bilderbuchi

could you maybe post them in the thread, so other people could also learn some coding efficiency tricks? :)

sry4lazyness.... 3-5€ apiece, bleh...

would maybe be nice for a monome-style BLM, although not soft silicone. how much do they cost? :(

ehm, no there isn't.... ;) at least if you count reports from people who know what they're talking about (i mean i can blather on all day about chevy truck transmission boxes, only i know next to nothing about them)... it's mostly that the deniers get bigger press coverage which leads to this skewed perception.... i recommend reading www.realclimate.org

so any updates on this nebula?

nice timing! :) i'd need the 2 preburned PIC18F4685 and the Core for my project http://www.midibox.org/forum/index.php/topic,13094.msg111912.html#msg111912 , and please also add one Din and one Dout kit. would ship to vienna p.s.: maybe you can also drop in one of the sampled pic 4685, as backup in case anything goes horribly wrong, or i start yet another project :P

thanks for the infos guys! i think my path is clearly outlined now :P

thx. yes two cores linked by CAN seems to be the most prudent and elegant way. :) a quick check of the specsheets yesterday night revealed that the pins for CAN on the 458 are occupied by the LCD connection in the core pcb. if i understand that right, i couldn't use a 458 on a normal core module? or is it possible to internally reassign CAN to another pin in the 458? i'll be looking into that MB_NET stuff.

for 5 coax multicore, you could use RGBHV video cable. dunno if they sell it in so short length, i've only ever seen the 50m looms, or maybe 20m ones :-P maybe you got to pay attention to impedance etc., i think these may be relevant to you: http://www.midibox.org/forum/index.php?PHPSESSID=c43cebc93f105690a49ba30c21b1f6a9&topic=6570.0 http://www.midibox.org/forum/index.php/topic,13094.0

yeah i kinda expected that answer.;-) was just hoping to plan accordingly. if it comes out in two months i'll wait and go stm, and build a midirouter/simple step seq beforehand), if it takes another year that's too late for me, that's why i asked. anyway, point is that, as i understand it, the STM core has a CAN transceiver (and presumably the routines to access it easily) built-in, that's the advantage i gain. with the old one i'd have to learn how to rig a transceiver together myself, and program the routines to be able to use it (in C since i don't speak asm). i'd have liked to spare myself... well, rumor mill, it's posted on the ucapps mainpage, after all...if that's no tease, i don't know what is ;-) cimo: sounds reasonable, i'll look into it! at least 9 cables is less than the 24 (or 16-18 depending on wiring) which are necessary if i physically pull the switches into the stage box - this is the case with the current control box for 8 projs of a different make (short press: fwd, long press: rew), uses an LPT cable... fat and stiff, a real joy to rig. :P

lylehaze: projectors are not close together (~10-20m, depends on the setup), but this does not matter because the distributor box only closes an electrical connection (via reed relais) in the projectors. moving the midibox up into the distributor would be a possibility, but i would lose the display and control via (hardware) buttons onstage. yes, i could add another midi controller or control solely with software, but both options i don't like. also, afaik midi connections are limited to 15m cable length? CAN is a neat idea! a quick search showed these interfaces are quite expensive, though (1-200€?). or can i use the CAN capabilities of the upcoming STM32 core (as the spec sheet seems to indicate)? CAN would be beautiful because it would be possible to use XLR cables! :) also, maybe SPI would be better? or is CAN the optimal bus for my application? edit: seems the cost is not so high after all? the IC's themselves are quite cheap. i have to see how easily they would be integrated into my system. well, more reading necesssary.... thanks for clearing that up about TTL. now i know i can drop the initial concept. stryd: thanks for the confirmation. a PC at the distributor end is unfeasible i fear (too big, clunky and sensitive), i need a sturdy little box. if CAN of the new core is not possible maybe i find an alternative protocol is usable (XLR/DMX or OSC). well, thanks guys for the input, i'll read up about CAN and the new core. is there an estimate when will it be available to us mere mortals? i will definitely keep you posted, and i hope i will have footage of some nice projector-synching goodness at the end :)

Hi folks! I've been lurking on these forums a couple months now, but now I've run into problems I can't solve by just reading (which i did alot lately)... Intro: I am currently building a Midibox as a midified remote control for a couple of old slide projectors (for event/visuals use). button press or sending midi notes should actuate fwd/rew on the slide projectors via reed relais. The Midibox resides in a stagebox with a couple buttons, and a distributor box (with a Dout and an array of reed relais) should be up in the rigging near the projectors. Sending the Dout signals over the long connection shall enable us to only use a 5-lead cable instead of the 24 leads (or more if scaled up) needed for 8 projectors. i have estimated that a cable length of up to ~50m is necessary for our purposes. The box is built already, i got the software operational so far, but now i've run into a problem i didn't expect: Problem: Everything works as intended if i only have a short (30cm) connection, but now i have connected the core and dout with a 100m long 5x0.14 braid shielded cable (still coiled up if that matters; resistance of one lead is about 10ohms; i didn't cut it just for trying out because it's (still) a friend's cable), and now nothing works anymore. i can't actuate the reed relais and leds i connected anymore (sometimes only flickering), and even the Din readings of my box seem to be affected if i interpret the LCD readings right. Troubleshooting: I figure that the long cable length somehow interferes with the Dout signals. i found an old thread talking about cable impedances etc http://www.midibox.org/forum/index.php?PHPSESSID=c43cebc93f105690a49ba30c21b1f6a9&topic=6570.0 , i am currently reading the linked pdf. i could kick myself that i didn't think about this earlier. :( I reconnected pins, so that power is fed to the Dout over a short cable directly from the core's J2, figured that maybe the long run let the voltage drop too much, didn't help. I connected the cable's braid shield to the power grid's earth, changed the behaviour slightly (more flickering, chattering relais) but still erratic. It should be earth for shielding, right? not the Core's ground? I have no easy access to an o-scope so i can't check the signals before and after the long cable. Question: Is it even possible to let a Dout signal set run over such a long distance? Would it be possible to reduce the update/multiplexing/SR frequency or whatnot of the Dout signals? I don't need ms-accurate timings, ~half a second would be way enough, so maybe this could be relaxed? Do i maybe need special circuitry for boosting the signal before and after the cable? or would terminating the cable correctly resolve this problem? Is it feasible to use other circuitry to enable another/slower switching method for driving 16 up to 32 reed relais? meaning, stuff the Dout into the stagebox, use the dout to drive some other elements, which are more robust with regard to long lines, while also minimizing the number of leads needed for the "long haul"? Well, thanks so far for any answers! edit: tried sending the Dout (-power) over 10m XLR cable, works much better, although some strange things happening (1 button actuates two leds, one flickering), so i think signal degradation is already setting in here. the reason i also want to transmit power of the long cable is that, while i could build a PSU into the distributor box, i fear that different ground levels (with switching psu's, or if the distributor is connected to another phase of the power grid) would also cripple operation. using 3-pin XLR cable would be best because of its ubiquity, but the ground level question remains. isolation with optocouplers is not feasible because there is no return path available in an XLR cable (3 leads, no shield).