raem promotion photo

Diary - Production/Build 2009

Start the build
Posted on July 18th, 2009 Heres a picture of a generic lcd screen purchased via ebay, its a vga 240 volt ac model. Pictured on the left is the step down transformer and inverter board while on the right is the logic board. After some research via google its been determined that this model takes 12 dc as main supply with a 3v on switch. Theres a dimmer circuit that goes from 0 to 5 volts but as we be on battery power the lcd will have to run at lowest setting, ie 0 volts for the dimmer. Presently this lcd is using 2 cold cathode fluorescent lamps to provide back lighting, this may be reduced to one.
Mighty atom
Posted on July 18th, 2009 This is the Intel Atom motherboard that will provide one half of the audio generating capabilities of RAEM and be carried aboard elevators in whats being called the Suitcase. Its low powered duo core and has been tested at 12 volts with about 1.5 amp drain. This represents a whole new design for us with the inclusion of the atom board as well as a redesign of the amplifier speaker system so that we can provide a richer sound within elevators. The previous version while being of suitable volume didnt provide the quality of audio that we would prefer. The upgraded gear will be taken on tour so it will need to be mains powered, have a line out for audio into a PA as well as being completely self contained and self powered for playing in elevators or other unusual locations.
Less is more
Posted on July 18th, 2009 Here is a test circuit for powering the lcd screen and logic board. The idea with this component as with all the RAEM gear we take onboard elevators is that it all has to be powered only by the batteries we take with us and the performance time is limited to how much power we can take with us and how quickly it gets used. In the 2006 Fringe Festival our total time per battery charge cycle was about 3 hours and for this year we hope to increase that to around 5 hours.
First battery test
Posted on July 18th, 2009 The final test rig for the lcd screen and logic board is ready for a battery test. The battery is a 12 volt 7.2 aH sealed lead acid type of which we used 3 to power the gear during the last fringe festival. Not too sure how the components will behave when they go into undervolt, especially the logic board, but then this is why we do tests at every stage. The screen is at 0 volts dim and is still readable in sunny daylight which is a good thing because some of the elevators we have been in were quite well lit inside. After this test I may try and run just one ccfl to see if theres a noticeable drop in battery drain, which there should be, and that the screen is still readable. There is an old broken laptop lcd standing by that uses only one ccfl so it may be of use, if i can drive the inverter and get the lvds signals figured out that is.
End of test one
Posted on July 18th, 2009 The first battery test is nearing completion with some unexpected results. The logic board for the lcd started switching off at about 10.40 volts after nearly 4 hours use. The inverter board which I expected to be the main drawer of current is actually still running now with just the ccfls lit after 7 hours down to about 10.11 volts. The logic board may be doing more than an initial glance at the datasheet revealed. An undervolt warning circuit has been made to install on the next lcd battery test but I think tomorrow will be busy with a test of the atom board and battery.
Next battery test
Posted on July 19th, 2009 After yesterdays test of the lcd screen, today sees the first test of the atom board running off a battery along side a second lcd test. An undervolt warning circuit has been added to the lcd battery thats designed to light a red led at about 11 volts which should leave 5 minutes to power down the screen. To the left is the lcd boards, screen and battery while on the right from the top is the second battery, psu board, atom board and a hard drive.
Partial end test 2
Posted on July 19th, 2009 After an hour and a half the atom board was manually shutdown following a 15 minute run on the voltage that saw a drop from 10 volts to 7 volts. The datasheet for the psu states it can have min 6 volts but at this stage there was no point in risking damage to either it or the atom board. One possible cause of this is a weak battery so a follow up test will be run using one of the others that have proven to run well for about 4 hours. The lcd screen is continuing to run at the rate expected.
Third test
Posted on July 19th, 2009 This time the atom board has been configured differently with hyper threading turned off as well putting the hard drive to idle after 5 minutes. Battery 1 is used but the results are similar if not worse than before and saw the voltage drop occur after just one hour. This battery is the same one that had a dropout for the lcd logic board after 3.5 hours so maybe the battery isnt in prime condition. Meanwhile battery 2 has finished powering the lcd after 6.5 hours and battery 4 is on recharge after the first atom test that ran for 1.5 hours. The power supply unit used is rated at 150 watts which may be overkill as the intel atom board can run at 80 watts according to the datasheet, so this may need to be replaced with a Pico psu that outputs just 90 watts. Trying to undervolt the cpu at the bios level seems impossible and probably not worth much in savings as the intention is to run the cpu to max load during performance. Another option is to have two batteries in parallel to power the atom board but that will only be looked at after the atom is down to its most efficient power use. Time for some planning…
Atom tests
Posted on July 20th, 2009 A new batch of tests are being run to try various battery configurations, firstly just to reference battery 2 which only lasted 1.3 hours and next to try two batteries in series giving 24.94 volts. This last will mainly test the psu and how it handles the conversion to 12 volts over a period of time. The next test will have the batteries in parallel. Ultimately the goal is to have a powered runtime of four hours with both the lcd screen and the atom motherboard at max load whether it requires two or three batteries.
Two battery test
Posted on July 21st, 2009 This test using battery 1 and 4 in series ended after about 3.3 hours. The main characteristic of this test was that when total voltage hit about 12 volts there was a rapid run down to 8 volts over a period of under 3 minutes. In this case too fast for a proper shutdown and so the psu went into undervolt, the vga signal was lost and no contact with the atom board via network was possible. The atom still seemed to be powered so not too sure what was actually happening other than maybe a shutdown of peripherals but the atom remains fine and the psu handled the undervolt well. Whatever happened, this test replicated the previous fringe festival running conditions including the rapid voltage drop, so next is to try and increase possible performance time by a test of two batteries in parallel.
Another day of tests
Posted on July 22nd, 2009 Another day offers more coffee and also a visit to the post office to pick up a parcel from Japan that has a little synth in it. The customs declaration might say “toy(doll)” but dont let that fool you, its a musical instrument. The last test saw a runtime of 3.3 hours for the atom so today will see a similar test but with batteries in parallel to see if more power is gained from sticking to 12 volts input.
Final battery test
Posted on July 25th, 2009 Well lets hope so. Here is a test of all the Suitcase components being powered at full load by all three of the 12 volt batteries hooked up in parallel. Although only two previous tests were performed on the lcd alone, the results are indicative of low power consumption compared to the atom board. A parallel circuit seems to offer the best running conditions in that voltage drop off is slow and steady as opposed to the rapid drop seen when the batteries are in series. Its hoped that at least 4 hours runtime is achieved with this test.
Twitter test
Posted on July 25th, 2009 Adding a few new functions to the website, in this case blog to tweet.
All is well
Posted on July 25th, 2009 So far all three batteries running well. Hoping that some of the web updates work.
End of first three battery test
Posted on July 25th, 2009 After just under three hours of running time at max load for the atom and the lcd, the latter went into undervolt at then 10 volt mark, dropping to 9.93 volts as the lcd logic board cycled through standby and power up mode a few times. As the atom boards psu can manage a drop to 6 volts it appears to be necessary to find a way of making the lcd draw less power. The logic board is the key here and it may need to be modified somehow. More planning and more tests to follow over the next few days.
Draw me a graph
Posted on July 26th, 2009 Sometimes its easier to figure things out with a drawing and here is a graph of the main battery tests performed so far. The 3 lines that are mostly horizontal show the battery, lcd and atom voltages over a 3 hour period with 3 batteries in parallel while the top line that drops down rapidly shows 2 batteries in series and the line that ends at the bottom shows 2 batteries in parallel. As a performance characteristic its better to have a gradual drop in voltage than to have to deal with a rapid drop that could effect the music we are playing. More playing time could be possible if some way is found to deal with the lcd logic board going into undervolt at 10v and shutting down.
Circuit building
Posted on July 27th, 2009 With the main power tests over its time to start building a prototype circuit board that will have the primary 12V and secondary 3 – 5V power buses as well as controls and meters necessary to run the atom and lcd system. Had a bit of a look at the logic board schematics to try and find ways of reducing its power consumption and apart from reducing the number of voltage inputs from 4 to 2 there isnt much scope for hacks that dont risk permanent damage to the board. Further investigation will be performed if time permits.
Pico power quick test
Posted on July 29th, 2009 A quick test now of the pico psu undervolt characteristics if its used to provide power to the atom board. Not a long test as after about ten minutes the battery voltage dropped to about the 11.20v mark and the old pentium board used for the test went into shutdown complete with bad bios beeps. So much for the pico spec that quoted 10.50v as undervolt but then its intended use is for a stable current in, unlike the M2 psu that will be used for our gear which is designed to be used in cars etc where power sources are variable.
New video
Posted on August 5th, 2009 We have a new video uploaded to the Video section that is a film we made for the Random Acts of Elevator Music track entitled Lunchtime Meditation Session. The film is an homage to office workers everywhere and the nature of working in an office environment. This film has been shown at the Electundra Music Festival in Melbourne. Also, presently comments have been temporarily disabled, again, due to the tedium of inarticulate spam posts taking up valuable time. This will be rectified as soon as possible.
Solder on
Posted on August 8th, 2009 This is the main circuit for the Suitcase that will carry the lcd and atom board. It consists of 2 dc power buses rated at 12 volts and 3.x volts that provide power to the atom psu, lcd logic board and the ccfl inverter board. Points are provided for switches, voltimeter, lights and the undervolt circuit. A quick 1 volt soak test proved that all is good with the soldering.
Looming task
Posted on August 14th, 2009 The circuit boards have been tested and are beginning to be assembled into their configuration in the suitcase. Here the power bus board is fixed on top of the lcd logic board. The lcd screen which will be free of any components has also had its ccfl leads and video signal leads extended by about a metre, all 20 of them. After all this somewhat boring construction work its time to play the software synths as our first gig is coming up on 1st September, in an elevator, in the city but more on this later.
Updates and news
Posted on August 16th, 2009 And yet another photo of a jumble of wires and stuff but there wont be many more as the construction of our briefcases and music gear is nearing an end. The website has been changed to be dominated by the RAEM logo and information so that its easier to figure out whats going on here. Also we have a new audio clip up that is of an interview we did with John Safran on JJJ in 2006 for that years Fringe Festival.
Plan first then build
Posted on August 24th, 2009 Sometimes you gotta just do it the proper way and this case its working out the basics of the main suitcase that will house the mixer, amplifier, speakers, power and control circuits. The design also has to take into account that there’s finite space available as the main Casio keyboard has to slide over the front of it when stowed. The layout is similar to the one used in the 2006 briefcase and contains the same components with the addition of 12v and 9v power distribution buses. After experience gained from the previous Fringe it became unnecessary to have access to most controls when actually performing as most settings can remain fixed other than a master battery on switch.
So build it already
Posted on August 25th, 2009 With our first performance nearing its time to get building and this here is the main amp, mixer and speaker cabinet that will go in the larger suitcase. Its all a bit raw looking as it still needs to be sealed so that it is a proper accoustic chamber and somewhat weather proof. Next to be built is the atom and lcd cabinet for the smaller suitcase.
The other case
Posted on August 29th, 2009 After a couple of long days building, soldering, wiring and a bit of hot glue later the second case housing the atom computer and all its parts along with the lcd screen have been assembled and tested. All thats left now is to fit this and the amplifier and mixer to their cases and then have a final test. Tomorrow is the day for that as well as another rehearsal for our first gig with the gear at the 2009 fringe festival media launch this Tuesday morning.

Onwards to the tour of 2009