Please help with the development of Peaberry CW

[AE9RB]

SDR has always had a bad reputation with CW. Various excuses have been used over the years. I think Peaberry CW has proven this is simply because we never had the right software architecture. There were technical issues for which I had no precedent or published research to build on. But all the difficult problems have been solved.

It took about four months to get Peaberry CW where it is today. Imagine the salary of a software engineer and divide by three. That's what it cost to make this Open Source software available. There's still a lot I'd like to do. Unfortunately, I have expenses like food, housing, health insurance, and so on. Selling radios doesn't come anywhere near covering these expenses. The Peaberry projects are very much a labor of love.

Right now I'm looking for a job. Development will slow substantially. However, you may be able to help create a different reality. Here's some thoughts...

Spread the word. Let your club, mailing lists, forums, twitter followers, and facebook friends know this project exists. It's possible there's an untapped market of CW enthusiasts who can support development a while longer with the purchase of radios. I'm sure these folks aren't actively looking for SDR solutions because the bad reputation has been around for more than 15 years. So we have to go find them.

I've been looking at crowdsourcing (kickstarter/indiegogo/etc.) but this isn't my area of expertise. I've never seen a ham radio project get funded. Hacker SDRs get funded all the time but they have a substantially larger market than hams who need FCC part 97. If you have experience with successful crowdsourcing and want to run a campaign, please let me know.

Sometimes, open source projects get grants or fellowships. Maybe you are or know of a philanthropist who likes what I'm doing and wants to write a big check.

Flex has all but abandoned the low-end SDR market. Their $700 entry level radio can't even do CW as well as a Peaberry now. Perhaps there's an opportunity to turn the Peaberry project into a real company. I think an all band Peaberry with 5-10W output would be extremely popular but it would have to be factory-built which I can't afford to do. I'd need a business guy, an investor, and so on. I've started several conversations along these lines over the years but they never went anywhere.

If you know C++ and Qt you could contribute code. It's fairly advanced code so you'd probably learn some interesting things along the way. For example, all concurrency is lockless; not a single mutex in the entire application.

[Insert your idea here.] Whoa. I hadn't considered that.

[N2TOH]

My first question is this, how to CW operators what to experience their operations?

Do they want to use their ears or eyes to receive the code? perhaps both maybe.

How do they want to send the Morse code?

Do they want to pound out the dits and dahs manually with their favorite key? or perhaps a more modern keyboard is called for.

After those questions have answers, maybe then software based CW operations will improve.

[k1qn]

I have a long history of CW. The short version is, I was a ham for 20 years before I worked a country on SSB without first working it on CW.

I worked 1 QSO with the Peaberry using cw software, typing and reading. I found it an empty and unsatisfying experience. I like the key and the sound.

I desperately miss CW "the old way".

[N2TOH]

Hi Al,

So if I'm understanding you correctly you want the CW experience to sound and feels like you remember and are familiar with?

What do you think of retaining that original CW operator experience with other digital modes? with the computer outputting audible CW and your key as the input, so regardless of the mode operated it all sounds like CW.

I have a long history of CW. The short version is, I was a ham for 20 years before I worked a country on SSB without first working it on CW.

I worked 1 QSO with the Peaberry using cw software, typing and reading. I found it an empty and unsatisfying experience. I like the key and the sound.

I desperately miss CW "the old way".

[AE9RB]

This initial version has a minimal set of features to prove beyond a doubt that real QSK and software keying is usable by fast operators. If that turned out to be unsolvable, I was going to scrap the whole thing. Only now that it works is it time to add convenience features.

Currently, Peaberry CW does "traditional CW" just fine. Plus, you get a band scope. Several people are already enjoying it on the air.

As far as what features get added next, that's simple. Whoever does the work or pays for the work to be done gets to decide. The things I was planning to do next are listed in the "Download Peaberry CW" thread.

[tannebri]

Selectable Audio output would be nice for those of us with multiple sound card options.


Brian

[W4MMP]

Here is my take on what I would like to see (or perhaps do in part) with the Peaberry environment. Below are design goals (first draft). This is not specific to Peaberry CW but I thought I would post it here.

1) Provide an architecture that allows multiple Peaberrys appear as one rig.
This I believe will be a combination of hardware and software. An interface adapter will be needed that will provide a one common interface to the host computer. All ports of individual Peaberrys will be attached to this adapter including the USB port. The adapter will accept up to four Peaberrys. Currently a combination of three Peaberrys covers all the HF bands. The fourth attachment is for future expansion should David decide to create a 6 Meter version. Optionally a 5 Watt amplifier with low pass filters could be included on the adapter. The adapter will be micro processor controlled that provides routing of the individual Peaberry ports to the common host interface ports. Under micro processor control USB data will be routed to between the common USB port to a selected Peaberry USB port. Routing will be directed by the ATU port on each Peaberry as the ATU port currently provides band data. Most likely a few tweaks will be need to the firmware

2) Preserve current user financial investment.
Taking the above approach, protects the current investment owners have in their Peaberry(s) versus purchasing an all band Peaberry.

3) Expandability.
This is modular approach to building a complete station. Hams may get their feet wet with a purchase of a single Peaberry. If they like what the Peaberry provides, then an adapter and second and/or third Peaberry may be purchased to build a complete all band rig. With this approach the ham can do an incremental approach to funding the build out of the rig.

4) Backwards compatible.
As always any changes to Peaberry firmware will be completely backwards compatible with the current Peaberry firmware. This the approach I have taken with the Peaberry CW firmware I provide.

5) Open source host application.
One of the biggest rubs I have had with the SDR environment is the "free as in beer" approach of the SDR host application providers. I have been unsuccessful (perhaps due to my sterling personality ) with getting any changes I would like to see to better support the Peaberry. I believe we as a ham community need a true open source host application ("free as in speech"). Optimally the host application will run on either Windows and Linux. Quisk is one option but is Python script based and has certain performance issues, particularly smaller micro processor implementations (such as the Raspberry PI 2). One application that has promise is OpenExpertSDR that I have been investigating. It is Qt based and there is currently a port for Linux. Placing the source on Github or Codeplex will provide (hopefully) a robust development environment allowing for quality and functionality enhancements.


Funding: Perhaps one method for funding this is to use some type of crowd funding. I believe I could fund the initial interface adapter prototype and create the firmware needed to run it. However getting the host application off the ground is a challenge for me. I am out of my league with that level of complexity. Having a developer dedicated full time is (IMHO) needed. Crowd funding the initial release may be pie in the sky, but it's an idea.

So, for what it's worth, this my take.

[N2TOH]

Ron,

if we are going to compile a list of requirements, I'd like to mention my frustration with the this one but not that one type of support. So it would be nice if there were at least Win/OSX/linux/UNIX support with the same common features between all supported operating systems.

an application that supports multiple radios simultaneous should mitigate the need for additional hardware.

the ability for the end user to write their own interfacing scripts should help users, for example the ability to implement both USB and legacy PC serial and LPT ports for transmit/receive change over for the VHF+ operators. or what ever specialty needs a specific operator may need. be it power management or remote control antenna switching.

[bubnikv]

David.

The direct CW mode is a great idea. I had a similar plan for a smooth CW QSK SDR solution by extending a SoftRock RX Ensamble firmware to generate the CW carrier by the Si570 synthetizer directly without having to mix it with a sound card audio. Your solution is general and works great indeed! My idea would reqire a lot more software support.

In regard to a SDR software development, I would recommend to talk to the author of the SoftRock software to integrate a support of your CW mode. If I were you, I would send him one unit and I bet he will enjoy it. You would then have more time for the development of a 5 or 10W version of your new peaberry, for which I would be a happy customer.

73, Vojtech OK1IAK

[parthsane]

I'm interested on getting the peaberry v2 on OS X . As of now OS X detects the peaberry radio audio interfaces. I have experience with programming just not on mac. I am an Android developer and more fluent in Java. It would be nice if someone can point me in the right direction. What I have understood from using the peaberry on Windows with HDSDR:-

1.VFO control is independent from the audio processing<br>
2.RX TX modulation and demodulation algorithms must be implemented and sent to the audio interfaces.

Pls guide me so I can make peaberry software on mac.