Shortwave Radio Picks Up Sideband

With the push to having most of a radio receiver as part of a PC, it might seem odd to have a standalone communication receiver, but [OM0ET] reviews the latest one he picked up, an ATS25. Inside isn’t much: a battery, a speaker, an encoder, and a Si4732 that provides the RF muscle.

It appears the receiver is pretty broadband which could be a problem. [OM0ET] suggests adding selectivity in the antenna or adding an extra board to use as a bandpass filter.

The design is simple enough, we are sure you could easily hack the unit to do different things. Most coverage stops at 30MHz, but there is an FM band, so we wondered if you could get the thing to work on other frequencies, too.

Clearly, the Arduino portion would be easily hackable. For the price, we were both impressed with the touchscreen and build, but maybe less impressed with the RF filtering. On the other hand, the small form factor would be great for backpacking or portable use and it isn’t that expensive. It does seem to work pretty well in practice.

We’ve seen similar homebrew radios using the same chipset or, at least, a similar one.

14 thoughts on “Shortwave Radio Picks Up Sideband

  1. I’m no RF expert, but I suspect there is no need for any added selectivity, and in fact, that would make it more difficult to tune, as you’d have the continuously tune that “added selectivity” to the right frequency. I have a Sony all-band receiver that’s over 30 years old that uses digital tuning and has no added selectivity of any sort. I have never felt that it was needed.

    1. Virtualy no traditional receivers have no selectivity between the antenna and first stage.

      A Racal from the fifties that upconverted had the option of just a lowpass filter, so it only “saw” 0-30MHz. I have no idea if that was used much. It had a traditional preselector too, one could switch between the two.

      Fifty years ago, a NE561 PLL was shown as a project for a simple receiver. Generally no tuned circuit at tge front. Nobody mentioned problems , but I have my doubts. They weren’t intended to be high performance.

      For a long time, superheterodyne receivers used one control to tune the front end and local oscillator. When higher first IFs came along, in the MHz range rather than 455KHz, a lot of receivers separated the two controls, so “tuning” and a separate preselector.

      But then it swung back again. So your shortwave receiver does both at the same time. You just don’t see it.

      For traditional receivers, you either need really high dynamic range active devices, or decent front end selectivity. Some go straight to a passive mixer, followed by a narrow filter, but that’s still prone to overload problems.

      If an SDR, and these are a specific subset, have enough bits, that improves dynamic range. I don’t think these have it. They try to get away without preselection, but I gather there are limits.

      Even a simple preselector cuts down on what the input sees. The less sharp, the less it needs to be retuned as the frequency changes. Some of the strongest signals a receiver will see, unless they live next to a transmitter, are AM and FM and TV transmitters. Keeping those out help a lot.

      People used to use some really cheap and bad receivers. So a separate preselector was not an uncommon accessory. So back to two knobs.

    2. It depends on quality. The wider the band admitted to the amplifier sections, the more noise you are putting into the signal chain. Many receivers, like most Racals, put a bandpass filter in front of the receiver to limit the spectrum being dealt with. If you don’t have this selectivity, you must design the receiver to reject noise of a much wider band which can complicate the design of components throughout the entire receiver. If you put say a 0 – 30 Mhz filter in front of your receiver, you only have to design around that band.

      A lot of multiband receivers are in fact several receivers in one box with filters to limit each receiver to seeing only the frequencies they are designed for.

  2. I am very wary of SDR and software radio in general. Seeing that generating a sine wave vs. a square wave mess of harmonics that gets you into an FCC mess if transmitting, the local oscillator of a receiver has to be just as pure.

    This means more processing much faster or analog circuits in the middle. I’d like to test one of these on the FM band against a recent model Pioneer car radio.

  3. Ahhh, Si473X receivers again; one of my favorite subjects. Buckle up…


    @Allan-H said: “The Si4732 appears to be obsolete. Silicon Labs recently sold that division to Skyworks, who list a number of related tuner ICs on their website but not that exact part number.”

    Speculation is the Si4732 chips in the Chinese radios are fake, but I don’t know that for a fact. New homebrew builds typically use PU2CLR’s Arduino firmware for the Si473X series parts [1][2], usually the Si4735 [3] which is current, not the Si4732. The biggest problem right now is that due to the chip shortage, good luck finding a Si4735 to buy [4]. Same goes for the Si4732 [5].

    * References:

    1. PU2CLR SI4735/473X Library for Arduino [Home]:

    2. PU2CLR SI4735 Library for Arduino [Master]:

    3. Si4735 @ SKYWORKS:

    4. Si4735 Unobtanium @ Octopart:

    5. Si4732 Unobtanium @ Octopart:


    @Al Williams said: “…[OM0ET] reviews the latest one he picked up, an ATS25.”

    There is supposedly a functionally-equivalent Chinese Si4732 receiver to the ATS25 [1] receiver shown in [OM0ET]’s video review. That alternative Si4732 receiver is smaller, has a monochrome non-touch display, and it costs a lot less money [2].

    Comparing [OM0ET]’s video review of the Chinese ATS25 Si4732 receiver to the inside-view of the smaller Chinese Si4732 receiver [3] one thing jumps out at me, the PCB on the smaller receiver is carrying a standard Arduino Nano board on it [4], but the larger ATS25 Si4732 receiver does not have a standard Arduino board of any type on it.

    This poses a problem to me. If I wanted to change the firmware on the smaller receiver with my own, I can remove the Arduino Nano board with the factory firmware and preserve it while replacing it with a new Arduino Nano programmed with my own firmware. But I cannot do that with the larger ATS25 receiver. Once I program the ATS25 receiver with my own firmware, the factory firmware is erased and likely gone forever. I doubt the receiver manufacturer is going to send me a replacement .ino source code file and any attendant libraries.

    * References:

    1. ATS25 Si4732 Receiver, $106.19 Buy It Now, free shipping from China:

    2. Alternative Si4732 Receiver, $55.72 Buy It Now, free shipping from USA (Wow!):

    3. Alternative Si4732 Receiver, inside view:

    4. Arduino Nano:


    @Al Williams said: “For the price, we were both impressed with the touchscreen and build, but maybe less impressed with the RF filtering.”

    I agree with the concerns regarding lack of RF filtering in the Chinese radios. There needs to be at least some bandpass filtering ahead of the on-die LNA. I have been looking at adding the auto-switched receive bandpass filter bank used in the SoftRock Ensemble RX III direct down converting software defined radio (DCSDR) design. The schematic for the SoftRock Ensemble RX III is shown in [1]. The SoftRock SDRs are sold by Tony Parks, KB9YIG on the FiveDashInc site [2], but I do not see the Ensemble RX III on the FiveDash site yet, maybe it is too new.

    The Ensemble RX III receive filter bank has four bands:

    – Band group 0: 160m – 1.8 to 4.0 MHz (attenuated)
    – Band group 1: 80m and 40m – 4.0 to 8.0 MHz (attenuated)
    – Band group 2: 30m, 20m, and 17m – 8.0 to 16 MHz
    – Band group 3: 15m, 12m, and 10m – 16 to 30 MHz (preAmplified)

    So 1.8-30.0 MHz in four bands is still pretty wide on a per band basis, but it is far better than having no filtering at all. Narrower bands at the lower frequencies is good as that’s where most of the noise and QRM exists. As for the low-frequency attenuation and high frequency pre-amplification in the Ensemble RX III design, I will leave those out at first. The on-die LNA in the Si473X chips is variable-gain in a user-controllable AGC loop. It may be possible some additional attenuation will be needed on the lower frequencies for night operation, but I doubt another pre-amplifier will be needed through 30 MHz.

    * References:

    1. WB5RVZ.Org – Documentation, Ensemble RX III:

    2. FiveDashInc – Buy SoftRock SDRs:

        1. I can’t speak for a Baofeng but I have a $5 DVB-T dongle, a more expensive one, an upconverter and a pile of crappy old analog radios ranging from merely obsolete to antique.

          I can swap between them all on the same antennas, the more expensive sdr stick is better than the cheap one but the “junk” radios all pick up way more signals than either. I have found this to be true for HF, VHF and UHF although the lower the frequency the more true it is.

          i’m not saying a decent SDR wouldn’t be better, only that you get what you pay for.

    1. I have the monchome version, and its OK, but they all use the example firmware written by PI2CLR, which looks like it was never intended to be used in a production radio.

      I presume the manufacturer just put a SMD ATMega chip on the PCB to make things smaller and cheaper.
      I’m surprised the monochome version doesn’t do the same thing.

      The ATMega probably still has the Arduino bootloader in it, so even if the PCB doesnt have a USB to serial adaptor on it, you could still use an external USB to serial and connect to the pins on the ATMega.
      Albeit not using the soldering iron the guy uses in the video, which looks like is more suitable for plumbing than electronics.

      Also… No mention of the reverse engineering of the secret microcode taken from a Tecsun radio that allows the SSB reception, and how these chips are probably capable of doing other things, but the manufacturer doesn’t release information on how to create your own microcode.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.