That's exactly the kind of FET I meant. It can be supplied straight from a sound card or computer microphone input. Apparently it also has an antiparallel diode built in to protect the gate.
Edit: according to DigiKey it is obsolete and no longer manufactured... Replaced by an SMD version maybe?
Edit again: 2SK208-R should do the trick. It has no built-in antiparallel diode.
That could be an interesting option if the simpler ones would prove insufficient for some reason.
Last edited:
Yes very similar recording now if not even better so well done you for making the effort. I only chose 516 Hz because it is in the Audacity dropdown menu. I have no idea if there is a conventionally accepted frequency for the Morse sound?
To be fair several people have contributed on this thread and we have possibly added to a simple and easily constructed way to receive SAQ. As you live in Sweden and "talk the language" as it were, maybe you could communicate with the Alexander Association to propose this addition please?
As far as I am concerned anything I myself have posted on this thread can be published by the Alexander Association (as is or adapted for their needs) freely, unconditionally with no need to acknowledge anything. That's as long as no third party copyrights etc are infringed by so doing, also that no rules of the diyAudio forum are broken.
Oh dear Farnell UK have 5.7 million of them in stock, have a look at this link:- Farnell UK 2SK596S
On Semi still have the pdf datasheet on their website:- 2SK956S datasheet
Not to worry there are several similar JFET's listed.
Last edited:
Re: the AD8021 How you work that out? Datasheet gives "2.1 nV/√Hz input voltage noise" "2.1 pA/√Hz input current noise"
Oh put a JFET in the feedback loop as per the TI circuit of post #177 . The OPA202 TI suggest is not much faster than LM358.
Or even a very low Gate capacitance MOSFET.
Last edited:
It's a bit unfortunate they sell them in multiples of 6500. 2SK208-R is an alternative, most other JFETs have too high IDSS to connect them straight to a sound card microphone input.
Either that, or just a discrete JFET used open loop, or any reasonable audio op-amp with JFET or MOSFET input, or a triode. The difference in noise performance at frequencies close to 17.2 kHz will be quite small, but it will make some difference for the noise at frequencies far from 17.2 kHz (that you could filter off anyway).
For low noise amplification one has to eliminate the noise contributed by (feedback) resistors. IOW instead of resistors one has to apply transformers. In RF design this is well known, for instance: https://www.semanticscholar.org/pap...Peng/2019391fa8f405c280f5d38be7f4b48f382f22e9
However this concerns a low impedance source. For high impedance (mos)fet gate input, feedback has to be applied to the source via a transformer. When done properly, it also decreases the effective input capacitance. The crucial part is the transformer.
However this concerns a low impedance source. For high impedance (mos)fet gate input, feedback has to be applied to the source via a transformer. When done properly, it also decreases the effective input capacitance. The crucial part is the transformer.
Conceptually transformers are better, but you can also keep the noise of the feedback network negligible by using sufficiently low resistor values for series feedback or sufficiently high resistor values for shunt feedback, or use capacitive feedback.
RF ambient noise at low frequencies is really high, see this link:
https://www.sciencedirect.com/topics/computer-science/frequency-noise
There is no point obsessing with low noise receivers unless your antenna is tiny and unsuitable.
There has even been a suggestion to harvest this noise to power wearables.
https://www.sciencedirect.com/topics/computer-science/frequency-noise
There is no point obsessing with low noise receivers unless your antenna is tiny and unsuitable.
There has even been a suggestion to harvest this noise to power wearables.
Thank you for your contributions to this thread, they helped confirm some things for me (and most likely others also).
Noise can be cancelled with a loop when arriving from a single direction. If that direction almost coincides with the direction of the wanted source, the induced voltage can be low enough to apply a low noise 1st stage. This has been the classical case for so called DXers for decades.
Although that's nulling, rather than cancellation to be pedantic, since the noise source doesn't link to the loop at the nulling point there's nothing to cancel. With two or more sources of noise you can't null them all, but you can eliminate the worst one.
Can nulling work at this wavelength?
I read one report using a VLF receiver that detected the signal from a lightning strike four times, each way around the world and a second lap both ways. The propagation delays were massive, more than 0.1 seconds
I read one report using a VLF receiver that detected the signal from a lightning strike four times, each way around the world and a second lap both ways. The propagation delays were massive, more than 0.1 seconds
Indeed, nulling only works for a single source and it's why loop antennas are used despite the low efficiency, especially when untuned. One can see Q as amplification factor but a tuned circuit, excited by a pulse signal, stretches it and more so when Q is higher. So in a high noise environment like the tropics it's best to avoid that.
With legacy radio receivers in CW mode the resulting audio frequency is easily adjusted with the tuning knob, so it is up to the listener, SDR applications seem to produce output at 700 Hz or thereabout. But maybe there is some frequency that is optimal in demanding situations?
I don't know how Audacity (and other software) does the pitch conversion but its a clever bit of maths. I tried dividing 17200 by 516 - all I can say is that an exact integer number of cycles at 17200 fits into three cycles at 516!
I'm guessing that Morse reading ability went with musical ability so maybe an experienced Morse operator would choose a musical note he or she liked?
C 261.63 C 523.25
C# 277.18 C# 554.37
D 293.66 D 587.33
D# 311.13 D# 622.25
E 329.63 E 659.26
F 349.23 F 698.46
F# 369.99 F# 739.99
G 392 G 783.99
G# 415.3 G# 830.61
A 440 A 880
A# 466.16 A# 932.33
B 493.88 B 987.77
Maybe someone else knows if there were conventionally used Morse code reading frequencies, or frequencies optimal for demanding situations?
I'm guessing that Morse reading ability went with musical ability so maybe an experienced Morse operator would choose a musical note he or she liked?
C 261.63 C 523.25
C# 277.18 C# 554.37
D 293.66 D 587.33
D# 311.13 D# 622.25
E 329.63 E 659.26
F 349.23 F 698.46
F# 369.99 F# 739.99
G 392 G 783.99
G# 415.3 G# 830.61
A 440 A 880
A# 466.16 A# 932.33
B 493.88 B 987.77
Maybe someone else knows if there were conventionally used Morse code reading frequencies, or frequencies optimal for demanding situations?
Last edited:
The issue of optimum frequencies was discussed by hams, and they too more or less suggested what you did.
I am guessing that the carrier freq does not affect the frequency of the chirps we hear during playback. The chirps represent the rate of rise and fall of signal intensity. I don't know exactly how that is possible. A 30khz carrier would produce the same sound if it had the same power envelope.