Why so much noise on 16 MHz

Hello,

I have made 100 nanowatt cw breadboard transmitter on 16mhz "computer" crystal.

The signal is picked up by dc breadboard receiver also based on comuter 16mhz crystal.

I can hear the signal at distances of about tens of meters, but when I am trying bigger distances combined with increased sensitivities, I am getting very strong noise.

I tried to find origins of the noise by putting dummy load (50-500 ohms resistors) between ground and receiver antenna entry. Dummy load does not produce strong noise. So my conclusion is that origin of this noise is either

1) noise in the 16MHz band, or

2) interference between antenna and receiver (the longer is antenna, the bigger is the noise).

Which one of two could be correct cause?

Should I connect antenna through long coaxial cable to get the answer?

Thank you in advance
Truffaldino

Probably both.

If there is no noise around 16MHz it can't be picked up anyway but the larger antenna is likely to pick up more signal across all frequencies and so feed more to your receiver.

Please show the schematic of your receiver, it sounds like it has poor selectivity and may be prone to other frequencies breaking through. Moving the receive antenna further away and using a screened feeder might help but only if the interference comes from a local source and you are moving further away from it. The co-ax itself will not improve matters and bear in mind that you could be moving the antenna CLOSER to the source of interference and making matters worse!

Brian.

Helo Brian,

Here is schematic and picture of receiver

Breadboard is a worst choice for RF or High Frequency circuits.Try to build up your circuit on a -at least- single sided PCB.

I have already did it.

It does not make much difference in my case. So I prefer breadbord.

Sounds like you have already proven that the noise is
coming in on the antenna. Now the question is, whether
this is ambient noise or self-induced. How about box-
shielding the receiver as another "diagnostic cut" to
prove / disprove local coupling?

I don't see anything like a spectrum analyzer plot from
the antenna, to say just how filthy the airwaves are.
Or whether this "noise" is a locally generated interferer.
If you had such equipment, and could "sniff" the antenna
port (splitter?) with batteries in and out, that might also
be useful insight.

Hi Dick,

It have put the breadboards in the metal candy box shown on the picture above. It does not reduce the noise. Perhaps this kind of box is not sufficient for proper shielding? Also I have tried it in different places (several kilometers away). The result is always the same.

So to recap:
Longer antenna makes more noise (amplitude)?
Shielded box does not reduce noise
This implies external pickup.

I think an independent means of measuring in-band noise
from the antenna would be good. Might also be good to,
rather than "long" and "short", try a wavelength-matched
antenna for a bit more front end selectivity and gain?

But still you have not said anything about the characteristics
of the "noise" (perhaps you do not have the capability to
measure).

Breadboarding at 16MHz is always going to give problems but I suspect the main one is the input filter network. You can't treat a quartz crystal as a series pass element unless it is carefully impedance matched at both sides. You should at least try to make the circuit symmetrical by adding a tuned circuit at the antenna side as well.

It's worth noting that the capacitance between strips on the breadboard is probably enough to let most signals through unhindered anyway.

Your other problem is the mixer stage, if I understand what you are doing, the crystal there is either 15.999MHz or 16.001MHZ (possibly a 'pulled' 16MHz crystal?) which is so close to the input that much of the oscillation will be picked up by the input stage.

The NE592 is running with a voltage gain of less than unity so it serves no useful purpose. You need to add a resistor across the gain setting pins for it to amplify. You can also make a huge improvement to the input noise level by modifying it's input bias circuit. At the moment, you are differentially amplifying (if you add the gain resistor!) the supply line and ground noise. If you join the two 4.7K resistors to make a potential divider, maybe adding a 10nF capacitor at their junction to ground, then connect it to a center tap on the input coil it will null the supply noise instead of amplifying it.

Brian.