I wired up a simple astable multivibrator 555 timer circuit (which outputs a series of 5 volt pulses) whose frequency I could control with a potentiometer. I got a 6ft lamp extension cord, cleaved the two wires in two and cut one to put the relay in line.

The primary reason to do the test was to see how much latency the 100 watt incandescent bulbs display at various rates. The bulbs take a little while to get to full brightness after current starts to flow and even longer to become totally dark after it stops. A little bit of this is okay, and the trailing edge will actually produce a neat ghosting effect as the meter falls off. Too much latency though, especially in reaching brightness, will mean a dim top edge to the meter and a disappointing disconnect between sound and visual.

By playing around with rapid pulsing and watching it closely I don’t think it will be a problem. It’s difficult to tell without seeing several of these in a row though, so I’m going to try to measure the brightness curve from the following video and see if I can write a quick simulation (using processing?) of what the completed meter would look like with these bulbs. The camera’s auto-exposure function sensor latency will skew the data a bit, but it should be enough for a rough idea. Alternatively, when I get an oscilloscope (what can’t they do!), I could use a photoresistor to measure the brightness curve more precisely.

I’ve started a photoset for this project here.

This is Patterson Office Tower (POT) on The University of Kentucky campus. It’s home to a small cafe/mezzanine, an obscure mathematics library, and the offices of all the professors and bureaucrats who don’t fit in the endearingly inefficient buildings elsewhere on campus. If, as civil engineers like to say, architecture is the art of wasting space, it’s oft-scorned brutalist style seems to be a good example of what would happen if we eschewed the discipline altogether.

On top of this mercilessly orthogonal concrete form is the broadcast tower of Radio Free Lexington. Despite it’s “dim bulb” 250 watt transmission power, the station currently reaches most of Lexington’s 275,000 or so residents via terrestrial broadcast thanks to the height of the tower and the hill it sits on. The FCC has approved WRFL for an upgrade to 7900 watts, which will let us reach a much larger swath of the central bluegrass.

So I have this idea for an installation I think would work well to draw attention to the station either in celebration of the upgrade or as a means to gather support for it. Following in the steps of the green building hack at MIT and the Blinkenlights Project I’d like to turn POT in to a giant real time volume display of the station’s broadcast.

To be more specific, I want to take a row of windows (or more, as money allows) in the building and place lamps in them to form a vertical bar VU meter. The overall effect will look like the video below, but 16 stories tall and visible from miles away. You could tune in to the station during the evening and watch this huge column of light on the face of POT leap up and down in synchrony with the music. I’m grinning just thinking about it.

How could that be done? Well the Blinkenlights and Green Building projects were both able to run significant amounts of cabling from large power supplies to relays across several stories/rooms. POT is in constant use so cost and difficulty considerations aside we can rule out any scheme involving huge cable lays.

Happily, the signal we want to represent is already available as radio broadcast on every floor. So instead of switching a huge number of lights from a central system, we can put a radio on each floor and have each set of lights switched autonomously in response to the signal it receives.

So the basic plan is to have a simple lamp with, lets say, three 100 watt bulbs in each window. You have a relay (actually, this relay) that can switch the bulbs on and off quickly in response to a low voltage signal. This signal is provided by a simple circuit that detects how loud the signal coming from the radio is, and while it’s above a certain threshold, activates the relay to turn on the lamps. By setting that threshold slightly higher on each successive floor, you’ll get that VU meter effect.

That circuit to sit between the radio and the relay will be an envelope follower and a comparator (probably an LM339). The envelope follower may be as simple as a diode to rectify the signal and a RC low pass filter. The whole apparatus (lamps included) should cost less than $40 (that’s a high estimate) per floor. Assuming we do 16 floors, and a few red ones on the top floor for a cool peaking effect, that’s around $800. Probably less. But let’s say $1000.

I’m currently working on a prototype of a single unit. I’ve already got one of the relays to play around with. The problems remaining are…

• Finding some radios that can run off a wall power supply, yet are cheap and can be had by the dozen
• Designing and testing the envelope follower with an oscilloscope. The big question here is what kind of time constant should be used to prevent flickering but not take too long to fall off.
• Finding good potentiometers for adjusting the threshold of the comparator.
• Packaging all of this, wiring everything safely, and in a duplicable way.

Then it’s just a matter of finding funding and convincing the powers that be (who do you even ask? the building super? each individual office holder?) to let us set it up. It could easily be reused or expanded (add some band pass filters for a frequency analyzer), and different locations provide for some interesting opportunities. For live events, you could use a low power FM transmitter to drive them instead.