The Ubuntu Platform team (the people Canonical employs for Ubuntu) is having a sprint in Portland. Portland is a nice city where you can be wandering down the road and come across 100 tweed wearing cyclists coming the other way.

My how they've grown

Kees' collection of every shipit CD ever nears completion

Voodoo doughnut, a local speciality. I got half way through eating this before gaining diabetes
February 03, 2010 11:04 PM

Late last year the Electron Club’s Roy Mohan Shearer sent an E-mail to the list asking if people would be interested in testing a small FM transmitter he was developing. Whilst at first this might not sound that exciting, this was an exercise in open source product design, rather than just putting together a simple circuit. Radio and open source ticked two big boxes for me, so how could I resist…
This blog post serves as my own notes on initial experiences with the Niftymitter and as a vehicle for feedback. To date my use has been limited to basic testing, however, I have a few hacks in mind and hope to find the time to try these out over the next month or so.
Testing

First of all I hooked the Niftymitter up to my laptop, played MP3s via Songbird and tuned in via a radio in the bedroom next door. The distance was probably about 10 feet, and with the signal having to penetrate a set of densely packed shelves carrying vinyl records and an interior wall. Somewhat surprisingly, given the obstacles in the path, the received signal quality was excellent.

Next up I connected the Niftymitter to the output of my record player, or to be more precise, to the attached “phono stage” (pre-amplifier). By this point a pirate radio station had surfaced on the frequency the transmitter was originally tuned to, and so I had to find a clear spot on the radio dial and re-tune it. Following which I experienced another small problem: overmodulation, resulting in distortion, due to the signal from the phono stage being too high.

This was quickly addressed by connecting a passive volume control (attenuator) into the path between the phono stage and the Niftymitter. The volume control could then be set to achieve the best sound quality at the receiver. The radio was situated some 20 feet or so away in my workshop, and again the sound quality was excellent.
Wish list
There will be use cases where a signal source has no level control, e.g. a HiFi ’seperates’ CD player. If the signal level happens to be too high it will result in distorted reception. So, an input level control on the Niftymitter might be handy, i.e. a small preset variable resistor, that like the tuning control is normally hidden.
I’d also quite like to see an AM version of this, although I’m confident that I am in the minority. The reason for this is that most of my radios are old valve sets that don’t have FM, and commercial low power AM transmitters are massively overpriced for what they are, despite being just as simple as a similarly sized FM transmitter. In fact, I quite like the idea of a Niftymitter sized FM -> AM transponder (re-broadcaster), that integrates a simple mono, tunable FM receiver and tunable AM transmitter. This would be great for listening to FM-only stations on old valve sets.
Conclusions
On first impression an open source FM transmitter project seems like an odd thing to do, when you can buy them ready made for probably less, and with digital tuning and stereo encoding. Furthermore, many fun hacks that immediately spring to mind are likely illegal, e.g. attaching an external antenna or boosting transmitter output. However, as a product it is something that is of immediate utility to most, employs a circuit that is not too difficult to grasp, should be easy to make and costs little to acquire the parts for. Furthermore, there is something fun about radio and setting up your own little transmitter, even when you have to comply with radio licensing laws which severely limit the amount of transmitter power you have on hand.

I think a radio in a matching or even interlocking case would be a fun companion. Together they would be great learning aids, and with an AM variant of the Niftymitter you could also then make an FM -> AM transponder. And I really like the construction of the enclosure - it not only looks smart but means that you can nicely house the transmitter without the need for tools to drill the case etc (although the laser cut labels were a nice touch).
More to come as I get the time to try out a few hacks…
Share This
January 31, 2010 04:37 PM
Back in December 2007 I tried a bit of cycle commuting on a borrowed mountain bike. I stopped when I realised that it wasn’t really saving me any money, and because I thought I was about to move to a job closer to home, whereupon I would start again. Sadly, the job never materialised, and I never started up cycling again, despite having done over 200 miles at that point.
Fast-forward to 2009, where I got 2 offers – a new job at Central and a new (to me) bike via an online friend. He had been given an old 10-speed road-bike, but wasn’t allowed by his wife to keep it, so I said I would take it.
I took delivery of an old Peugeot Carbolite 103 (actually the name of the frame, the model is an obscure number on a sticker near the bottom bracket that is a bit torn
) in November, but sadly there is no such thing as a free bike. New tyres were a good start, but when I went for my first serious test ride, the chain broke. A new chain then enabled me to see that the saddle was woefully uncomfortable, so a new saddle arrived for Christmas. I’m still in need of some mud guards, and I suspect a rack at some point, and some additional cycling clothes, etc, etc, etc. You get the picture
Me collecting the bike:

A few alterations later:

Of course, then there are the alterations that need doing. The brakes are shoddy and need a bit more work, if not replacing. The brake levers have been moved, but need moved a lot more to make the drops anything more than decoration. I have bar tape to fit (matching the saddle…). Plus it now needs a good clean and oiling.
However, I’ve done 100 miles in January, having ridden every work day, and I intend to ride at least 200 work days in 2010, which would be about 1000 miles of cycling. It will also save me in the region of £400, although I would need to minus expenses off that…… now – where’s that cycling mag….
mrBen
January 28, 2010 09:44 PM