Hi- I was able to Mcgyver something together and now I’m getting passable SSB audio. The mic gnd on the audio hat is now the ground return for the line out l,r and mic in. PTT also grounds through the same point. With mic gain at 0 setting, I can hear myself through a SDRPlay attached with about 70 db attenuation.?

There is a decaying echo with the audio- One (one one one) with about a second between echos, 3 or 4 as it fades. Above gain 10, things get wonky.I suspect??the mic comp code for both symptoms, and returning the mic and line grounds to mic gnd on the hat fixed the noise with no mic input but gained up.?

Larry
KB3CUF

Hi Ciccio,
Try this:

On 2023-09-07 15:48, K9HZ wrote:

The attenuator board has 64 dB of attenuation in 0.5 dB steps. That
will get you from milliwatts to one-hundred watts with my PA modules.
It's I2C so really easy to incorporate.

*** Is it 3.3V compatible? I know that I2C is open-collector, but it
does have those pullup resistors...

- Jerry, KF6VB

Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ
PJ2/K9HZ VP2EHZ
Owner - Operator
Big Signal Ranch – K9ZC
Staunton, Illinois
Owner – Operator
Villa Grand Piton – J68HZ
Soufriere, St. Lucia W.I.
Rent it: www.VillaGrandPiton.com
Moderator: North American QRO Group at Groups.IO.
Moderator: Amateur Radio Builders Group at Groups.IO.
email: bill@...
-----Original Message-----
From: [email protected]
<[email protected]> On Behalf Of jerry-KF6VB
Sent: Thursday, September 7, 2023 5:11 PM
To: [email protected]
Cc: K9HZ <bill@...>
Subject: Re: [SoftwareControlledHamRadio] T41 Transmitting (noise in ssb)
On 2023-09-07 14:56, K9HZ wrote:

That’s why I build the digital attenuator board to use with the PAs…
Use the full range of the A/D then… and attenuate the power where it
should be attenuated.

*** Definitely the way to do it. Every last bit of A/D resolution is precious.
For my own use case, I only need 10W. Because my linear will turn
that into 800W. I do need to be able to output 15W or so just for
tuning. I have an MFJ998 remote autotuner at the base of my antenna,
and it takes 15W to urge it to tune. Hitting it with the full 800W
when it's not properly tuned can fry it.
Setting the T41 to a quarter watt for tune should get me a
reasonable output out of the linear.
So - the two power levels I need are 10W and 1/4W.
- Jerry, KF6VB

DR. WILLIAM J. SCHMIDT - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ
PJ2/K9HZ VP2EHZ
Owner - Operator
Big Signal Ranch – K9ZC
Staunton, Illinois
Owner – Operator
Villa Grand Piton – J68HZ
Soufriere, St. Lucia W.I.
Rent it: www.VillaGrandPiton.com [1]
Moderator: North American QRO Group at Groups.IO.
Moderator: Amateur Radio Builders Group at Groups.IO.
email: bill@...
FROM: [email protected]
<[email protected]> ON BEHALF OF Greg KF5N
SENT: Thursday, September 7, 2023 11:00 AM
TO: [email protected]
SUBJECT: Re: [SoftwareControlledHamRadio] T41 Transmitting (noise in
ssb)
It will take some calculating to discover the optimum amplitudes out
of the Audio Hat.
Voice peaks need to be accounted for. So you can't drive too hard or
the peaks will get clipped.
The T41 power control is done by varying the I and Q amplitudes. So
that is another consideration.
If you want power varied from 1 watt to 20 watts that soaks up a lot
of the DAC range.
And then you have to add the voice peak problem on top of that.
--
73 Greg KF5N
Links:
------
[1]
[2] /g/SoftwareControlledHamRadio/message/24140
[3] /mt/101160969/243852
[4] /g/SoftwareControlledHamRadio/post
[5] /g/SoftwareControlledHamRadio/editsub/243852
[6]
/g/SoftwareControlledHamRadio/leave/10484476/243852/1
943518115/xyzzy

The attenuator board has 64 dB of attenuation in 0.5 dB steps. That will get you from milliwatts to one-hundred watts with my PA modules. It's I2C so really easy to incorporate.


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ VP2EHZ

Owner - Operator
Big Signal Ranch – K9ZC
Staunton, Illinois

Owner – Operator
Villa Grand Piton – J68HZ
Soufriere, St. Lucia W.I.
Rent it: www.VillaGrandPiton.com

Moderator: North American QRO Group at Groups.IO.
Moderator: Amateur Radio Builders Group at Groups.IO.

email: bill@...

On 2023-09-07 14:56, K9HZ wrote:

That’s why I build the digital attenuator board to use with the
PAs… Use the full range of the A/D then… and attenuate the power
where it should be attenuated.

*** Definitely the way to do it. Every last bit of A/D resolution is
precious.

For my own use case, I only need 10W. Because my linear will turn that
into 800W. I do need to be able to output 15W or so just for tuning. I have
an MFJ998 remote autotuner at the base of my antenna, and it takes 15W
to urge it to tune. Hitting it with the full 800W when it's not properly tuned
can fry it.

Setting the T41 to a quarter watt for tune should get me a reasonable output
out of the linear.

So - the two power levels I need are 10W and 1/4W.

- Jerry, KF6VB

DR. WILLIAM J. SCHMIDT - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ
PJ2/K9HZ VP2EHZ
Owner - Operator
Big Signal Ranch – K9ZC
Staunton, Illinois
Owner – Operator
Villa Grand Piton – J68HZ
Soufriere, St. Lucia W.I.
Rent it: www.VillaGrandPiton.com [1]
Moderator: North American QRO Group at Groups.IO.
Moderator: Amateur Radio Builders Group at Groups.IO.
email: bill@...
FROM: [email protected]
<[email protected]> ON BEHALF OF Greg KF5N
SENT: Thursday, September 7, 2023 11:00 AM
TO: [email protected]
SUBJECT: Re: [SoftwareControlledHamRadio] T41 Transmitting (noise in
ssb)
It will take some calculating to discover the optimum amplitudes out
of the Audio Hat.
Voice peaks need to be accounted for. So you can't drive too hard or
the peaks will get clipped.
The T41 power control is done by varying the I and Q amplitudes. So
that is another consideration.
If you want power varied from 1 watt to 20 watts that soaks up a lot
of the DAC range.
And then you have to add the voice peak problem on top of that.
--
73 Greg KF5N
Links:
------
[1]
[2] /g/SoftwareControlledHamRadio/message/24140
[3] /mt/101160969/243852
[4] /g/SoftwareControlledHamRadio/post
[5] /g/SoftwareControlledHamRadio/editsub/243852
[6]
/g/SoftwareControlledHamRadio/leave/10484476/243852/1943518115/xyzzy

Hi Bill, yes that is a very good function to add to the transmitter.

--
73 Greg KF5N

开云体育

Hi Harry, microphone compression has a couple of bugs.? Those problems will be resolved in the next release.

--
73 Greg KF5N

Greg and Larry

Re voice peaks I was monitoring the audio lines into the multiplexer on a scope and the waveforms appeared to be limiting at 400mV pk-pk no matter how I shouted.

I'm getting 40dB plus carrier suppression up to 20m and in the 30dBs for higher bands at over 10W.? That will presumably get worse at lower powers.

I'm getting significant noise on TX at low frequency end of the spectrum (no talk) . I tried to kill it with the mic comp (which sets up ok) but it doesn't seem to work on the noise.??

73 Harry? GM3RVL

On 2023-09-07 08:05, Larry Acklin wrote:

OK, this tracks with what Greg suggested- additional attenuation at
the PA and raise the gains to compensate. The bassy audio quality
could be the microphone- Electret mics are variable in quality- or

I have gotten multiple reports of "muffled audio". I tried to address it by raising the
highs with the graphic equalizer function. My microphone is not fancy - it's a Baofeng that
came with a UBITX radio kit. I did replace it with a Yaesu desk mike. Which has much smoother
lows - but still muffled.

- Jerry, KF6VB

The SG5000 codec shows .6V RMS max out of the line out. That translates to well over 1V P-P.?

Perhaps the calibration of SSB PA level will help raise the output in conjunction with your increased mic gain. The SSB Calibration of PA output obviously is another setting in software. It seems to make a difference for me, and is quite sensitive in setting levels- to get the PEP I wanted meant I had to go way down with that setting. Means the PA is too sensitive I think. I’m using the K9HZ (Bill Schmidt) PA. A -10db pad will be in my next fact finding session.?

Larry
KB3CUF



Larry

It will take some calculating to discover the optimum amplitudes out of the Audio Hat.
Voice peaks need to be accounted for.? So you can't drive too hard or the peaks will get clipped.
The T41 power control is done by varying the I and Q amplitudes.? So that is another consideration.
If you want power varied from 1 watt to 20 watts that soaks up a lot of the DAC range.
And then you have to add the voice peak problem on top of that.

--
73 Greg KF5N

OK, this tracks with what Greg suggested- additional attenuation at the PA and raise the gains to compensate. The bassy audio quality could be the microphone- Electret mics are variable in quality- or possibly RF making it’s way backwards. The spurs are what I have been working on, the noise with no microphone attached is next.?

Good progress for you!

Larry
KB3CUF

Well, that’s encouraging progress. How were you able to increase the drive levels from the audio adapter to the qse? I’m not sure but I haven’t attempted that specifically. I’m measuring the output of the QSE at the fundamental frequency (CW mode) and get a bit better than 1 mW into the PA. I also have scoped the I and Q just before the output mux and see relatively good waveforms.?

Larry
KB3CUF

Hi Larry
Thanks for the reply.
My QSE V10 signals are all getting through to the output multiplexer. I've managed to raise the multiplexer inputs to around 400mV pk-pk and that has made a big improvement to the QSE output.? The spurs I mentioned before are not visible - suspicious but not complaining!
In parallel with that I've increased the attenuation on the PA input reducing the PA drive by about 10dB.? Long term, if this works, I'll change the preamp.to give less gain.
With those changes I've just had a 5&5 contact on SSB with a station in Kent (I'm in Edinburgh).? Conditions were not good but I got through OK with about 10W on 40m. He commented that the audio was very good except that it was very bassy which fits with what I'm seeing on the SDRPlay/SDRUno.? Ive tried to alter the audio by using the equalisation option.? It sets up fine but I don't think it's doing anything.
The noise is still there on TX although a lot smaller in proportion because I'm using a bigger drive signal.
I don't have any filter in the connection from QSE to PA.
I would like to raise the 400mV signal into the multiplexer.? I tried changing the scaling in
arm_scale_f32(float_buffer_L_EX, 20, float_buffer_L_EX, 2048); //Scale to compensate for losses in Interpolation??
I changed 20 to 40 and higher but it didn't seem to make any difference.? As they say it was worth a try!
I would like to raise the 400mV to probably 1.5V or even more.
Once I maximise the output of the QSE I'll reduce the PA preamp gain.
Thanks?
Harry? GM3RVL