Page 2 of 3

Re: testing the CW program

PostPosted: Sat Feb 07, 2015 10:49 pm
by AE9RB
Good to hear the keyer sidetone is better. Transmitting will be no different.

It might be some of the distortion you hear comes from the operating system changing sample rates. I'm currently rewriting all the audio to get direct to the hardware (like I did for the sidetone). Attempting to use cross-platform abstractions and OS conveniences turned out to be a fool's errand.

Can you check if there are newer video drivers for the Atom computer? The visualization is done with OpenGL and I think some of the early Intel HD drivers were broken. It'd be a shame to switch to the CPU for drawing since a GPU is far more power efficient.

Re: testing the CW program

PostPosted: Sat Feb 07, 2015 10:54 pm
by AE9RB
Also, you can calibrate the S-meter in the settings.

Re: testing the CW program

PostPosted: Sun Feb 08, 2015 6:42 pm
by kz1x
I now see the S meter calibration setting; thanks for the pointer. I will go find my XG3 and calibrate it.

I was looking for the dBm reference, that's how I missed the cal setting, since an S meter (to me) is kind of meaningless.

The spectrum display needs to be able to scale better; Y axis. Auto or manual, doesn't matter to me.

Also ... If I cal the 'S meter' at, say, S9, the Y axis in dBm should line up at -73 dBm.

After extensive playing with this v0.2 code, using multiple keys, etc., I can find no glaring fault with the keyer. It's really quite nice.

Regards the sidetone comment you made... Do you mean to say the sidetone is also going to be QSK RF as heard now?

If so, that's simply remarkable. I could easily copy between elements at max speed. This is TenTec territory.
You'll quickly find a small but vocal set of QSK diehard fanboi groupies. It'll need to have another 15% or so faster top speed.
Right now it goes to 40 WPM. I didn't measure it, but it felt like it was correct at that speed setting.

If you're able to address the funky RX sound, this would be a killer QRP CW rig setup.

It'd be ideal for relatively poor efficiency portable antennas since the PBv2 RX is hotter than your usual portable SDR but the effective RX dynamic range right now is rather challenged.

The AGC just runs out of breath, fast, with any other nearby strong signals ... or even (and especially) the one you're listening to. And, the received note just has too many artifacts on it to listen to for any length of time. It's like a DC RX with low frequency noisy FM on the LO.

The spectrum display on the Atom powered netbook works fine ... or not, it seems to be intermittent. The driver needed is clearly present. If I reboot the computer, it usually starts just fine. But restarting the app later often shows some other screen-attribute background where the spectrum display should be.

This doesn't happen on the i5 box.

Re: testing the CW program

PostPosted: Sun Feb 08, 2015 7:42 pm
by W4MMP
It seems Peaberry CW can't find the Peaberry audio device on my PC. I performed all the standard remedial trouble shooting (reboot, checked the device for existence, ...). Not much to go on as this is the only error reported. I will check the windows event viewer next but I need a clue as to which subsystem to examine.
This is on a Win 7 PC (I7, 6GB) with all the latest updates.


Screenshot 2015-02-08 14.31.08_cr.jpg
Screenshot 2015-02-08 14.31.08_cr.jpg (206.89 KiB) Viewed 9834 times

Re: testing the CW program

PostPosted: Sun Feb 08, 2015 8:15 pm
by W4MMP
David,
I just realized I changed the name of the Peaberry devices for reasons not related to Peaberry CW since I first tested Peaberry CW. Are you storing the Peaberry audio device name somewhere (system registry or an initialization file)?

Re: testing the CW program

PostPosted: Mon Feb 09, 2015 12:54 am
by kb3cuf
I agree the keyer is even better- and I liked it before :D I think it is an improvement, and I hear no lag from key to side tone. I cant send reliably above 30 but I was able to get chars out up to there. I cant copy that fast anyway.

I also hear the AGC pumping on really strong sigs next to the station I am listening to. There may be limitations to that more as a function of selectivity. It started out with no setting in the AGC window and boy that sounded terrible- super hard clicks and distortuon. I was listening to a pileup and after I cycled through the AGC settings once. it behaved better.

If the QSK is this fast, then it it a winner. I am eager to try this.

I also agree that a click to QSY would be awesome. Now I need to think about a QSK amp.

W8.1 4 GB RAM, AMD somehing or other-and no issues with the OS at all.

Thanks for the great work so far.

73
Larry

Re: testing the CW program

PostPosted: Mon Feb 09, 2015 4:01 am
by AE9RB
W4MMP wrote:I just realized I changed the name of the Peaberry

You'll have to change it back for now. Maybe make it shorter. The problem is that MME truncates device names. This shouldn't happen in the next release which will be 100% WASAPI.

Re: testing the CW program

PostPosted: Mon Feb 09, 2015 4:09 am
by AE9RB
kb3cuf wrote:If the QSK is this fast, then it it a winner.

kz1x wrote:Do you mean to say the sidetone is also going to be QSK RF as heard now?

I think real QSK is going to happen. Still rewriting the Windows audio but the Mac version is done and about 50ms from antenna to speaker.

Re: testing the CW program

PostPosted: Mon Feb 09, 2015 4:14 pm
by W4MMP
AE9RB wrote:
W4MMP wrote:I just realized I changed the name of the Peaberry

You'll have to change it back for now. Maybe make it shorter. The problem is that MME truncates device names. This shouldn't happen in the next release which will be 100% WASAPI.


Thanks, but I will wait until the WASAPI release is available.

Re: testing the CW program

PostPosted: Mon Feb 09, 2015 4:52 pm
by kb3cuf
One note- which may or may not be a problem for users-

I can not run this program with anything else. If I minimise the screen, and open Chrome, the audio starts to stutter. After a few things opened, the PC crashed (no BSOD shows now, it just rebooted itself)

So I think HRD or other data loggers would be on another PC.

Personally this is not a problem. I have enough going on with a CW QSO that I don't need to check email also :roll:

Larry