OK, this gets really geeky. But I think it

OK, this gets really geeky. But I think it's interesting and a bit like a history lesson in syntbyggeri. When the Japanese instrument builders Roland released their first MIDI synth JX3p 1983 saw many it afterwards as an omen - the analog era was over. JX3p had no regular "analog" parametersliders where each slider influenced a parameter. bree daniels Instead, it was replaced with a complicated editing spools systems where the buttons chose the parameter to be edited and then pulled it in editslidern. This and the digital standard MIDI led many to accuse JX3p to be cool and edgy. Not nearly as hot as the half-brother of Juno-60 that had come out the year before. And certainly bree daniels sounds different, no doubt about it. Then there are a number of myths and misunderstandings about why they are different. bree daniels One of them - to JX3p would have DCOer ie, digital bree daniels oscillators vs Juno 60ns analog VCOs - is the myth I thought put under the microscope here. Already in the overview description (above) of the Juno-60n oscillator will the myth of the all-analog VCO to shame - it says "DCO" as part of the sound generation. DCO in this case means Numerically controlled oscillator and is in fact the further integrated programmable counter circuit 8253 which includes three 16-bit counter that clocks the Juno 60ns VCOs. Juno-60n VCO is thus governed by both an analog voltage - CV for the control voltage - AND clocked by a digital clock in the shape calculator 8253rd This was a very important discovery bree daniels that I think Roland was almost bree daniels the first and Juno-60n is often described as The first polyphonic synth, you could turn on and start playing immediately, without waiting for it to warm up to the mood would hold. It is therefore not surprising that Roland re-used the technique in his later polysyntar. Here (above) is the corresponding elements described in JX3pns service manual. First, the Juno-60 already familiar description of the digital clocking meddelst counter circuit 8253 and then a close-up of the wiring diagram that shows how VCOerna clocked and CV controlled 4558orna. Here it is interesting to see how the Roland techniques, presumably to save money, skipped the entire building with variable pulsgenereing based sågtandsvågformen by a comparator, etc. that appear in the description above. Instead clocks to the digital pulse from 8253an through a NAND gate for achieving a symmetric pulse. And indeed: In JX3p are true-PWM is not really on the map, but it occurs rather than as an additional fixed "tighter" pulse waveform which is probably generated by varying the bells against visible CV, or you can "cheat" presented on the oscillator 2 by sync it to oscillator 1. Then there are some other interesting "save points" as Roland put on JX3p vs Juno-60. It uses the example, not Roland's proprietary VCA BA662 they really sprinkled with Juno-60 bree daniels but uses a less expensive (and certainly less dynamic) transistor solution instead. A little inspiration and thoughts BankAbout this post, I initially downloaded from Sequence15-blog's briefing on the Juno-60 vs Juno-106 where a bit the same thoughts when it comes to the VCO vs DCO are represented.
Power sequence15 ps3bloggen Patrick Fridh blog Olaf's blog musictechy Neriks Blog My syrras blog Matrix Synth little-scale Lisa's blog Jean-Michel Jarre For musicans only EMS EAM snippets densistabloggen Brusa little higher on the Analogue Sweden
2011 (1) January bree daniels (1) 2010 (6) October (1) August (1) June (1) February (3) 2009 (25) November (1) September (3) August (1) June (3) May (4) teaser UASSO The Droids vs Moloko Magnetic Fields Juno VCO 60 vs JX3p DCO April (3) March (4) February (5) January (1) 2008 (38) December (1) November (2) October (3) September (7) August (2) July (1) June (3) May (4) April (4) March (5) February (2) January (4) 2007 (65) December (3) November (8) October (3) September (3) August (1) June (7) May (4) April (5) March (9) February (12) January (10) 2006 (32) December (9) November (23)