Audio recovery

I4LEC
Posts: 28
Joined: Sun Apr 09, 2017 2:58 pm

Audio recovery

Postby I4LEC » Fri Jul 16, 2021 2:30 pm

The recent tested K4D, highlighted its excellent audio recovery, I noticed a huge dynamic range from the base band noise and the incoming signals.
While on a clear frequency, no hiss could be heard, once tuned on a even weak signal, this comes out very loud, apparently from no where, listening is very pleasant.

I brought the subject up, as I felt my ANAN, expecially on the low bands, lacks a bit on this regard.

The first noticeable difference between the two is the ADC driver, the ANAN radios have the LNA always ON (gain slightly reduced on the Orion II boards), the ATT can be inserted at 1dB step, to lower the gain, if I recall correctly, Phil mentioned about choosing this solution as the preferred one, do not know if he specified why?

Given the K4 MDS at -121dBm (500 Hz), I tried to reduce the sensitivity of my ANAN to the same level (about 12dB), given my rural bands noise, suitable from 17m down, above it, those 12dB were deem necessary, matter of fact, K4 Pre1 (12dB) had to be turned ON as well.

Even at equal gain level, I could not replicate the no signal, signal enhanced effect, I thus diverted my attention to the AGC, which I thought could be the culprit, we all know the importance of adjusting AGC threshold along with AF volume for the best signal to noise ratio, tweaking it helped a bit, I also tried to change the AGC slope, without really being able to reach the same result.

How could eventually the audio recovery be improved? Any chance from developers to eventually have a look at the code?

73, Clay
User avatar
w-u-2-o
Posts: 5539
Joined: Fri Mar 10, 2017 1:47 pm

Re: Audio recovery

Postby w-u-2-o » Sat Jul 17, 2021 2:17 pm

Clay,

Sorry I didn't reply to this earlier, I was not home all yesterday.

Also, I only know two languages: English and Bad English. So you are going to have to carry the load there :)

I'm not entirely sure what you mean by "audio recovery", but I'm going to assume you mean something like "audio quality" or "audio SNR".

Comparing audio between two receivers in a fair and objective way is often a difficult task.

The LNA in the openHPSDR architecture design is a fixed gain stage required by the ADC. You will find many such fixed gain stages in other receiver designs, including the K4 and K3. These fixed gain stages may be for an IF (such as in the K3) or for the ADC (such as in any ANAN series hardware and undoubtedly in the K4 as well). Their purpose is not to provide an adjustable, switchable additional gain stage. Indeed, in any well designed, modern, HF (<30MHz RF) SDR direct sampling receiver, providing a switchable gain stage is generally unnecessary as any competent design will provide more dynamic range and IIP performance than necessary. Why Elecraft thinks it is necessary in the K4 is a mystery. However there is much that is unknown about the K4 design, I have not seen so much as a block diagram myself.

It is good that you are adjusting the key parameters of input attenuation, RF gain (so-called AGC gain in openHPSDR), and other AGC parameters in order to try to create an equivalent test. However this can be difficult because many of these controls are uncalibrated and thus you wind up doing this by ear. One thing that should be relatively easy to set is input attenuation. Simply match the two radios. If Elecraft only provides the traditional 10 and 20dB attenuation, then you are forced to follow suit with the ANAN.

Are both receivers receiving the same signals from the same antenna simultaneously, thereby minimizing any differences caused by changing band conditions and any subjective memory effects?

Are you also listening in matched passbands? E.g. 100 to 2900Hz? With matched passband slopes?

Are you also listening using the same speaker or headphones and at the exact same volume level?

It would be better if you did not perform the comparison by ear. Capturing the audio output via a digital audio connection to the receiver, making a digital recording of that output, and analyzing it using an audio spectrum analyzer would allow a more objective measurements and a better understanding of the differences.

73,

Scott
I4LEC
Posts: 28
Joined: Sun Apr 09, 2017 2:58 pm

Re: Audio recovery

Postby I4LEC » Sun Jul 18, 2021 5:21 pm

Hi Scott,

thanks for the reply, no hurry, I was nor even sure to bring the subject up.

As to the term “audio recovery” this came from down under, when a friend, VK4SX told me about the differences he noticed while using several different DS radios, it is not audio quality, which is excellent on both radios.

Frankly, I wasn’t really sure what Dennis meant until I had the K4 on my hands, “audio SNR” is probably more appropriate, as the effect is almost like having the SQL ON when on a clear frequency, the ANAN is very quiet too, but does not seems to have the same high dynamic when a signal comes into the passband, as well as leaving a little hiss on the background when without a signal, unfortunately my old HP 3400a failed on me after few seconds from power-up (needle went negative), otherwise, I would have attempted to make some measurements.

As to comparing audio, I agree with you, I didn't mean to compare anything, neither I had both radios side by side, the perception was so neat simply by tuning the radio, when I went back to the ANAN, I knew exactly what I was looking for, tweaking improved a bit but not as much as clearing the little background noise, totally absent on the other one, and yes, the headphone were the same, as well as the BW upper/lower cuts, cannot tell about the slopes.

As a side note, the effect, if any, was much less noticeable on CW, the narrow BW probably mitigated the difference.

Given the MDS at -121dBm (500 Hz) I wrongly thought it could be reached without any fix gain, few dB still be necessary to get there, their Pre1 gains about 12dB, and Pre2 about 5 more, ATT can be selected at 3dB steps up to 18dB, my ANAN 10, with its fix gain (which should be 20dB) always on, has a measured MDS of -134dBm on 20m (-133dBm on the other bands).

73, Clay
User avatar
I2NOY
Posts: 89
Joined: Thu Mar 25, 2021 10:44 am
Location: Milano - Italy

Re: Audio recovery

Postby I2NOY » Mon Jul 19, 2021 8:48 am

Dears, I understood the same of Clay from replies of K4 designers in some email groups.
They say that between passive front-end filters and ADC there is not an active amplifier always ON, like in IC-7610 and Anan 7000.
The MDS is enough in lower HF bands and you can switch ON the preamplifiers, if you need to improve MDS. This is what designers say.
On last firmware they develop an algorithm to switch automatically ON and OFF the preamplifiers, depending on band you are using.

This is my understanding. ;)

Return to “Everything Else: Antennas, Relays, Switches, Power, Grounding, Cooling, etc.”