433.92 Noise Is Different Across The Same Type Of Devices

Question

I have two 433.92 RF receivers of type XY-MK-5V and one of type RXB12 . The question that brings we here is this:

Lets give the receivers names:

XY-MK-5V is Y as yellow
XY-MK-5V is O as orange
RXB12 is B as blue

When I connect them to an oscilloscope, B and Y behave in the same way, constantly giving the background noise and when I try sending a signal from a remote control, t is not so clear when the signal begins but I can distinguish it from the noise. So far, so good.

When I connect O, there is almost no background noise and when I try sending a signal from a remote control I can clearly see it. There is almost no noise at all.

So, could someone help me with understanding why O behaves like it has a built-in nose filter. I like and would like to have this type of hardware filter but since only 1 of 3 has it and this one from the group of two and another one does not have this filter, it sounds like there is a problem somewhere. Both were bought on the same day from the same supplier and came in the same package.

Here is the same thing in pictures. 3 readings are taken from 3 receivers via Blue, Yellow and Orange cables:

I am connecting 3 receivers to the oscilloscope, start recording, then press a button on the receiver and stopping the recording.

So, why does the Orange filter out the noise?

Thanks.

edit:

XY-MK-5V Documentation URL (Technical Parameters) Document

document content:

Technical Parameters

Operating voltage (V): DC5V
Quiescent Current (mA): 4MA
Modulation: AM (OOK)
Operating temperature: -10 ? ~ 70 ?
Receiver sensitivity (dBm):-105DB
Operating frequency (MHz): 315,433.92 MHz (266-433MHZ frequency band can
be optional)
Dimensions (LWH): 30 * 14 * 7mm

Requirements such as the pro distance farther, then 1/4 wavelength antenna,
generally use 50 ohm single-core wire, the length of the antenna 315M of
approximately 23cm, 433M is about 17cm;

Reception antenna location on the module also affects the installation,
the antenna as far as possible straight away from the shield, high pressure,
and interference source place; 

receiving frequency used, decoding and
oscillation resistor should match with the launch
Pin and instructions for use:

The ANT connected antenna side
2 VCC power supply positive
3,4 the DATA data output
5 GND negative power
The receiver module has four external interfaces, VCC positive power
supply, "DATA" Output "GND" indicates the power supply negative (products
labeled in English).

Uses: remote control switch, receiver module, motorcycle, automobile
anti-theft products, household anti-theft products, electric doors,
shutter doors, windows, remote control socket, remote control the LED,
remote control stereo, remote control electric gate, garage door remote
control, remote control retractable doors, remote control volume gate ,
sliding door, remote control door opener, door closing device control
system, remote control curtains, alarm system, alarm, remote control
motorcycle, remote control electric cars, remote control MP3, remote
lights, remote control cars, security and other remote areas of civil and
industrial facilities

Quality characteristics:

1, the receiver module and fixed code on the market, the same frequency
remote control to learn the code any ancillary use, can be equipped with
all wireless remote control of the shop.

2, the super-regenerative receiver module LC oscillator circuit
containing zoom in shaping the output data signal is TTL level and can
be directly to the decoder is extremely easy to use, and inexpensive, so
widely used. Small size, high sensitivity, easy debugging; frequency,
short lead times; consistency of product quality, cost-effective, the
current is the amount of wireless remote control market is the largest
and most widely used form of high-frequency receiver module. Receiver
module has a wide receiver bandwidth, typically ± 10MHz, factory general
tone in the 315MHz or 433.92MHZ (if there are special requirements for
adjustable frequency, adjustment of the frequency range 266MHz ~ 433MHz.)
Receiver module with DC5V power supply generally, any special
requirements to adjust the voltage, the voltage adjustment range of 3 ~
8V.

For 4mA quiescent operating current of the receiver module factory
generally, if there are special requirements can reduce the current,
minimum current is adjustable to 1.5mA, but the receiver sensitivity will
be reduced. Factory receiver module output noise output, if any special
requirements can be changed without the noise output, but will reduce the
receiver sensitivity.

Answer 1

The RXB12, due to the presence of a crystal oscillator, is necessarily a superheterodyne receiver. The XY-MK-5V, on the other hand, can be surmised to be a super regenerative receiver. A quick guess based simply on the lack of any crystal is really all you needed, but the bare inductor coil and tunable component are additional dead giveaways. And I have several XY-MK-5V modules (and none of them have quite the same PCB layout, hilariously!), and I manually pieced together the circuit, and can personally confirm that they are definitely all some minor variation of the same super regenerative receiver circuit.

It is not that the RXB12 is low noise per se. And the XY-MK-5V is not high noise. The signal looks nice and clean to me, I see no noise on the actual signal, just nice sharp square waves right where you'd expect them to be. What you're seeing is not really noise, at least, in the sense of noise within a communication channel. Super regenerative receivers are very capable and indeed, all 3 receivers receive the intended signal loud and clear.

However, super regenerative receivers have a fairly undesirable property of producing something called 'no signal noise'. Basically, the output will just send out garbage until a real signal is received. Super regenerative receivers work by having a self-quenching oscillator with positive feedback (usually taken from the output) fed back into it, allowing a huge amount of gain with a single stage. The oscillator restarts after quenching, and if it self-quenches, you end up just amplifying noise instead. If there is no signal, then the oscillator is always being forced to quench and restart, so you get that characteristic nonsense on the output. Once a signal is actually being received though, the oscillator is stabilized, and assuming the signal changes power states faster than the maximum run time of the oscillator (in other words, before it self-quenches), then you get a clean output signal. So that means that you can't modulate DC, and indeed, need to transmit a signal with some minimum rate of state change or these types of receivers will return to their default garbage output mode.

Super regenerative receivers are very old, and weren't really ever used because they're pretty crappy and can radiate noise when there is no signal, but very recently, they are seeing use in very low power and low data content applications, like these receivers, or garage door openers, that kind of thing. They are extremely low cost, in fact, they're pretty much the cheapest possible way you can build a useful radio receiver.

It's the sort of circuit that works, and is cheap, but no one would describe as 'good'. They are also extremely sensitive to things like the angle something is bent at, the direction your tongue is pointing, that sort of thing. There are actually stories of car wireless FOB receiver modules (the part in the car, not the thing on the keychain) where the manufacturer had to include factory jigs for bending some little TO-92 transistor to just the right angle as part of the manufacturing process, because that TO-92 transistor actually had to be soldered 9mm above the board and bent at a 17° angle (or whatever), or the super regen circuit wouldn't oscillate. But they were probably making a few million of these, and that will amplify small differences in cost into something meaningful very quickly - hence the reason they were using such a delightfully awful circuit to begin with. Don't get me wrong, super regens are kind of ingenious, but in the same way cup of noodles is ingenious.

The RXB12, being a superheterodyne, has no such no signal noise property. It works by mixing the RF signal with a locally generated frequency (graciously provided by the onboard crystal) to produce a beat frequency called the intermediate frequency, which in the case of ASK, will simply be something saying 'is there meaningful amounts of power right now at 433.92MHz (or whatever we're mixing in)'. So when there isn't any meaningful power, you just get a nice constant 'low' on the output. No signal looks like no signal, in other words. There is also no minimum modulation requirement, so theoretically you could modulate DC and receive it correctly (for example, if you wanted to simply emulate a button press but over RF). But don't quote me on that - I have none of those modules, so that might not be correct. But I would be surprised if it was not correct.

Superheterodyne receivers also offer better sensitivity, selectivity, lower noise, etc. The only downside is cost. Historically, a superheterodyne receiver would be expected to cost 2-3 times (at least) more than the same receiver implemented using super regen topology.

So, the extreme simplified answer to your question is, simply, "because the RXB12 is better."

You may be thinking, "but it costs 89 cents shipped on ebay, I thought it was supposed to be more expensive? And why wasn't it the norm to begin with?".

You can thank large scale integration for this. Someone ran some numbers, and determined that it was worth the investment to actually make an integrated circuit version of superheterodyne receiver for 433.92MHz ASK applications. Integrated circuits are literally made out of dirt (well, sand/silicon), and as you know, dirt is basically free. I have access to large amounts of no cost dirt at this very moment.

I jest, but basically, when the market conditions are right, an integrated circuit can take something that normally would have been more expensive, turn it into a single die, produce a few million of them, and now suddenly the better, more robust option is actually cheaper than some other crappier option that was done before. In fact, that's kind of the point of integrated circuits. Yay technology!

Anyway, whoever is making those ICs is hopefully making a killing, because I'm going to go buy a bucket of these guys right now. The super regen receivers were always a pain to use, because you'd constantly be having to try and sync with a real signal vs. noise, and would randomly miss the first byte so you'd have to allow for this in your transmitting device. Its not really that big of a deal, but it just adds up to little gotchyas that make things more complex and pretty much require a micro controller on both ends, or at least those Holtek modulator chips (which work great, and based on the price, Holtek knows it).

These RXB12 modules are much easier to work with, and I would guess have superior range and performance, although probably not by a huge margin or anything.