Will they work? These are the little dongle devices with 2 jacks for Mic and Earphone.

Yes, they will work. But you do get what you pay for. I use a Tascam US-100 for audio and it definitely performs better than generic PC audio. For most people it probably doesn't matter...

73

-Jim
NU0C

On Mon, 13 Mar 2023 20:48:32 -0700
"Bill" <bill333@...> wrote:

Will they work? These are the little dongle devices with 2 jacks for Mic and Earphone.

Yes. They work fine

Yes, it works. WSJT-x does not need a HiFi device...

Pietro I2OIM



Il giorno mar 14 mar 2023 alle ore 04:48 Bill <bill333@...> ha
scritto:

Mornin' Bill:

Regarding the concern about hi-fi adapters, I have used, extensively, for almost 2 years, 2 SABRENT Aluminum USB External Stereo Sound Adapters, from Amazon. They have worked flawlessly, with 2 ancient Buck Rogers Rascals, with an ancient TS-790A on vhf and uhf, and an ancient FT-990 on all hf bands.

I can only assume that if I did not configure "stuff" correctly, that I might have justifiable complaints.

AS always, YMMV.

Pete/wa4hei

Yes, and I’ve found it works well. I use it with my FT-891 along with a Cat
cable on my Dell laptop. Makes for a portable operating system for POTA as
well. For this set up, I’ve found this YouTube vid helpful:
https://youtu.be/XTPbE--_Dw0
Go for it.



On Mon, Mar 13, 2023 at 23:48 Bill <bill333@...> wrote:

Will they work? These are the little dongle devices with 2 jacks for Mic
and Earphone.





--

Randy
KO4WTM <https://www.qrz.com/db/KO4WTM>
West Palm Beach Amateur Radio Group http://wpbarg.com/
Jupiter Tequesta Repeater Group https://jtrg.org/
Straight Key Century Club #26062
Long Island CW Club #3482
roaringkelly.com

It depends on what your station is and what your goals are.

If you are going out into the field, using modest antennas, and are just happy to copy what you can and work that, they are a great solution.

If you're primarily operating EME where the signals are teeny weeny, they do a fine job.

But, if you have large antennas scooping up lots of loud signals and you want to copy the weaker ones, you may need something better in terms of distortion performance. I've decoded as many as 97 stations recently when 10M was open. Not all were blasting in, but certainly many were pretty loud. That was just with a dipole, too. The math for the combined signal power in the audio band for all those combined signals is pretty straightforward.

K9YC is a (mostly retired) professional in this field and he wrote up a summary of what he learned in his own station: http://k9yc.com/USB_Interfaces.pdf

FWIW, I bought a higher quality ADC/DAC for my own station and found the improvement was well worth it. It was still much less than $100 (can't remember what I paid).

Very interesting paper. My question is what about radios that have built in soundcards. My TS590SG seems to work well on FT8 and I often see decodes indicates as -20 or even -24 so that would seem to agree with your paper. It is so much easier to use the hardware of the radio. I had use SIGNALINK and have to agree with you that it is far from a great device but at the time I did nor have a radio with Sound Card.

Outlook LT Gil W0MN
Hierro Candente Batir de Repente
44.08226 N 92.51265 W EN34rb

FWIW, I bought a higher quality ADC/DAC for my own station and found the improvement was well worth it. It was still much less than $100 (can't remember what I paid).

Since WSJTX limits its internal audio handling to 16 bits, a 24 or 32
bit sound card add nothing (the additional bits are reduced before the
audio samples even reach WSJTX). The real key is the care in which the
sound card designer used in filtering/bypassing the power supply,
decoupling the reference voltage from any digital signals, and the
quality/stability of the reference. Each of those items add low level
noise in the sound card which reduces the *available* dynamic range of
the sound card.

A good (bad?) example of failure to address these filtering/bypassing
issues is the earliest models of the Signalink. The first units had
no filtering on the USB power line. The CODEC reference was a simple
voltage divider from USB power with no bypassing, no filtering and no
regulation. In addition, many of the digital logic ICs have no
bypassing which results in digital noise coupling to the audio circuits
via the power supply (USB power).

Unfortunately, it is nearly impossible to determine in advance whether
a particular $10 USB sound device has the needed filtering and
regulation!

73,

... Joe, W4TV


On 3/14/2023 4:33 PM, The Greene Family wrote:

It depends on what your station is and what your goals are.
If you are going out into the field, using modest antennas, and are just happy to copy what you can and work that, they are a great solution.
If you're primarily operating EME where the signals are teeny weeny, they do a fine job.
But, if you have large antennas scooping up lots of loud signals and you want to copy the weaker ones, you may need something better in terms of distortion performance. I've decoded as many as 97 stations recently when 10M was open. Not all were blasting in, but certainly many were pretty loud. That was just with a dipole, too. The math for the combined signal power in the audio band for all those combined signals is pretty straightforward.
K9YC is a (mostly retired) professional in this field and he wrote up a summary of what he learned in his own station: http://k9yc.com/USB_Interfaces.pdf
FWIW, I bought a higher quality ADC/DAC for my own station and found the improvement was well worth it. It was still much less than $100 (can't remember what I paid).

That's were I got the tip to look for a Tascam US-100. They can be found on the auction site for reasonable money with a little patience. I liked it so much I have three of them now.

73

-Jim
NU0C

On Tue, 14 Mar 2023 13:33:07 -0700
"The Greene Family" <cvgreene@...> wrote:

K9YC is a (mostly retired) professional in this field and he wrote up a summary of what he learned in his own station: http://k9yc.com/USB_Interfaces.pdf

I suggest using one that someone has found good.
I had very good resuklt with this one https://www.amazon.it/gp/product/B01J3QGU50/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
I think you can find it in your national Amazon store, too

Pietro I2OIM

I follow this discussion with interest. As important facts I noted:

Am 14.03.2023 um 22:08 schrieb Joe Subich, W4TV:

Since WSJTX limits its internal audio handling to 16 bits, a 24 or 32
bit sound card add nothing (the additional bits are reduced before the
audio samples even reach WSJTX).  ....

and the reqirement of 48 Hz as setting for the sound routing that K9YC
mentions in his PDF document.

Regarding the impact of

... filtering/bypassing the power supply, decoupling the reference
voltage from any digital signals, and the quality/stability of the
reference.

can I assume that my setup is fine in this regard when I observe signals
with a SNR of -25 dB sometimes?

And the last question: Another decoding software claims "improvements of
FT8/4 decoder for new internal data transfer, 24-bit (raw data)". I
don't want to discuss that software in the WSJT-X list but I am curious
what others think about a 24 bit routing from a technical perspective.

vy 73,
Willi, DJ6JZ

https://www.presonus.com/learn/technical-articles/sample-rate-and-bit-depth

/*
The most important practical effect of bit depth is that it determines the dynamic range of the signal. In theory, 24-bit digital audio has a maximum dynamic range of 144 dB, compared to 96 dB for 16-bit but today’s digital audio converter technology cannot come close to that upper limit. As of this writing, the 24-bit Burr-Brown converters in StudioLive RM-series rackmount mixers ( http://www.presonus.com/products/studiolive-ai-rack-mixers ) , and Studio 192-series audio interfaces ( http://www.presonus.com/products/studio-192-usb-30-interfaces ) offer a dynamic range of 118 dB. The high-speed converters in Quantum-series audio interfaces ( https://www.presonus.com/products/Quantum-Thunderbolt-2-Interfaces ) provide a dynamic range of 120 dB.
*/

So my guess is that it’s not going to make a lot of difference, so you could probably pull more signals out if your audio game was turned way way way down. 🤷‍♂️

73, Willie N1JBJ

This one works well too: https://www.amazon.it/Sabrent-Esterno-Adattatore-necessit%C3%A0-AU-MMSA/dp/B00IRVQ0F8/ref=sr_1_20?__mk_it_IT=%C3%85M%C3%85%C5%BD%C3%95%C3%91&crid=3T4258C8O4M5K&keywords=sabrent+usb&qid=1678894735&s=pc&sprefix=sobrent+usb%2Ccomputers%2C238&sr=1-20

It's on Amazon everywhere.

Thanks Fred I have one on order

24-bit gives more dynamic range of about 138.5dB vs  90.3 dB for 16-bit -- but not significantly more impact on SNR that one could notice or measure.
Prior testing has shown 30dB of signal overcomes quantization error and is the absolute minimum one should have on the signal meter in WSJT-X.We have found running WSJT-X at around 70dB (on a quiet spot) with slow AGC seems to work really well.  I suspect the signal subtraction might prefer the higher signal level since subtracting a signal from a low-level signal would leave a really-low-level signal.That is now the new recommendation in our paper on setting audio levels for WSJT-X now.
Mike W9MDB

Actually that article is wrong for sinusoidal signals oscillating around zero as 1 bit is used for the sign.

So 16-bit is actually 90.3dB maximum and 24 bit 138.5dB.  

I don't' know any rig that samples the output and provides a full 16 or 24-bit digitized result.  They all internally convert to a signed integer to put out a digital result on the USB serial port.

Mike W9MDB

can I assume that my setup is fine in this regard when I observe
signals with a SNR of -25 dB sometimes?

Not necessarily as the SNR in WSJTX is heavily influenced by the
level of antenna noise. A reported -25 dB is relative to the
noise in the DSP passband thus one would be more likely to see
very low SNR reports with higher total system noise.

And the last question: Another decoding software claims "improvements
of FT8/4 decoder for new internal data transfer, 24-bit (raw data)".
I don't want to discuss that software in the WSJT-X list but I am
curious what others think about a 24 bit routing from a technical
perspective.

Total BS in the real world. A well designed 16 bit sound card will
have an effective dynamic range (internal noise floor to clip) within
10 dB (or better) of the theoretical 96 dB (20 Log(2^16)). Professional
24 bit sound cards are only generating 118-120 dB dynamic range in the
labs.

Still, the best amateur receivers are only generating 100 dB (or less)
of narrow spaced dynamic range with internal noise floors around -140
dBm. With "antenna noise" in urban/suburban areas that runs more than
-120 dBm, even the best receivers in "quiet" locations can't take
advantage of more than 80 to 90 dB of dynamic range. With -120 dBm
antenna noise, -25 dB SNR, 90 dB Dynamic range, that works out to:
-120 -25 +90 = -55dBm (or clipping level of S9+20 dBm).

73,

... Joe, W4TV

On 3/15/2023 7:47 AM, Willi Passmann wrote:

I follow this discussion with interest. As important facts I noted:
Am 14.03.2023 um 22:08 schrieb Joe Subich, W4TV:

Since WSJTX limits its internal audio handling to 16 bits, a 24 or 32
bit sound card add nothing (the additional bits are reduced before the
audio samples even reach WSJTX).  ....

and the reqirement of 48 Hz as setting for the sound routing that K9YC
mentions in his PDF document.
Regarding the impact of

... filtering/bypassing the power supply, decoupling the reference
voltage from any digital signals, and the quality/stability of the
reference.

can I assume that my setup is fine in this regard when I observe signals
with a SNR of -25 dB sometimes?
And the last question: Another decoding software claims "improvements of
FT8/4 decoder for new internal data transfer, 24-bit (raw data)". I
don't want to discuss that software in the WSJT-X list but I am curious
what others think about a 24 bit routing from a technical perspective.
vy 73,
Willi, DJ6JZ

That one does not have a volume adjustment like the other one.

It's important to adjust signal level in the analog space rather that in the digital space inside the computer.
So preferable to have that volume control on the device which I assume is working in analog space (since digital space would be more expensive than a simple rheostat)

Mike W9MDB

Many (most? all?) contemporary radios digitize the incoming analog signal at some point in the processing chain. It could be directly at the incoming frequency, at some intermediate frequency, or at audio frequencies. This allows for all those DSP functions that provide filtering and lots of cool marketing features.

What I do not get is why more radios don't just offer up that digitized signal on a USB interface, after processing. Yeah, there might need to be a sample rate conversion to get to 48 KHz suitable for WSJT-X, but that's pretty standard code these days. (You'd need to do the reverse for the transmit functions.) This would eliminate multiple conversions from and to the analog domain, which has to be a good thing. Operationally, it would also be great. Far less opportunities for dirty transmit signals, less chance for RFI problems, and so on.

It's also worth considering how these radios perform when decoding crowded bands. That 96 or so dB of dynamic range for an ideal 16 bit converter assumes a single signal within the band. As somebody sorta said a while back, Mo Signals, Mo Problems. If you're talking about weak signal VHF operation, you have one set of conditions mostly limited by noise. On crowded HF bands, the external noise is still a limiting factor, but IMD within the FT8 bands generated in the receiver may really be the limiting factor. Plus, the usual dirty transmitters and so on.

Clarke K1JX