() is used for normal muting. If you call it for squelch or band switching, it will unmute the receiver if it was muted. To prevent that, there is the tempMuteOn() that doesn't affect the mute state.

The libraries are not missing. You need to update the index: arduino-cli lib update-index

[Christian3x]

Sorry I was just curious and thanks for the lib update.

[Christian3x]

Now that the lid is off there is also a bug in CB channel calculation in findCBChannelByFreq(uint16_t freq); MIN_CB_FREQUENCY and MAX_CB_FREQUENCY are wrong so it shows channels in 10m band also, but change the def and it is correct.

// CB frequency range
#define MIN_CB_FREQUENCY 26960 // old value 26060
#define MAX_CB_FREQUENCY 27405 // old value 29665