lylehaze

Yes it appears you have done your homework, and done a clean job, too. Your display wiring is very neatly done. That improves project appearance, reliability, and makes troubleshooting easier, too. Regarding your pots.. The +5 and GND terminals will always have the same resistance between them, no matter where you turn the knob. The "wiper" is the moving part, and that's the terminal that goes to the AIN connection. I will GUESS that the "max" setting would be the +5 volt end, but that's an assumption. I haven't built one yet. Encoders: "Follow the documentation" is the correct answer. Or you could just try one and see which way it goes, then wire them all accordingly. That would be a sloppy, unprofessional, and way too easy way of dong it. :-) I continue to be impressed with the quality of panel builds that we see around here. Keep the pictures coming!

Yes. The dimple is pin 1 the notch is at the same end as Pin 1

Hi John, All your DIN and DOUT modules use shift registers to communicate with the core. This lets you get a LOT of IO from a few core pins. If you are looking for a hundred odd inputs and a hundred-odd outputs from a single core, you NEED those shift registers to make it happen. Enough preaching. Yeah, if you aren't using DIN or DOUT modules and you want to get those pins for your private use, you can call MIOS_SRIO_NumberSet with a value of Zero, and MIOS will stop scanning DIN and DOUT signals. That's what I needed for the MBMixer, and the thread we discussed it in is: http://www.midibox.org/forum/index.php/topic,11977.0.html Once that is done, you can re-assign pins as inputs or outputs with TRIS registers, and read the pin values directly from the PORT registers, or set output states from the LATCH registers. This is all straight PIC ASM code. At least, that's how I understand it.. If I've got it wrong, smarter people will step in and correct me. Have Fun, LyleHaze

I swear it's completely true. The classic Amigas have a program called "Soundscape" and one of the tools in there allowed very rapid fullscreen graphics changes mapped to MIDI events. And yes, if you've ever tried to Sync to a drummer that doesn't like to be sync'ed.. it can be damn challenging. Now how can I use this to make my next million? I'll have to sell twice as many if I need to share the profits with Jenna Jameson.

I'll take a few guesses.. You did not say whether the USB power worked AFTER the regulator seemed to die. If the 5 to 3.3 regulator still works on USB power but fails when powered from the 24 to 5 step-down, then you might need to look at the 24 to 5 step-down circuit. First thing to check might be the output filter capacitor. I see they specified tantalum, those can get tricky if damaged.. If the 3.3 regulator does NOT work on USB power after the failure, then it has been damaged. The most likely cause would be the output filter capacitor of the 24 to 5 volt step-down circuit. Either way, I know where I'd be looking first. :-) If you really want to know what's going on, a scope would tell everything rather quickly. Unfortunately, not everyone has one. Good Luck, LyleHaze

http://www.midibox.org/forum/index.php/topic,8954.0.html

Under some conditions, a tempo LED can be very useful. Ever try to get a drummer to "sync up" with a sequencer playback? For some it's difficult, for some it's just impossible. and audio cues cannot be heard over the live drums. I once hooked up a "super-jumbo" LED that must have been almost an inch across. I triggered it from one of the individual outputs of a D-110 rack synth. Yes, there were driver electronics to detect the audio and drive the LED. It worked just as I had hoped, but the drummer wasn't interested in being synched to a machine. I finally solved the problem by looping 24 frames of a short video clip of questionable subject matter, then driving the playback from the sequencers MIDI clock. It's not the kind of thing I could sell commercially, but it worked wonders to to keep the live drummer on the beat, even through tempo changes. I guess anything is possible if you find the right motivation. :-) Getting back to your question: Any software that watches the MIDI clock can control your LED, but a separate LED flash circuit with an audio trigger is more programmable, if you have a spare audio output to drive it with. That way, the flashes can be programmed into the sequence, and you won't need to worry about different time signatures or MIDI clock resolution settings. Or you can try the video method. ;-) LyleHaze

You are right. I have edited the wiki, and it now says: I should probably include a link to his shop, but I'm kinda busy today, and I have little time to figure out how to add external links to the WIKI page. Maybe tonight if I get some free time. I would like to add that Smash has cleaned up the design a bit, and added a LOT of time tracking down better parts for these than I used in the original. This is a great benefit to everyone that will be building with these. It's not a simple project, and some of the parts are unusual and hard to find. Thanks for your interest, LyleHaze

I have seen that kind of failure when RB5/PGM/Pin 38 is allowed to float during programming, but here it is supposed to go to the LCD, which some people leave disconnected anyway. The CMOS inputs will _usually_ float high all by themselves, but it not reliable. RB5 low can make the chip enter "Low Voltage Programming Mode" during initial programming. If I were adding pullup resistors, that would be the first one, ESPECIALLY on a programmer board.. those few moments between erasing a chip and re-flashing it with the "LVP Disabled" option are the trickiest.

Quiet Noisy?? Sounds a bit confusing. :-) Switchers are great for efficiency, but they can create electrical noise that might be heard in an audio signal. Since most of the bipolar supplies needed around here are for audio, I prefer to use linear regulators. But that's just my opinion. I'm really glad it works for you. One thing I really like about 7815/7915 or 7812/7912 is that they are not adjustable, so there's a few less things that need to be adjusted. I just like to keep it simple when I can. Have Fun, LyleHaze

Yes, that helps, Thank You. Interesting to see, as a few people have asked about more than 16 channels of mixer, and my software has been "stuck" in what I thought was the proper way to do it. What you showed me is certainly one way around the 16 channel limit. It's still "absolute", but it gives me a better idea about how others are getting around it. For now, I think my remote control to MIDI receiver will have to use NRPNs to work, and it will only work for MBMixer boards. As before when I made the mixer, it will be enough to suit my own needs. It is a shame that the standard doesn't work for so many "standard" needs, especially where audio mixers are concerned. Thanks Again, LyleHaze

I am adding IR remote control support for the MBMixer. The format of IR remote signals are basically a bunch of "Volume Up" and/or "Volume Down" commands, just keep them coming until you reach the correct level. It is not generally possible for the remote to know or care what the absolute position of the setting is. After looking over the MIDI standard, it would appear that the "Data Increment" and "Data Decrement" messages are exactly what I need. at the first remote key press, I would need to send the RPN or NRPN prefix to indicate that the following increments and decrements are intended to be volume controls. First bad news, there appears to be no RPN for "Volume Control". I can use an NRPN, but then my method will only work with MBMixer boards. Especially since the IR to MIDI receiver will be a separate module, it would be a lot more useful if it could work with other equipment too. I don 't have enough experience with studio mixing boards to know if there is already some standard out there for this. If there is already a "normal" way of doing this, I'll go along with that. Creating a private NRPN is the last choice. I suppose another way out would be to use SysEx to poll the mixer for the current levels, and manage a absolute settings from there. But that assumes that there is a standard way to poll for level settings. So my question: Is there a common/standard/manufacturer specific way to manage this? If so, where can I get more information? Is there one brand of MIDI-automated mixer that is considered the "Standard"? I can make it work "my way", but the more compatible I am with the rest of the world, the better. Thanks, LyleHaze

OK,I figured a way, with a bit of hardware wiring, to do this: The first channel board has the guitar input pre, and it's outputs are Left and Right clean, and the FX outputs feed two possible effects chains. Simple enough.. Now it gets twisted.. Each of the additional channel boards serves one effect, using the FX1 and FX2 controls to mix the SEND (to the effect) between clean and the FX output of the previous stage.. the Left and Right outputs control the mix of the RETURN(from effect) to the audio mix. This effect return is also passed to the FX input of the next stage. Because only the first channel board actually needs its input preamps, we have four "available" op-amps on each channel board unused. This hack will use one of those to combine the clean and FX input signals into the SEND for the effect. Nice thing is, it will work with the existing "full mixer" software, no changes needed there. There will be a few jumper wires between the channel boards, but not too bad. Controls will work something like this: Channel 1 Level and pan are "clean" guitar into the mix. Channel 1 FX1 and FX2 allow playing with the starting levels of the two effects chains.. For each effects channel (2-16): Main Level and pan will control mix from the return from that effect into the outputs (left and right) as you'd expect. FX1 and FX2 will allow you to select the balance of input, from "clean" on FX1 to the output of the previous effect with FX2. Instead of controlling outputs, these are actually mixing up the signal going out to the effect for this channel..(weird, but useful) So, If you set all the effect channels main levels to Zero, Set all FX1 to zero, and turn up all FX2s, you'll get all your effects in series, one after the other. You WILL have a gain setting between each one, using FX2.(turn up main level on the last effect to hear the result) If you turn off all the FX2s and turn on all the FX1s, then each effect will get a "clean" input, and you can control the input drive for each effect with FX1, and combine all the outputs using the main level controls. But the best part is that you can combine and re-arrange the above, doing everything except re-order the effects chain. Not too bad. Now "play" with the settings live with a sequencer.. MIDI control rules! I think this will be the most flexible option. You should be able to remove one wire and handle stereo effects returns too.. as long as you don't need to swap left and right in software.. But we'd need to sum the stereo returns to create the feed to the next effect. One possible problem, some effects have a inverting output, some are non-inverting. Combining their outputs might start cancelling each other out.. could sound weird. Easy enough to fix it, but that is another topic. This could really kick ass.. I hope I have explained it well enough.. Thoughts?? LyleHaze

I have a GIF that I drew, not sure how to insert it in the forum.. already I have other ideas too. LyleHaze

Cool. I get it now. You are setting up all your effects pedals in parallel (side by side). So your "input stage" from the guitar will be driving 20-odd different input channels. Well, you'll have a better chance of doing that if you get to build it yourself, so you don't have to deal with 20-odd impedance resistors in parallel. That part should work. What you will be missing is the chance to connect your effects in series. As I understand it, that's how it's usually done. "Clean" guitar goes into effect 1, effect 1 out goes to effect 2 in, etc.. I figured you could do something similar, with each effect output having a mix into the mains (at that point) and also 2 outputs into the "next" effect(s).. That would be some kind of Pyramid thing.. Channel board 1 feeds into Left (clean), Right(clean), and the FX1 and FX2 outputs each feed clean into two different effects inputs.. Say, for example, a compressor and a reverb unit. Each of those effects outputs(mono) can be mixed into Left, Right, and branches into Two more possible effects chains... Not every effect would HAVE to split wider, but the potential is there. You would end up summing a wide variety of effects levels, from pure, clean from just the first channel board, to a endless combination of effects feeds.. All of this could be done by just removing the stackpins from the C and D mix outputs(FX1 and FX2) and wiring each boards FX outputs to other effects. What you could NOT do is to re-arrange the order of the effects, they would stay in the same sequence unless you manually re-connect them. Either way, it would be a lot of fun. Discussion Please? I'm not a guitar player.. but it seems like it could get really diverse. Have Fun, LyleHaze

OK, I finally had a chance to try and follow your signal path.. One thing you my want to change: You said that some of the mixer outputs would be going to the guitar preamps. I think.. A "proper" guitar preamp will take a low level (high impedance) signal direct from the pickups, and will output a "line level" (low impedance) signal. All the signal inside the mixer are line level. We can mix high impedance and balanced inputs only after converting them to line level. That's why the input sections are so configurable. Taking an output of the mixer (line level, balanced or unbalanced) back to a preamp may not be what you intended. As we talked about before.. getting your signal paths figured out will take some work. I provide setup for a "line mixer" and a "Full Mixer" and you can mix and match those as desired. Figuring out all the options for what you want might take a bit more work. To review, for others just reading this thread: A "line mixer" mixes a stereo pair (Left and Right) into the main Left and Right outputs, no effects. You can handle two pairs per channel board. You'll use all four input sections, output to just A and B. A "Full Mixer" takes a single input, and mixes it between Left, Right, Effects1 and Effects2. This requires a channel board for each input. You'll use just ONE input section, and all four outputs (ABCD) Your may mix "line mixer" and "full mixer" boards in the same stack, and configure them by MIDI when you start the board up. If you look closely at the "Channel Board" diagram, it has (up to) four input connectors, "-1+, -2+, -3+, -4+", up to four input sections,each with five points to configure them, four volume controls (inside the PGA chip), and four possible destinations to "mix" into. You wire the inputs and outputs as desired, and configure the rest with control change commands. All four mix destinations (A,B,C,D, usually left, right, FX1, and FX2) are usually ganged together by stackpins, but the user can separate them when building if you want. Enough review. These are things I expect to be answering a lot of questions about, so I wanted to get it posted here. If there are suggestions to how I can make it more clear, speak up! Have Fun, LyleHaze

@Orkspalter The boards are sized to fit in a 1U case, no matter how many are stacked together. Each stack layer adds less than an inch to the total size, which then LAYS DOWN in the case. That's why I made the boards so narrow. Notice these boards are "mounted" with double sided tape.. it works great. of course, my boards are not yet available, I can't fix that today. The signal routing sounds pretty confusing, I'll be back to read more after breakfast and mowing the lawn. :-) Using the same transformer should be fine.. just use separate 5 volt regulators for analog and digital. I think the two windings would be positive and negative, though the positive will also be feeding the digital stuff. So in this picture.. both CT connect together (Ground), you'll feed both 7805s from +, and feed the 7905 from -. 2200 uf should be great, I always shoot a bit over on filter capacitors. Have Fun, LyleHaze

very nice.. Having the compartments in the case will be good for shielding, keeping the digital noise out of the analog signals. That display looks HUGE! Since you're making a mixer, I have a display question.. What do you get from character codes 0x10 to 0x18? (dec 16 to 24). If you get vertical bars moving from left to right, you'll be able to use the graphic level display. There's also a Db display that will work on all display types. You CAN run the mixer software before connecting the PGA chips, if you want to check it out. You may also connect a encoder if you make a DINX1 module (just one chip). Looks like a great start. LyleHaze

I'd say the most difficult part would be choosing the parts.. When you build a panel, you get to wire everything as needed back to the modules. If you want to use a PCB for that, you run the risk of making the PCBs useless if one of the parts you use becomes unavailable. Including DIN and DOUTs should be workable, many here have done the same for their projects. On the bright side, it's easier now than ever to get high quality PCBs made for a hobbyists budget. Do you plan on building a prototype first to make sure you have the features and wiring all worked out before committing to circuit boards? Have Fun, LyleHaze

From the photo and link you posted, it looks like a fairly standard bipolar supply. They appear to be using 7812 and 7912 regulators, and a wall transformer that is capable of about 300ma. If you need + and - 12 volts at 300 ma or less, yes. However, you'll probably need a separate power connection for the core +5 volts. I have not looked over the CV project too closely, but I assume it has a core, and maybe a display too? If you choose a transformer with larger current capacity, you could power everything off of it, including this board. Most bipolar (positive and negative) supplies require a center tapped secondary on the transformer. The link you provided says :Price: $41.45 Then avoid nILS. No, seriously.. PAiA have a good reputation. I think $41 is a bit expensive, but I'm sure the product will do what it claims to do. Have Fun, LyleHaze

I'm not sure if you are designing a mixer or a control surface. I believe the MB64 or 64E would lend itself well to the control surface, and you could configure that to match any digital mixer you like, as long as it accepts MIDI messages. If you are actually designing a digital mixer, I think that pic chips are probably a bit underpowered for the job. Some fool around here was talking about designing a solid state analog mixer with digital controls and no pots or mechanical parts. I believe he went insane before the project was finished. Circuit board design is something that a lot of us have learned around here. I'm not the best, but I have managed to get a few working designs out. Many people start with the free version of Eagle.Eagle is only free as long as the board is small, no more than two layers, and it's not for profit. Others use.. umm, protel, I think (I'm not sure). Have Fun, LyleHaze

Greetings.. I regret that I don't have time to thoroughly answer your questions.. But I can toss you a quick answer.. Somewhere in the mios source tree is a file called macros.h On this machine the end of the path is MIOS\mios_base_v1_0b\include\asm\macros.h (OK, so I probably have an old version..) This file defines a variety of useful macros, including skpz #define skpz btfss STATUS,Z It's worth reading the file to see what else has been done for you. Have Fun, LyleHaze

I would leave out everything from J1 to IC3. Connect the power supply GND and +5 directly to J2. BEWARE: DO NOT connect it to J2 backwards. It would be wise to use a fuse too, as 6 Amps is WAY more than your core should ever need. Just my opinion, there are many people here smarter than I am. :-) LyleHaze [edit] If you want to add a "safety", get a rectifier diode (any common type, maybe 1n4004) and install it in pin2 and pin3 of the 7805 holes, with the cathode (striped end) on pin 3. This will have no effect if the power supply is connected right, but it will blow the core fuse and protect the core from damage if the power supply is connected backwards.

If you're looking for isolation, optos are hard to beat. If you want to "logically OR" more than one trigger source into a single trigger input, you can also try open collector circuits. They allow "mixing" without output overcurrent issues. If you need a voltage conversion, and optional polarity inversion,you might want to use an op-amp wired as a comparator.. someone else did that here recently to trigger a family organ drum unit. Have Fun, LyleHaze

ASSUMING: this is to feed a 7805 regulator, depending on the brand of 7805 used, the minimum acceptable input voltage might be as high as 7.5 volts. LyleHaze