V4XXL on my wishlist :-)

[DB9JG]

Since Christmas time is coming soon, there’s a nice German tradition of well-behaved children writing wishlists to the “Christkind.”

Maybe my wishlist will inspire not only the elves at the North Pole, but also those in China?

My V4XXL wishlist 😉

• The Theremin effect to disappear — helped by a new ferrite bead (FB) and a 10 pF capacitor at the LPF input,

• a headphone antenna circuit that can be detached with just one solder joint — no more FM interference,

• the telescopic antenna reaching 15 cm more out,

• a lighter capacitive load on both the antenna and the JFET caused by the FM path (18 pF → 10 pF),

• that unnecessary 24 MHz impedance dip finally gone,

• a higher input impedance above 15 MHz by moving the inductor from the JFET’s source back to the front,

• the RTTY/CW audio solder pads now extended with two extra pads for the R and C of an LPF, and

• maybe an improved shielding can for the SI4732 — its tone pitch surely deserves it!

Any comments and ideas appreciated.
But please, no comments now about whether I’ve been well-behaved or not 😉

Peter, DB9JG

[DB9JG]

I didn’t want to make my wishlist too long, but…
there are two additional wishes now from dear users:

"The metal tuning knob is a nice feature, especially together with a metallic case.
However, with the ABS molded housing, there is an annoying RF effect if the radio’s ground connection relies solely on the metallic tuning knob and your fingers."

and ...

"Dear elves,

Today, H.J. Berndt slipped a very modest wish onto my list:

Could you please provide a default link for the two existing solder pads between the amplifier and IO11? Thousands of users without a soldering station or microscope would greatly appreciate being able to use RTTY/CW—and perhaps even more future decoders.

TIA
Peter, DB9JG"

[G8PTN]

Hi Peter,
I noticed your note on one of the FB groups regarding the need for filtering of the audio signal feeding the ESP32.

I think it is probably a good idea to add the filter anyway to remove high frequency noise, but the NS4160 should be operating in Class AB mode.

On the original firmware the CTRL line only transitioned from LOW to HIGH on power up. However, in the ats-mini current firmware, it is also used for a workaround for some ATS-MINI hardware variants that do not have the audio clamp transistors to suppress the clicks when the SI4732 goes through a reset phase (e.g. switching mode).

The firmware now also uses the CTRL signal for this function, and it is also used for the squelch options. The firmware should respect the timing to ensure that the NS4160 is not operating in Class D mode.

I am not sure what happens on the hardware with the additional headphone amplifier that does not have the transistor clamps. If the CTRL line is not extended to this chip, clicks will still be present on the headphone output when the SI4732-A10 goes through a reset phase. I think some of the headphone amplifiers operate in Class D mode as well.

In addition, I am not sure what happens with the CTRL control signal on other firmware variants.

Partially translated information from the NS4160 datasheet is shown below.

73, Dave

[DB9JG]

Hi Dave,

thank you very much for clarifying.

We discussed this topic when investigating possible sources of "birdies" — I remembered your hints very well ;-)

I have to admit that the idea of inserting that RC filter came up when I looked at CW/RTTY signals on the IO11 ADC input with an oscilloscope and noticed the class-D fragments ... but that was actually while running the H.J. Berndt firmware.

H.J. Berndt’s tone detector runs at 1 kHz.
Unfortunately, the SI4732 doesn’t have a true CW filter — all audio filters start at 300 Hz (or is it actually 100 Hz?).
The 1 kHz bandwidth setting, which in fact seems to be 1.2 kHz as the 1 and 1.2 values appear swapped, effectively cuts the higher frequencies from the SI4732 output.

That makes CW detection a bit more susceptible to low-frequency noise between the Morse tones. With RTTY, the constant carrier helps to suppress these effects, so noise isn’t as problematic.

My latest filter suggestion (1 kΩ + 100 nF LPF) was intended to cut frequencies above 2 kHz for my preferred firmware and to keep the design ready for whatever comes next ... AND FOR THE ELVES IN CHINA;-)

I still hope for WEFAX detection — not because it will look beautiful on the Mini, but because it could be done entirely within this one MiniRX, without an external smartphone or other device :-)

Thanks again for your thoughts!
73, 55
Peter, DB9JG

[G8PTN]

Hi Peter,
Last week I noticed an Si4732 Mini receiver called “ATS-mini V3” on special offer for only £13.38 delivered to the UK, so I thought I better get one for spares and ongoing experimental purposes!! For my reference, I have called this hardware V3(B).

I think it is the same generation as the V2(B), which has the headphone amplifier fitted on the screen side of the PCB.
The V3(B) is effectively the same with the addition of the Hi-Z circuit.
The speaker is mounted behind the headphone socket, which is the opposite side to most other models.
It does not have the transistor audio clamps, but the workaround in the ats-mini firmware takes care of this.

For interest, I have had a quick look at the main circuit side of the PCB.

The JFET used for the Hi-Z buffer is marked “2AW”, which appears to be a BF826 or similar.

The ESD protection device connected to the antenna input is marked “AF”, which appears to be a PESD24VL1BA or similar 24V TVS diode.
I think the TVS diode is similar to that used on the V4 hardware.

I am not sure if the TVS diode will provide adequate protection for the SI4732 RF pins?

73, Dave

[G8PTN]

With my modded V3 and now the V4, I opted for the method that maximises shortwave reception and allows FM to suffice for local stations. This even effectively eliminates the Theremin effect. Therefore, I connected AMi and FMi to the JFET’s source.

Hi Peter,
The V3(B) also suffers from the VCO detuning as expected, although no worse than the two other V1's that I have.
I plan to modify one of them to become a simple VHF/UHF (FM only) version using the BK4802 chip.
#277 (comment)
73, Dave

[DB9JG]

Thanks Dave, that's a very interesting combination ... making things easy :-)

[DB9JG]

I did re-read the Si4732 containing several hints how to avoid the VCO signal being "radiated from pin x" etc
Simplified they recommended to provide a LPF at each and every pin :-)

And there are again some hints for a future better diplexer at the input:

The AM-path-inductance of this circuitry has been moved to the output of the JFET impedance changer in V4 now and creates a strange impedance dip at 24MHz , where this should be the other way around- lift the impedance at higher frequencies.

[G8PTN]

Hi Peter,
Yes, AN383 does imply that there is coupling of the VCO to most of the pins.

For the ATS-MINI, it really only becomes a problem in two cases, coupling capacitance variation to the device itself and the FMI signal path.

73, Dave

[DB9JG]

Yes Dave, my thoughts are inspired by the latest V4’s “Theremin” problem despite the presence of an LPF for VCO suppression. Furthermore, it features a long PCB trace for the headphone antenna which is ideal for radiating the VCO signal into the circuitry. Therefore, removing the headphone antenna capacitor could also be a good idea in this context.

And there are remarks about "noise" possibly caused by lack of current limiting resistors at some other pins. That made me suspicious even more :-))

[yo2naa]

I would add a switch (or internal jumper) to choose high-Z antenna (with internat FET amplifier) or 50 ohm external antenna, bypass the FET.

[DB9JG]

A jumper to the original AM input pin?
The Si4732 lacks a 50-ohm input.
It already has a high-impedance input with a 7pF||7kΩ impedance. This high-impedance design is intended for portable use with a telescopic whip of a maximum length of 1 metre.
If a 50-ohm termination is needed for an external passive or active antenna an un-un transformer, 3dB attenuator or simply a 50-ohm resistor will suffice for matching.