Sunday 24 March 2013

Arduino VFO

The m0xpd wallet has been dusted off - to the extent that I actually invested in a genuine Arduino (!!!) 

Nothing against the penny-pinching WotduinoTM - but I need to be able to show true compatibility for some "public appearances" in the future and there's nothing more convincing than hearing the tune played on a genuine Stradivarius fiddle.

As opening application for the kosher Arduino, I rustled up a VFO system - much in the style of the excellent pa0klt VFO I got from Jan at SDR-Kits a few years back.

I took another of my Arduino DDS shields (this time made up with a socket for the DDS module, as I wanted to be able to plug different units in for testing). Then I took my code for driving the DDS module and mixed in some rotary encoder interfacing and some calls to the Arduino LiquidCrystal library to drive a 16*4 LCD.

Here's the finished result...


 Like my Arduino QRP rig, the VFO currently jumps the gaps between the amateur bands - although there's nothing to stop me having a continuously-variable frequency, with no inter-band gaps if I choose. That's the beauty of this approach - it is flexible and entirely configurable. Something other VFOs can't match.

It's also inexpensive - even if you insist on buying a pukka Arduino, you still might get change from £25.

Cheap chips!

...-.- de m0xpd


Sunday 17 March 2013

Arduino Honeymoon Over

After a long and enjoyable honeymoon period with the Wotduino, in which she could do no wrong, I woke up this morning to see her for what she is - warts and all...

The problem was that I had made some modifications to the code (what the Arduino community call the "sketch") for my little QRP Rig and I had committed the ultimate sin - I hadn't saved it!

 

Go ahead - laugh at me. Call me all sorts of names. Tell me I should never do that.

Fact is - any other IDE for a micro-controller I've ever worked with (as far as I can remember) saves the code when it is built. So, naively, I expected Arduino would do the same. The sort of assumption you are likely to make in the first flush of romance!

No sir! Not the Arduino!

Turns out that this is a weakness which trips up many newcomers - a quick search on the net returns reassuring comments like "Lots of people ask this. Quick answer - no." (on the Arduino forum, no less). Fortunately, the quick answer is that its IS possible to recover code that has been uploaded to the Arduino, even when you didn't go through the stone-age step of saving the sketch at the time (or even when you closed the Arduino IDE).

I was put onto the scent by Robert's useful post, which had me searching for a folder of the form "buildxxxxxxxxxx.tmp" and which, in my case turned out to be:

C:\Users\Paul\AppData\Local\Temp\build6490475155612533273.tmp.

You need to remember the (approximate) date at which it was created / modified to find out which is the correct folder.

Then you look for the file with the same name as your sketch and a C++ extension - in this case, the file highlighted...

 

Opening this file (in Notepad, if on a windows machine) will reassure you that you're on the right track...

 

Then, it can simply be copied and pasted back into the Arduino editor window where it belongs...

   

Here I changed the name of the sketch and made damned sure to save it - I won't be going through this fiasco again!

So - there IS a way to rescue all your typing and escape the disappointments and inadequacies of the Arduino "Development Environment". Perhaps she's worth sticking with for a little while longer!

...-.- de m0xpd

Saturday 9 March 2013

Arduino QRP Transceiver

Meet my new rig - a QRP CW Transceiver for the lower HF bands, built around the Arduino "open-source electronics prototyping platform".

   

Readers have seen in recent posts how I developed a multi-mode beacon using the Arduino - but that was only a waypoint on the journey towards a full ham radio transceiver.

We started the journey with my homebrew version of the Arduino - the "Wotduino" and then wrote code to make the Wotduino work as a Keyer. Then the Wotduino was given the ability to generate radio frequency signals, using an AD9850 DDS chip on an inexpensive module from eBay, mounted on a homebrew DDS shield. Then we added a QRP transmitter, enabling the system to work as a beacon. The last leg on the journey is a receiver, as seen in the photo above.

My new receiver is derived directly from the Rx stage of Roy Lewallen, w7el's beautiful and definitive Optimised Transceiver for 40m.

I've used this as a start point before, both in the Rx for my "Funster Plus" rig and as a QRSS/WSPR receiver.

My new circuit is implemented on an Arduino shield, as seen below...

 

The signal path from the antenna goes to a double-balanced modulator, where it is mixed with the LO signal from the DDS module. This generates audio frequency (by direct conversion), which is amplified in a common base stage and then passed through a 2nd-order Sallen-Key bandbass filter. Next, a further gain stage controls volume and limits the voltage swing (ironically using the same back-to-back diode method exploited in the Tube Screamer). The receiver is muted during transmission by a FET pass-gate, under the control of the Arduino. Finally, a little audio power stage drives headphones or small speakers.

The receiver sits in the stack of "shields" on top of the Wotduino...


All this adds up to a viable transceiver, as explained in the block diagram of "architecture" below...


The Wotduino controls everything through what we might call a "control bus" formed by the stackable header plugs on either side of the boards.

To this bus is connected a paddle and a straight key, interacting with the keyer code. The DDS module also sits on the control bus, through which it is instructed to generate the appropriate frequency. This RF is carried on an additional "RF bus", implemented on a line of 6-way stackable headers which link the DDS, Tx and Rx shields. The DDS signal is amplified and buffered by a circuit on the Tx shield, from where it is passed to the new Rx shield by another line on the RF bus.

Also hanging off the control bus is a display to show the operating frequency. This is an old 12-character, 7-segment serial input LCD module which I picked up at some long forgotten rally for pennies.

The frequency is adjusted using a rotary encoder - which also selects the band. The rotary encoder has a push-button, which allows me to adjust the frequency in 10Hz, 100Hz or 1kHz increments (10Hz is the lowest resolution necessary, whilst 1kHz lets me scan across the band - or change bands - quickly). When you adjust the frequency above the top band edge (more exactly above the top of the CW segment of the band), it automatically steps up to the next band. Similarly for downward QSYs.

Here's the rig in the habitual lashed-up state (perhaps I should say "Kludged" or "Kloodged" state) on the bench, back in the day when I was still using a potentiometer for frequency adjustment...

 

It is connected to speakers from a roadkill computer and an old straight key for testing. I don't know about you, but whenever there's a viable rig testing on the bench, I just can't keep my hand off the key and the antenna coupled to a dummy load - I always end up having a QSO. So much for testing!

In spite of all my high intentions of making a case, etc, etc, I couldn't stop myself from answering Andy dm5mu in Leipzig and getting a 559. Then Miroslav, yu1dw, in Cicevac, Serbia for a 579. Not bad for my first two attempts on 40m with my 500mW. After these initial QSOs I added a "send/return" connection so I could route the audio signal through my variable bandwidth CW filter (SPRAT v.146 p20), which makes the rig much easier to use when the bands are noisy or congested. The rig is set to cover top band through 20m - but so far I've only had QSOs on 80 (including Ian, g3roo's QRSnet) and 40.

Right - now let's see about some kind of enclosure to pretty the whole thing up a bit...

...-.- de m0xpd