tonyn

OK. All connections from terminal strips to Dins play notes (I used a jumper wire to connect the Din side of the terminal points to the grounds of the barrier strips, bypassing the push button switches). So there maybe some bad switches or I need to make some adjustments with the key hammers hitting the push bottons correctly, etc. Damn version 2 with touch switches(my thoughts on this are back a few posts) may be next. But right now I need to find out why my syx files are not updating the core... They compiled right and seamed to upload with both midi-ox and midosstudio, but maybe not.. It can't be a bad midi input to the core, since the midibox128 setup hex file uploaded, and I wouldn't have note presses if it didn't, etc.. I need to look over the midiosstudio and midi-ox configurations again.. I am off to bed now. If you guys have any inputs or suggestions for me, I would appreciate them.

OK both dins are working. I had to shift the IDC connectors to the right as in these pictures. I knew I should have used 5 pin female connectors for this connection. Well, I can do that later. Right now: #1 there are missing keypresses on both dins(i.e. some keys work, others do not, this is verified in the monitor in of midi-ox). This may just be bad push botton contacts(I'll test this out later), etc. #2 The notes that play in Miditzer are not in the right places, etc. #3 I can't seam to update the syx file on the core. Right now the lowest C note on the keyboard maps to octive 7 note G# (this note plays, but is beyond the Miditzer keyboard keys). The highest C note on the keyboard maps to octive 3 note E The other notes are scattered(some are in order, some show up in weird places).. The highest note C, maps to pin 5 in midibox. The lowest C on the keyboard maps to pin 57 in midibox. The main thing is being able to update the midibox syx file on the core. But it also seams that the Dins are out of order as to how the core sees them (i.e. the first din J6 is the higest pin in midibox, the second Din J3 is the lowest pin, etc.) So I may have to think on how to route them different remount the Dins (i.e. flip them), and/or remount the core, etc.. I guess the order I planned didn't work out as I hoped. There is enough cable length, etc., to change things around a bit. But right now I just want to update the midibox syx file on the core to remap the keys for now. I am exhausted. I may call it quits for today and aproach it tomarrow. In the meantime: Any ideas what's wrong guys?

humm... the midibox128.syx file doesn't seam to be uploading? No updates to the file on the core.

OK. I am going to start by reconfiguring the midibox file to map the key presses right...

Talked too soon. I just connected up my first din to the second and don't have anything from it.... The second din still works with key presses(but not all notes play) to the core with the first din connected to it, but no note presses from the first din. Humm.. Maybe it's just the mapping in the midibox_syx file.. Maybe I should make long IDC cable for testing to just connect up the first din to the core to see if it actually works. I am confident it does. But it won't hurt to make sure. If so, it's either in the connection between the Dins, or the maping in the midibox128_syx file. So it partially works.. The keys aren't mapped right for Miditzer, so it must be the mapping.

OK. Take note: Don't get an underpowered PSU! That was what it was! Now to finish mapping out the keys and we have a playable keyboard for Christmass! I'll share my midibox codes with you too once I finish them.

We have a virtual organ guys(well at least one of my keyboards works with it for now)! I have Miditzer Style 216 running with the Miditzer keyboards set to channel 1. I can play some notes guys! This is only with the one din, and I may need to reconfigure the midibox128 file, etc. So now to connect up the other din.. My son is happy, I am happy!!!!!!!! He is playing the keyboard right now. :hyper:

OK. Core is remounted, using my computer wires. These are 22 gauge, and should work fine. I should have just did this to start. Those wires are more flexible to route. I will remount the fuse holder and fuse later... As you can see, only one din is connected right now. It's all powered up, connected to the computer, with a good sysex message. So all is well so far. I'll upload a midibox128_syx file next, and test out the keyboard. Once one din works, etc., I'll connect up the other. One step at a time...

Time to remount the core onto the switch board, connect up the dins to the core, and see if it still works.I am also risking doing this unfused right now too. But I also have a feeling that the reason the fuse blew was due to the PSU itself being overloaded because it couldn't handle the current draw, etc. I'll refuse it shortly(after the dins are all connected up, and everything is working, I'll check the total current draw and put in an apropiate fuse). BTW: The first optocoupler had been reversed, so I can't trust it, and it's going in the trash can. Even if it was good, it may have been stressed out with what I did to it. So it may go eventually anyway. So, to play it safe, I consider it bad and it gos in the trash, so I don't inavertingly use it.

Somehow I had to set the id to 1(that's because this PIC is preprogrammed for channel 1). I couldn't use the "wait for upload request". But I was able to use "Smart Mode - use feedback from core" So it has to communicating with the PIC OK. Seams to have uploaded the hex file...

Great!!!!!!!!!!!!!!!!! My core seams to be fully functional! :frantics: I took a risk by using the American PSU, just popping in the ICs, and connecting it to my organ computer with midi-ox open( with the monitor in open), and I have a sysex code from the core! So the problem was that that other PSU wasn't putting out enough current! But I am worried about the 12v part. Well sometimes it is something simple. Now to see if I can upload the hex file to the core... I'll change over to midiosstudio for that, and set it for automatic etc., to see if it gets the requests from the core, etc.

I have a funny feeling that it's a PSU! So: I am going to risk this: I am going to use the first core, which may or not be bad. Hook it up to the American PSU and test voltages, and see if under load it puts out 9v. If all is well, then I will proceed with that one and hope my problems were the PSU. OK: That PSU under load still puts out 12v. But my 5vs are good. So, do I Continue with it or?

Here's my test setup. The red wire is the + side of the power adpator. Here is something strange that I found out that has nothing to do with the core, but... The right power adaptor has been the one I have been using. It is rated 1200ma MAX, and can be set to different voltages. I did have it set to 9v. It does put out 9v non load and 9v load. But it is possible it doesn't put out enough current etc... The left power adaptor was a so called 9v 1A that I picked up at American for the cores, eventually. I checked it out non load, and it outputs 12v! On it it says output 9v 1a! So????????? Well I'll set it aside for now.. I don't want to risk using it if it is too much for the 7805s Damn! Well voltages and grounds test out fine(btw this is the second core I built). I will use this core first, and if it is all good, then I will know it is good, and then go back to test the bad one. Get one good core running, and then... I will check out currents next. put in a PIC, test current. Put in opto(cross fingers), test current. It is JUST possible that the power adaptor isn't putting out enough current? maybe that max is for the 3.5v, and when set to 9v it isn't putting out enough? I know the one I got from Amaerican(because it's heftier) should put out enough. But at 12v? Will it put out 9v when under load? Do I chance testing it out with this core, just to have it blow something ?Delema...

OK. Just in case the first core has a bad power circuit that I couldn't verify with the ohm meter: I am swapping out cores for now(just in case the old core has bad 5v circuits, etc.,i don't want to blow another opto, if that happened). To make testing easier I just cut a 2 conductor female connector from an old computer to use. I will first power up without any ICs, test out the voltages and current draw. Pop in the PIC and check for current draw again. Put in opto, check for current draw again. Pull out PIC, follow trooubleshooting.

Thanks, that's good to know about the PICs. I hope that I didnt blow an opto. Anyway: I just finished comparing cores, ohming out all of the traces etc., and verifying that all components are good(except the ICs). As much as I can tell with a multimeter that is. Everything on the board looks good! ??? So now it's removing the ICs, applying power to it again, and testing the 5v power circuit (again, initially I did test it out, but this time it wil be more thorough)... I will also test for current draw. I will do this for both core boards(If the second is good, I may just swap it out first just in case the first has something blown that I can't tell) I am just wondering about the PSU. Although it shouldn't matter the polarity to J1, it is still a DC supply and it may matter? OR do I make some Rx/Tx loop backs to make sure what gos in should go out first? I am so parinoid at this point to insert an optocoupler just to have it blow. Can I bypass that to test it out first? I want the LCD part working, but that will take time to build... But I feel a bit blind without it now.

OK. Here is what I am doing right now: I am desoldering the power connection to the J1 header pins(it's tacked onto the outside of them). I'll be pulling my core off of the switch baord and comparing it to my other built core, etc., off the switch board(good thing I made a second core board to use). All of my trouble shooting will be with the core itself not mounted. One I have it fixed and working, it can then be remounted. Fortunetely I also have a couple of non soldered core boards to look at the silk screen and use for tracing out the traces, etc. I may also switch over to my other core to swap out, and then go back to see what's blown, etc., on the bad core. Nice to have a board to swap out, etc. I will also do as many tests, etc., without the ICs in. I hate to think that I may have blown the PICs or optocouplers!I do have 2 non touched PICs and 2 non touched optocouplers. I did swap out 2 PICs and 2 optocouplers the other night. I also reversed one optocoupler thinking it wasn't oriented right(pin 1 is in the oposite direction as to pin 1 of the PIC on the baord, so I thought..).Yes, stupid me. The optocouplers have just a dot on them for pin 1, the PIC is notched, so the PICs are easy to figure out the orientation(so those were not reversed). But that optocoupler may for sure may be blown, so I have it set aside for now... If those are blown???? I will only use the non touched when I KNOW the rest is fine, etc. I hope no PIC is blown, since to order more of those, preprogrammed from avishowtech, would take time. I do have optocouplers on order. But I don't want any of those blown either.

My mind is always planning out a better mouse trap. So I need to put my thoughts down before I forget: You know what? I didn't have to use push botton switches for my contacts for my keyboards! The backs of the keys are metal and are screwed into metal. So in efect they are all tied together electrically for a ground side! I could have used thumb tacks on the switch boards for the other sides! The problems with the push button switches are that the springs in them are a bit stiff, and the keys are a bit stiff. If I take out the springs I have great feeling key presses. Using thumb tacks(or metal screw heads, etc.) would be a way to make cheap touch switches! Those would last forever too! See the keys already have springs in them, so adding extra springs isn't needed(nor wanted). The old contact boards had springs that weren't as stiff. My push bottons are a bit stiffer than the old contact board springs, but they aren't too stiff. Just a bit more than with the old contact boards. My organ guy said that baldwins were a bit stiff anyway. But without the push bottons my keyboards would feel more like piano keyboards, etc. Well like I said: This is version alpha. Once this is working I will be working on improving it. Just thoughts... Now back to my troubleshooting!

Wow, thanks for the inputs guys! I will try to answer them all in this post. My PSU is a cheap 1A wall adaptor with 9V DC. This should be fine for the 7805. My crsytal is soldered in, but I did use a hot soldering iron(it was set on 40w instead of 25w, which is recommended for circuit boards) When I did do the loopback it didn't work, but then at that time I may have been working with a blown fuse so my core may have been dead. Yesterday I went out to find an LCD module at American Science and Surplus(It's a nice place that has surplus parts that are cheap). American Science and Surplus bought out Edmunds Scientific years ago. Edmunds was around sicne I was a kid and had lots of lens, etc. Furtunetely it's located locally to me, but I think they have a website. Anyway I went through their parts bins of LCD modules. I saw a couple that came from cell phones, etc. Nothing that I could really use other than 2 that don't have driver circuits. I grabbed those for $.40 ea. They have nice legs that I can put into prototype boards, etc. Plus they have part nuumbers that I can use to look up the specs for the pinouts (I hope, else I will have to test them to see what's what). I'll see if I can use them with a minimum of parts to the PICs of the cores. If not I'll hunt around again for my LCD modules I KNOW I have somewhere. I have too much stuff that is unorganized in tons of boxes to go through! But they were supposed to be with a project that I was working on for an automatic seahorse feeder (one of my inventions). But they weren't. Those were complete modules with driver circuits I got from Parrlex I think. I want an LCD module at this time for troubleshooting, since even if I don't get an output to the midi and optocoupler, I should have a reading from the PIC on the LCD, telling me that it is at least working, right? Now I need to build an LCD module to use for troubleshooting! I was hoping plug and play, but I guess not. Here's what I got (if I can't use them for the cores, I can use them for something, at $.40 they were a steal):

I know, but thanks for the input. I thought 64=64 keys max, so midibox64. I did start with the midibox64 until I found out that midibox128 was more for organ consoles. So, yes, I did find that out and was working on the midibox128 instead. I do have a midibox128 ini file I am working on. I have midosstudio, midi-ox, active perl, and makefile script, all configured and setup. I even wrote a bat file that I inserted into the midibox128 tools folder where I have the makefile script and midibox128.ini file that makes it easier to make the syx file than opening a command prompt, and cding to the folder. I renamed the midibox128.ini to core001.ini Here's the code: You can make that simple file in notepad. Save it with a .bat extention and it is a simple executable file. Just double click on that file instead of using a command prompt, etc. I'll share some of my coding with you guys when I get back to that part. Batch files are nice little executable DOS files you can make to help make things easier. This shows how old I am(I have been programming with DOS/Windows since the beginning). DOS bat files have been a part of Microsoft since the beginning, and they still work with Vista(my organ computer is running Vista right now). There are a few commands for them, if you want I can share those. "pause" is one, and just pauses it so you can read it before it closes the command prompt. These only work on Windows/DOS operating systems. When you want to compile for another core, just make another core00x.ini file, edit the .bat file for that, etc. Makes life a bit easier :) That was until I found out that the core wsn't actually working. So right now I have to have my electronics tech hat on instead of the programmers hat :)

Great! Thanks for the input. Ok I can rule out having to top solder then. Yes I have been careful and methodical, which is why it kind of upsets me that I do have problems. Yes, I am following along with that trouble shooting guide, thanks. I have several browser windows open, and a couple of good pdf files that I grabbed and have been following for troubleshotting, assembly, etc., from this site, or the midwtzer, etc., on the organ PC(that PC for the organ also connects to the net). I am worried that I may have blown an optocoupler. Yes, I swapped them out, but that was last night when I was a bit exhuasted from getting to this point, that I may have blown the other one I swappped out too. I ordered 4 more just in case from a local vender. He has an order going out so I'll have 4 more spare by next week. But they aren't cheap: $5 ea! I am down to 2 spare from my other core kits that I can still use for swapping. But I need to be careful not to blow them. The crystals are a bit of a concern too since in the how to I don't think he soldered them, since he felt it would stress out the crystals. Just insert them without soldering, and they would make good contact? I don't have an O-Scope so I can't test for Frequencies, etc. Best I can do is what I can with a multimeter. To me I can't see the PIC being bad, at least I hope not. But it could be a cap, resistor, crystal, or the optocoupler. Or the board. The optocoupler is directly related to the midi out, so that area I will really investigate. But like I stated in my last post, I will thoroughly be testing everything out. In the trouble shooting guide it says that someone had problems with the ground on the midi out. There is a pull up jumper area on the circuit board. It's shorted now to keep the ground. But if I need to pull it I'll have to cut the trace. To reconnect it I'll have to insert header pins and use a jumper. So I don't want to do that unless I have to. Today I am going to American to also see if I can pick up a cheap LCD module to help with troubleshoting now. I have my core isolated from the Dins, but I want it isolated from the computer too just in case...

All of those tests last night may have been worthless since today I found that the fuse was blown. So I replaced it, and another blew. I can't see where it should draw more than .5a , even from the power adaptor side which could draw more than the 5v side. I temporarily bypassed the fuse, and had voltages, but no output. So it wasn't as simple as that(sometimes it is, but not this time). I may have damaged the optocoupler or PIC last night. I hope not. I have only a couple spare that I don't want to damage in my trouble shooting either, since it's a short supply. There is definately something causing this. Maybe the power circuit on the board is bad. I do remember bending the 7805 while doing something last night. Oh, well... What should I do next? I am about to the point of VERY THROUGHLY VERIFING MY SECOND CORE, TRACING EVERY RUN, OHMING OUT FOR SHORTS, OPENS, TESTING THE PASSIVE COMPONENTS, CONENCTING UP TO POWER, VERIFYING EVERY POWER POINT, PUTTING NEW ICS IN IT, CHECKING CURRENT DRAW, ETC. To this point I just did a minor checking. Now they will be thorough! At least I have another built to use to swap with. But I need to be careful to not blow anything on it, since the optocouplers are easy to get, but I don't want to have to blow any. The PICs are a different story... Why me? Everything was going good up until this point. I thought I was thorough in doing it right. Well I don't give up, I will get it working eventually. But at what cost in blown parts, etc.? I guess this is the time I need to share with you the bad parts and what I find I did wrong, etc. I'll let you know what I find that I did wrong so you know too. My mind doesn't work like it did 20-30 years ago when I was an electronics tech. Plus I am a bit rusty with the electronic part, since I have been mostly programming the last 20 years. So it takes me time now... So if you can see something I may have overlooked, please give me your inputs. Christmass is coming up and my wife wants me to clean up my mess in the living room to put up the tree, etc. So I have until tomarrow to find this problem and get it fixed, else I will have to pick up next week, etc. I was hoping to have the keyboards working by this weekend for Christamss too. Oh well...

Damn! You may have thought my next post would be a good one, but it isn't! What seamed liked a sysex message last night from the core wasn't. That upload went nowhere. I am not getting anything out of my core! I have been in the troubleshooting area since last night. My usb-midi is the turtle beach, and it was already working fine with miditzer, another kwyboard, etc. So that's ruled out. My software is all instaleld and configured right. I ruled out my software and usb-midi last night. I did the loopback tie of rx/tx and nothing. Swapped optocouplors out, PICs out, etc. Put an LED on the output pins of the midi and nothing. My voltages grounds are all there, no observable shorts or opens. I DID THOROUGHLY VISUALLY CHECK THIS CORE OUT BEFORE PUTTING IT ON THE SWITCH BOARD! I am confident in my soldering. So what the...!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! I am kind of thinking crystal? In the how to it said to not tack the crystal as not to stress it out. I soldered it in? But the jumping of Rx-TX shouldn't have needed a crystal. That should have just looped back? I don't have an O-Scope to test the cyrstal out other than swapping it out, etc. That LCD display module would have saved me a lot of time last night, and even now... These are simple circuts too. They should have plated through holes right? I did no top soldering assuming that. Now I neeed help! Any suggestions as to what could be wrong would be appreciated. In the meantime I'll wake up more today(I just woke up), try to clear my head and put on my technician hat.

OK. Step 1 seams to have worked. I had to upload the midibox64 hex file in manual mode. Now to configure it and install active perl.

I just downloaded Miosstudio and conencted up the core, and have the propper sysex message. So core is functional! Now to download midibox64(not midibox128, since I only have 2 dins) to the core...

OK. Nothing blew up! :) Just wait, I smell smoke! :shocked: OOPs that's my cigarette. :D I have 5v everywhere I need it for the ICs for the cores and Dins. Nothing seams shorted, so time to mount the ICs and connect the core to the computer! Now, where is that great midibox software located here for download... I guess I'll need active perl too eh? Well time to look for what I need to program the core now.