Noisy JFET supply voltage of the MiniRadio V3S

[DB9JG]

The AMNVOLT ATS MiniRadio V3S addresses the battery drain issue caused by the JFETs’ direct connection to the battery. However, even with the V3S hardware, the drain of the JFET still connects to the noisy Vcc_Bat.

I investigated the quality of the supply voltages. There’s a significant noise level at Vcc_Bat, which affects the impedance changer’s source. This noise originates from the PWM-backlight, the display IC, ESP32, and, in some cases, the class-D audio amplifier.

To eliminate this noisy supply voltage at the JFET, I suggest connecting the Drain to the 3.3V supply Vcc_Radio of the SI4732 (changes marked GREEN). After making this change, I could no longer measure the previously noticed noise floor at the AMI.

I reverse-engineered the antenna inputs and noticed that there are no capacitors to DC-block the JFET’s input or output.

To mitigate the lack of input voltage protection and the theremin effect in SSB mode, I added some components (RED).
While these changes don’t eliminate all “birdies” in the receiving spectrum, they are a remarkable piece of the puzzle.

[max-arnold]

I wonder whether this non-AMNVOLT V3 is also affected by the noisy Hi-Z power

More photos here https://t.me/talkradio/174172/233494

[DB9JG]

Hi Max
This is an interesting version!
It features a headphone amp on the backside, a high-Z stage with a protection diode, and a capacitor between the source and AMI. There’s another capacitor, but I can’t tell if it DC-blocks the gate or decouples the drain. The placement near the PWM transistor and its Vcc_Bat is a bit suspicious. I guess it uses Vcc_Bat - but who knows, without having access to such a device?

This will definitely demonstrate the theremin effect, as the FMI is not decoupled by a VCO blocking inductance, only a high-pass filter.

It’s very unfortunate that NONE of the manufacturers provide schematic diagrams, making it difficult to assess whether a design is better or worse. I suppose they don’t do that without intention, :-(

[max-arnold]

NONE of the manufacturers provide schematic diagrams

Yeah, and also they violate the CC BY-NC-SA 3.0 license used by Sunnygold for the original open source 1.0 receiver hardware...

[vegos]

I am having trouble with the one I almost destroy with the dremel and have spend almost 4 hours with no success yet.
No schematic, no gerber files, no everything :(

[max-arnold]

Another revision has been spotted. HiZ or preamp?

[max-arnold]

It is sold as a motherboard/display/antenna kit https://www.newegg.com/p/1W7-00W4-00Z64

The search keywords are "BD3CS-588 Si4732" (yeah, with a typo, although the board is marked as BD3SC)

[DB9JG]

It’s really nice, although we’re somehow back at V1. The antenna input seems to be improved, but I can’t see a headphone amplifier. A schematic diagram would be a kingdom!

[max-arnold]

BD3SC-588 in action https://so.douyin.com/s?douyin_web_id=7550085162957407780&enter_method=web_search&in_ogs=40&in_tfs=OS&keyword=%E8%A2%96%E7%8F%8D%E6%94%B6%E9%9F%B3%E6%9C%BA%E5%85%A8%E6%B3%A2%E6%AE%B5%E5%8D%AB%E6%98%9F&original_source=40&query_correct_type=1&search_entrance=overseas&traffic_source=ZY1112&is_no_width_reload=0&actv_aid=7546864978487201076&actv_atype=0&actv_agid=7546864978487201076

[DB9JG]

I’m not a fan of those non-tunable donut loops, but if he wants to compare them to a tunable ferrite rod, he should at least have turned it in the same direction as the wire loop around the ferrite rod.

I can see here a PCB without a headphone amplifier, battery, speaker, or USB-C cable for around €30. Let’s wait a bit or decide that this likely won’t be a version sold in large quantities.

[vegos]

@DB9JG my V2 has the headphones amp on the other side, maybe this is the same setup...