Iambic Firmware for the Peaberry (Release Thread)

General discussion and support for the Peaberry SDR V2.

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby KA1CFP » Tue Dec 09, 2014 10:50 pm

Hi Ron,

First, let me admit that I am a absolute amateur in both CW and use of paddles. I am just (re)learning, even after 30 years with my ticket. I amazes me that my 62-year-old brain is still able to learn new tricks!

I did a bit of testing after wiring up the paddles, loading your new firmware, and also building a PicoKeyer for comparison.

I took a look at the CW transmitted waveform from the Peaberry as monitored by my aging but always reliable Kenwood TS440SAT. Testing conditions were as follows: BY-2 paddles connected to the KEY input on the Peaberry, IAMBIC MODE ON, IAMBIC IAMBIC mode selected, CTRL-SHIFT-U enabled, transmitting on 40m into a low wire antenna. Receiver is the aforementioned TS440 on another co-located small antenna. Paddles set for 15wpm, both direct into the Peaberry and also through the PicoKeyer for comparison.

The CW transmitted by the Peaberry does not sound smooth to the (admittedly untrained) ear, so I dug out the scope and recorded the received audio from the TS440:

Peaberry V2.png
One "dit" from the Peaberry on the monitor receiver.
Peaberry V2.png (31.88 KiB) Viewed 8010 times
Note the transient just ahead of each transmitted dit or dah.

For comparison, I also recorded a few other close, strong CW signals on 40m, two of which are shown below:

40m CW1.png
40m CW1.png (43.52 KiB) Viewed 8010 times


40m CW2.png
40m CW2.png (32.25 KiB) Viewed 8010 times


I then tested with the PicoKeyer attached to the KEY input on the Peaberry, all other conditions being the same, except IAMBIC STRAIGHT mode was selected to accept the output of the outboard keyer. No change in the shape of the monitored output waveform was noted.

So, I'm wondering if the transient and keying shape is due to the way the transmit output is being gated by the firmware, or perhaps this is normal behavior for the Peaberry. Wiser minds than mine should probably weigh in on this one.

All in all, just love the firmware and the Peaberry. Let's keep the improvements coming!

Marshall
KA1CFP
KA1CFP
 
Posts: 4
Joined: Wed Dec 03, 2014 7:52 pm

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby W4MMP » Wed Dec 10, 2014 12:40 am

Hi Marshall,
This is exactly what I was hoping someone would perform, a more in depth analysis of the wave shape. I don't have an immediate answer for the transient. I do request that you do one more test. Instead of using the iambic straight key feature (function), please turn off iambic and only have CW mode set. Then key the Peaberry with your PicoKeyer. This will help isolate the problem area. When the Peaberry is in CW mode only, the firmware simply turns TX on and off with the key input. When in Iambic mode the firmware goes through a fair amount of code to generate the CW even if the "straight key" mode is selected. This will tell me if the transient is a hardware or software issue.
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby KA1CFP » Wed Dec 10, 2014 1:57 am

Hi Ron,

Great idea!

OK, here is the CW waveform with the BY-2 paddles feeding the PicoKeyer direct into the KEY jack on the Peaberry V2:

BY-2+PK+PB_KEY_IAMBIC_OFF_dit.png
W4MMP firmware on Peaberry V2, Paddles w/PicoKeyer into KEY input with IAMBIC function OFF, CW transmit mode on 40m
BY-2+PK+PB_KEY_IAMBIC_OFF_dit.png (27.82 KiB) Viewed 8007 times


Maybe I should load the original firmware back in as a test to see if it tracks with the firmware?

Marshall
KA1CFP
 
Posts: 4
Joined: Wed Dec 03, 2014 7:52 pm

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby W4MMP » Wed Dec 10, 2014 2:37 am

Hi Marshall,
Yes, if you don't mind taking the time, please fall back to the previous version. One thing this does tell me that it is not the Iambic routines that is causing the problem. Obviously I need to give this some thought. It looks like the de-bouncer(s) are not functioning properly.
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby AE9RB » Wed Dec 10, 2014 2:39 am

Did you turn off AGC and all other signal processing on the receiver?
73 David AE9RB
AE9RB
Site Admin
 
Posts: 511
Joined: Thu Nov 15, 2012 5:27 am
Location: Portland, Oregon, USA (CN85ql)

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby KA1CFP » Wed Dec 10, 2014 2:54 am

Doggone it, David, you were spot on:

TS-440SAT_AGC_SLOW.png
TS-440SAT_AGC_SLOW.png (13.32 KiB) Viewed 8004 times


TS-440SAT_AGC-FAST.png
TS-440SAT_AGC-FAST.png (14.03 KiB) Viewed 8004 times


Well, that's embarrassing. Wild goose and/or crow has been added to the Christmas menu. Good call.

Marshall
KA1CFP
 
Posts: 4
Joined: Wed Dec 03, 2014 7:52 pm

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby W4MMP » Wed Dec 10, 2014 3:10 am

Hi Marshall,

LOL :) But, thanks very much for examining the wave shape! I truly appreciate the effort.

I was about ready to capture a sample via HDSDR and find something to analyze the capture file. I think I will do that anyway.

And thanks very much David for your insight into this. I would have been chasing my tail on this one for a bit. It's good to know the de-bouncer component is doing its job.

(As a side note, I always run HDSDR on slow AGC. Anything else it just sounds too "hard").


OK, for everyone following this thread, except for bug fixes, this really is the last release of the firmware until I have developed a new ExtIO.DLL. As mentioned before, the plan is to move all the "magic number" stuff into the dll in the hopes of allowing for smother and cleaner operations.
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby W4MMP » Sat Dec 13, 2014 3:54 pm

Please find below a theory of operations for the firmware. I got the impression recently there may be a little confusing on how the firmware works particularly in relationship to HDSDR. So for those interested here it is ;). This will also be posted on the codeplex website.

A little background on the theory of operations.

For an I/Q based transceiver (Softrock RXTX, Peaberry, ...) to generate a CW signal the host application (HDSDR) applies an audio tone to the transceiver via the audio device connection. For the Peaberry this audio connection is listed in the Windows (Win 7) device manager under the "Sound, video and game controllers" category as "Peaberry Audio" The audio tone is applied to the mixer section, specifically in the case of the Peaberry V2 it is U8. The tone is mixed with a RF signal at the LO (local oscillator) frequency. The LO is provided by the Si570 (U9). The SUM of the two frequencies is then applied to the PA section which is the TUNE frequency as shown on the HDSDR display. There is some interesting and complex math also involved with the mixing of the two frequencies, but is outside the scope of this document.
The PA section is turned on by setting the Peaberry internal signal (TX) to the on state. The PA then amplifies the TUNE frequency and applies it to the antenna connector. The Peaberry is commanded by HDSDR to transmit a CW carrier signal by the user setting HDSDR to CW mode and then activating transmit by clicking on the TX button or tapping the space bar. This commands the Peaberry into transmit mode via the "Peaberry SDR" device as shown in the Windows (Win 7) device manager under the "libusb-win32 devices" category.
However HDSDR has very limited CW capabilities. As mentioned previously an audio tone must be applied to the Peaberry and this is accomplished by pressing and holding the keyboard "return" key. This obviously does not accommodate CW operations.

This is where the new firmware comes into play. By using a "magic number" the Peaberry is configured to "CW Mode". When in CW mode, the Peaberry intercepts the HDSDR command to set the Peaberry to transmit mode. In effect the firmware manages the internal TX line. It manages the TX line by interrogating the key paddle port. When either "KEY_0 or KEY_1" is closed the firmware turns on the TX line and thereby turns on the PA section thus generating a CW carrier signal. The paddles port may be attached to a regular straight key, an electronic keyer or directly to a paddle. In the latter case when the Peaberry is directly attached to a paddle, the Peaberry may be set to Iambic mode via the use of a "magic number". Please note and as mentioned previously, HDSDR must also be generating an audio signal. This is accomplished by pressing and holding the return key. However there is an alternative to this by entering a keyboard key sequence of "Control Shift U" simultaneously. This sets HDSDR to generate the audio signal whenever HDSDR is set to transmit mode and HDSDR is in CW mode. This is the method that the Peaberry in combination with HDSDR used to generate zero latency CW.

To return to NON CW mode, both HDSDR and the Peaberry must be taken out of CW mode. The Peaberry is taken out of CW mode by double clicking on a "magic number". Obviously HDSDR is simply set to another mode. The Peaberry must be taken out of CW mode for any other mode to function as it must be commanded not to intercept transmit mode. (I wish this was more straight forward, but I am limited to what I can develop due to the current existing software architecture of SDR applications and the ExtIO.dll).
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby W4MMP » Thu Dec 18, 2014 3:37 pm

Hi,
For those of you using or thinking of using the CW/Iambic firmware, I believe I have found a method for implementing semi break-in keying. However to use that feature would require building a simple interface device to attach the ATU port to a comport. Anyone interested? Would you take the time to build the interface? The first beta release will be available sometime after the first of the year.

(Full QSK is most likely not possible as the switch time of HDSDR is just to long even when using a comport to signal transmit on.)
73
Ron / W4MMP
W4MMP
 
Posts: 579
Joined: Fri Jan 03, 2014 3:31 pm
Location: Lovettsville, Virginia FM19EG

Re: Iambic Firmware for the Peaberry (Release Thread)

Postby WB8ENE » Thu Dec 18, 2014 4:28 pm

W4MMP wrote:Hi,
For those of you using or thinking of using the CW/Iambic firmware, I believe I have found a method for implementing semi break-in keying. However to use that feature would require building a simple interface device to attach the ATU port to a comport. Anyone interested? Would you take the time to build the interface? The first beta release will be available sometime after the first of the year.


Ron,

I'm definitely interested, assuming it works with a straight key, which is my usual method of transmitting CW. I'm not sure how much time I would have right now to build the interface though. I just received my Hardrock-50 kit, and the only thing I've been able to do with it so far is to inventory the parts. I haven't even had a chance to load your latest firmware. Thanks for all of your continued effort on this project.

73,
Art WB8ENE
WB8ENE
 
Posts: 47
Joined: Fri Jan 24, 2014 12:35 pm

PreviousNext

Return to Peaberry SDR V2

Who is online

Users browsing this forum: No registered users and 2 guests

cron