Version4 hardware at the horizont?

[DB9JG]

I've got a hint for a new V4 version that seems to have some fixed issues

https://www.banggood.com/AMNVOLT-V4-ATS-Mini-SI4732-Radio-ALL-Band-DSP-Radio-FM-AM-SSB-LSB-USB-Receiver-built-in-Hi-Z-Headphone-Amplifier-Circuit-with-Dual-Antenna-p-2042514.html?akmClientCountry=DE&cur_warehouse=CN&ID=6335577&rmmds=search&a=1761581806.8115&DCC=DE&currency=USD

V4 Update Content: 1. Added a shield to the Si4732 receiver to suppress electromagnetic interference; Add an ESD device to the antenna input to protect the chip.

2. The Hi-Z circuit has been moved to the front of the antenna path and uses a more stable VCC_RADIO power supply to preserve signal resonance and reduce noise.

3. Use wired headphones as an FM antenna (headphones are not included in the shipping package; this requires use in an environment with a strong FM signal and a long headphone cable for optimal performance).
4. The left and right audio channels have been swapped; the amplifier and speaker parameters have also been retuned to improve sound quality.
5. Replaced with more expensive and superior accessories: telescopic antenna and a non-slip knob cap with a better feel.

[maurilored-stack]

OM0ET
's video just released on YT

[EmEiBee]

... hmmm ... Maybe time for a new one ;)

[omixann]

Apparently (as per OM0ET) now there are pads to place a 0 ohm resistor and connect audio output back to the ESP32. This could allow things like CW/RTTY decoding.

[DB9JG]

Thats right ...but ...
That was my facebook post a few hours ago after my new V4 had arrived

V4 … It‘s your choice … but …

… I am a bit (?) disapointed now !

My VERY first impression of an AMNVOLT ATS MINI V4 :

• Firmware v2.33 Sep. 22, 2025

• Neatly crafted housing with well-cut, open slots for the speaker

• Robust telescopic antenna with a well-handling SMA plug

• Telescopic antenna shorter than expected but appropriate – only 40 cm instead of 60 cm

• Grippy aluminum tuning knob with recessed fastening screw. The screw hole could very well be used for an additional tuning lever

• Without external antenna or headphone cable:

  • FM stations are already received without any antenna, indicating apparently very poor shielding of the FM section
  • Changes in the display content cause audio interference in FM reception
  • Strong birdies without antenna in SSB mode within the ham bands, every 16 kHz (display PWM?)
  • Birdies frequencies shift slightly when the casing is touched and/or pressed (theremin effect)

With telescopic antenna:

• Very strong theremin effect when touching the base of the telescopic antenna
• Very strong theremin effect even just by shaking the receiver and telescopic antenna in the hand
• antenna is short compared to the formerly sold version. Likely worse capacitive probe effect and less signal strenght with this antenna - but: sensitivity not yet investigated

With headphone antenna:

• No strong shortwave signals coming through via headphone wire !!!
• 2 Hz ticking sound overlays FM stereo audio even at medium strong RF signals!!
• Clear sound but lacking bass frequencies on my good iPhone headphones though deemphasis is set correctly for EU!
• With headphone and telescopic antenna connected, both antennas create either an additive or canceling signal effect when one of them is moved
• Even stronger FM signals show a noisy interference obviously corresponding with activity of the ESP32 or the display

I did not look at the inner values until now.

And I did not compare it with my modded V3S attached to an antenna splitter.

Although …

… I have to take a deeeeeeep breath because of the not so well start 😞

EDIT, AFTER SOME MORE EXPERIMENTS:

Sorry, but I cannot recommend this version to anyone who is not able to open it and do some work to bring it up to an acceptable level. This is not a step forward!

trim.188F2077-11F4-46B9-BDC0-5506489779D5.MOV

trim.4F5446B9-1C3B-4A04-BBF8-45347FC5B2CE.MOV

Meanwhile I started modding this version!!!!

That more mechanically "creepy sound" issue was easy to fix—though I never expected to be doing that on a brand-new device!
Just placing a two-layer piece of solder wick on top of the Si4732 under the hood stopped it from flexing and producing those odd noises when shaken.
However, the electrical load sensitivity at the SMA connector still persists, of course. It’s not exactly convenient to apply my 7-turn coil modification under that shielding hood now.

[DB9JG]

some more hints and findings:

V4 - The First Mod

Maybe you noticed how disappointed I was with the first impression of my newly arrived AMNVOLT Mini Radio V4 (ABS housing).

One of the most annoying issues is a strong theremin-like effect observed in SSB modes. The whining sound occurs when mechanical stress bends the PCB or when a change in capacitive load is introduced at the SMA socket or the base of the telescopic antenna.

This effect appears even though there is now an additional filter at the FM input and a shielding hood installed over the SI4732.

I found that the shielding hood itself causes this behavior because it bends easily and acts as an efficient antenna for the internal 3.4 GHz oscillator.

The problem was easily fixed by placing a short piece of solder wick (two layers) between the hood and the surface of the SI4732. This copper wick fills the variable gap and stabilizes the capacitive load on the VCO, preventing unwanted coupling to either the hood or the SMA socket. I added a small amount of glue to keep the solder wick in place.

I haven’t soldered the hood at one corner yet, as I’m expecting to fix additional issues at the FM input.

The first issue is now resolved.

There’s a chance this device might win back my heart. ;-)

[DB9JG]

this is even more worse than I ever thought!

The 3.5GHz VCO leakage of this v4 miniradio at the SMA input is now again at -34.9dBm (I brought it to -54dBm at my V3S)

That is definitely BAD and an explanation for the strong theremin-like sound effects when touching the antenna - despite the new shielding hood and a TPF at FMi-pin!!!!

This will not be easy to fix, ax there is less space for an inductance.

The very brutal work around could be this (- changes in GREEN)

[G8PTN]

Hi Peter,
A few thoughts below.

I am not sure what component values have been used for the added low-pass filter on the ATS-MINI V4 hardware.

A simplified simulation with a 50Ω source impedance and ideal inductors is shown below, based on “Option 2” component values from AN383.

I wonder if the layout on the V4 hardware is non-optimal and the filter is not working as expected.

With the 7-turn air spaced inductor that you have used previously, the attenuation may have been higher, although the pass band attenuation may also have been greater.

If a solution is possible after the high impedance buffer, it may actually be better. Perhaps the HPF and LPF can remain in place?

73, Dave

[EmEiBee]

* The headphone antenna connection seems not to be at GND of Headphone socket (?). Headphone seems NOT lifted from ground !!!

Hmmm... That make no sense in my eyes...
In radios that use the headphones as an antenna, I've always seen that the headphone's ground (GND) is connected to device-ground via an RF choke, allowing the headphone's ground wire to be used as an antenna. Since headphones often use shielded cables (the cheaper ones usually don't), any other wire would be within the shield and therefore hardly usable as an antenna.

[DB9JG]

Yes. That’s how AN383 recommends it on page 18, 50, etc and how it should be now ...
I guess that component left of the headphone socket is the "lifting" component, but I do not really care.

I don’t like a headphone FM antenna or its PCB traces to be able to pick up signals from the digital surroundings. So, I’ll remove the coupling capacitor. The trace can be left open since it will only be connected to audio components.

My concern now is about how to insert that coil between an SMA pin and the first capacitor, respectively, the ESD diodes.
It seems impossible to lift the SMA pin, and I’m hesitant to cut or drill at that point.

[vegos]

Peter, desolder and remove the sma. Shorten the center tab, solder again...

[DB9JG]

Antonis, I understand that would be the logical way, but I’m still searching for a viable solution for others as well.

Also, I can’t remove the tuning knob – I can’t turn the screw. :(

[DB9JG]

#No 3.4GHz ... No Cry!!!

The -56dBm interims solution looks like this:

[G8PTN]

Hi Peter,
Great work. I did wonder about just adding the inductor externally.
I guess it still detunes if you touch around the adaptor.
It seems to show that the internal filter is not enough to block the 3.4G problem.
Perhaps tweaks to the SMD component values can also result in an adequate fix without seriously affecting the FM sensitivity.
73, Dave

[DB9JG]

Hi Dave,
This solution—or rather workaround—is the safe approach. In the best case, it only involves adding a ferrite bead to an existing un-un impedance transformer to properly match the 50-ohm coax to “whatever.”
I suppose increasing the SMD inductance at the FMI would be the ideal solution.
My test of shorting the FMI to ground resulted in only −50 dB, so there is probably a second path to the SMA, and keeping the coil there still seems like a good idea.
Glad I didn’t have to resort to my next plan: adding a 3.5 GHz λ/4 open piece of semirigid coax as a stub.

But the journey is the goal, and it’s satisfying to do the work for engineers who likely grew up with iPads in their hands and may not actually know what VNAs or TinySAs are used for. :-)
I’m afraid that no matter what we discover or recommend next, the Chinese business model probably won’t allow the implementation of a truly final and perfect version. So what comes next then?

73&55
Peter

[vegos]

What about replacing the values of the parts for the existing filters?

[DB9JG]

Having said that, I think I’ll try adding a coaxial suck-out circuit—just for fun. :-)

[DB9JG]

What about replacing the values of the parts for the existing filters?

Yes you are absolutely right - but did you see how tiny the three components are?
AND: I have SMD Cs and Rs but no Ls :-)

[EmEiBee]

... and we're back to "we're making our own circuit board as a community work" ...

[DB9JG]

... and we're back to "we're making our own circuit board as a community work" ...

No, no — it’s way more entertaining to watch the chaos others create than to be the mastermind behind it!

[G8PTN]

Hi Peter,
I suspect as previously discussed the ATS-MINI hardware will be a never ending cycle!
It would probably be best to completely isolate the external port from the FMI pin on the SI4732.
On the ATS20+ there is a cheap BP78108M of the shelf bandpass filtered followed by an LNA.
I am a little surprised that the VCO detuning had never been raised before, but it only becomes noticeable when listening to SSB, which was not the original use case for the SI4732 family of devices.
73, Dave

[DB9JG]

Hi Peter,

I suspect as previously discussed the ATS-MINI hardware will be a never ending cycle!

It would probably be best to completely isolate the external port from the FMI pin on the SI4732.

On the ATS20+ there is a cheap BP78108M of the shelf bandpass filtered followed by an LNA.

73, Dave

You’re right, but regarding the “MINI”-radio, the fun comes from being so small and simple yet effective. If the impedance converter could be designed for 0 > 100 MHz, preserving the high impedance at < 30 MHz, then the already discussed diplexer solution behind it could be simple and effective.

[DB9JG]

V4 – A Long Story Short… but Not Yet Complete

There is currently a lot of discussion about the AMNVOLT ATS MINI V4.

My very personal interim assessment:

• The FM mode is significantly impaired by the new headphone antenna feature. There is a ticking sound and signal suppression between the telescopic whip and the headphone cable, leading to increased internal and external interference. This effect even seems to cause some more birdies in SSB mode. The issue is not very easy to resolve, as it requires removing an SMD capacitor to disable the headphone antenna.

• Theremin effect due to mechanical stress: A shielding hood covering the SI4732 changes its distance from the chip’s surface, causing a wobbling sound in SSB modes. #### Solution: A piece of conductive solder wick can bridge the gap between the shield and the IC. WARNING: The hood is very close to the battery switch and may cause a dangerous short if moved.

• Theremin effect when listening to SSB stations and touching the antenna base or changing the load at the SMA socket. The wobbling sound occurs because an internal signal is not sufficiently suppressed and temporarily detunes the SSB signal. #### Solution: Difficult, as some SMD components appear to be poorly dimensioned and are not easy to modify. A small 7-turn inductor cannot be easily inserted at the SMA socket. An external workaround helps, but requires an interface circuit at the SMA socket or an un-un impedance transformer with an additional ferrit breed.

• In contrast to the new V4, my V3S exhibited only the Theremin issue, which was easy to fix in that version. If I could buy a V3S for less than a V4… 🙂

The photo shows my test setup. The 3.4 GHz VCO signal is now at –56 dBm instead of –35 dBm, which eliminates the Theremin effect.

I am still working on developing a DIY solution for the Theremin issue that doesn’t require advanced electronics skills. Any suggestions are greatly appreciated! ;-)

[DB9JG]

V4 - Open Heart Surgery

Today, I continued my investigation on the AMNVOLT MiniATS V4, focusing on a DIY solution suitable for medium skill levels to improve FM reception and eliminate both the Theremin effect and unwanted birdies.

I concluded that replacing the SMD components of the FMI LPF is not a practical approach. Instead, I explored ways to add the required inductance directly at the SMA pin outside of the shielding hood.

To achieve this, I used a 1mm drill and a Dremel to cut a gap into the SMA pin at half its length. To bridge this gap, I used a 12-turn air coil with a diameter of 1.2mm wound around a PVC core. Beware of the metal dust from drilling!!! Draw the baterie plug while working and cleaning!

I though about only lifting the SMA pin from the PCB as an alternative solution for the gap - but that was not possible.

It was not possible to cut the trace directly at the tip of the SMA pin because there’s a through hole leading to the ESD diode on the PCB’s backside.

Additionally, I removed the coupling capacitor from the headphone antenna circuit to eliminate the interferences and the double antenna signal issues.

##############################
As a result, the VCO leakage at 3.4GHz measured -70dBm at the SMA antenna socket of the V4, compared to the previous -35dBm.
The Theremin effect completely disappeared with this.
Shortwaves aren’t affected by this coil.
FM will perceive this as an elongation of the 40cm telescopic whip, making it even more resonant at 90MHz.
The headphone antenna is deactivated
##############################

Although making the cut in the SMA pin is not as straightforward as I had hoped, it remains a viable solution for users with some experience in electronics and mechanical work. Alternatively, an external damping coil, as previously described, can also be used.

Below are some photos of this “surgical” modification.

THIS SUGGESTION IS NOT SUITABLE FOR USERS WHO ARE INEXPERIENCED WITH WORKING ON ELECTRONIC CIRCUITS. PROCEED WITH DIY AT YOUR OWN RISK!

73&55

GOOD LUCK!

Peter, DB9JG

Below are some photos of this “surgical” modification.

[vegos]

Ok, with a little words, that's what I'll spend my tomorrow afternoon on.. :)

The theremin-effect is now reduced? Any other observations?

[centroparotjag]

What is the inductance of the added coil?

[DB9JG]

That is about 20nH , but it is very uncritically, as the low pass effect on the 3.4GHz counts.

The limit is the attenuating influence of this coil on the FM path.

Meanwhile I successfully used a small ferrit bread with only two turns of thin wire.