Hawkeye

Whoops, you are of course totally right, the common pin is +5V, not GND (not enough sleep, as usual, hehe :-))! Many greets! Peter

Hola! Wild guess, if you are using resistor networks on your DIN module, can you check that these are of the required "bussed" type? The resistor network needs one common pin connected to ground 5V, this pin is connected to every other pin via individual 10k resistances - you can measure that with a multimeter... The described behaviour sounds like you maybe have got the wrong type of resistor network, where the 10k resistances are actually between the pins (but not going to common ground 5V). Many greets, Peter

Ok, here is a simple example: 1) define the variable in exactly one module (e.g. in this case "hardware.c") file u8 led_startstop = 14; 2) declare the variable as "extern" for use in the corresponding header (e.g. in this case "hardware.h") included by other modules (other .c files) extern u8 led_startstop; Now you can use the variable in the original c file, as well as in all other .c files, that include the header. Good luck and many greets! Peter

Welcome back, Smithy, how is Ireland? ;-) Wild guess: WT1 BEG = WT 2 BEG, so both views map to the same memory byte, modifying one will also modify the other. Redemption: set WT2 BEG to a different value! :-) Many greets, Peter

Fantastic job! Happy chipsound + great vocals! :-) Many greets, Peter

Great idea! :-) Even if the Raystar OLEDs did hurt the wallet, you'll love those displays! :-) Many greets! Peter

Hi there, you should be able to do that when clicking on your profile icon on the upper right of your screen, and enter your status update on the next screen in the "whats on your mind" field. Many greets and enjoy! Peter

Sounds good. Everything is normal, this is the character of the 6581s :-)! Recommendation: add a noisegate per sid pair and glue on some IC heatsinks and increase air circulation! :-) Jjonas' work is great! (The 6581s can take some heat, but putting 4 or more of them in a MB6582 case causes more heat than in the original C64 case. The heatsinks help, if there is any kind of (passive or active) air circulation.) Many greets! Peter

Hmm... could you post a demo recording (without audio compressor/noisegate! :-)) of the MBSID default/init patch? Just one short note on/note off event and the tone, that emits afterwards... Normally, it should be silent enough to properly use a noisegate... :-) Does it occur in all engines, did you test multiple 6581s? Many greets, Peter

Latigid on: maybe I've found, what causes the problem on your side... the third icon from the left after the first horizontal spacer truly is the "code" icon. Best guess: your images are cached and outdated and should be reloaded. Can you clear your browser (image) cache or Shift-Reload the page? The problem should definitely go away if you install a new browser to test (e.g. Chrome, if you don't run that already, or Firefox if you run Chrome :-)). Many greets, Peter

Yes, there are still some forum software bugs present, methinks. Antix: Regarding the old image gallery problem. This is really bad... First, I've examined the HTML source below the lightbox settings. The thing is, that the image is present, it just is not handled correctly by the IPB forum browser. By tweaking a few HTML attributes, it is possible to display the image (not when clicking on the lightbox link). It just thinks it has a picture size of 0x0px, and therefore hides the image. So, the IPB gallery image database is kind of broken (probably happened during one of the last updates). To remedy this problem, I've googled a bit for other people with the same symptoms. The common recommendation here is to perform a "gallery rebuild" in the admin panel (ACP > Community > Gallery > Settings > blue link "rebuild existing images") , which I've tried twice. The problem is, that it crashes reproducably after about 25% of the images are processed. What is even worse, is that it takes up more and more diskspace after each run, so this is not recommended. So, my recommendation is, to report this issue to the makers of IPB (TK. paid for the license from the community donations) and hope for a bugfix for the image rebuilding. Latigid on: that is really a strange bug, too! Which browser are you using? It is not shifted here, running a new Firefox build... Can you try a different browser and report back? Many greets, Peter

Hmmm... yes, works here, too. Plain windows 10 installation with no additional drivers... also an old STM32F1... can you try to remove the GM5 drivers, if you have them installed? Many greets, Peter

... looking into it! :-) Many greets, Peter

Latigid on: up to eight from the Core J15A, four more from J28 (as you said) and up to 7x8 more from DOUT-equivalent modules connected to J28 - see these links: http://ucapps.de/mbhp/mbhp_lcd_ssd1306_single_mios32.pdf http://ucapps.de/mbhp/mbhp_lcd_ssd1306_multiple_mios32.pdf and the SSD1306 section of this page: http://ucapps.de/midibox_ng_manual_lcd.html Henkmeid: 7 pins are all that is needed, if you are positive, that these match with SDA, SCLK, RESET, CS, DC, GND and VCC, you are good to go. Maybe you need to resolder some onboard jumpers to select 4-wire SPI mode. (on my displays: D1 = SDA, D0 = SCLK, so it looks good. Buy only a few first to test, though!) Enjoy and many greets! Peter

I'd suspect the switchers are causing the spike, as they are unloaded. You could test-install a load and redo the test (e.g. as proposed, the 1k resistor between 5V and GND) - the voltage spike should not be so big then... pull the resistor, if it gets hot (but unlikely)... Many greets, Peter

Latigid on is right, some connections were optimized for the matrix routing, but don't worry too much :-) If your switches are "temporary on", you can use them, if they fit. There are 2 input pins and 2 output pins for every switch on the PCB. The input pins and output pins are usually connected internally within the standard switches, but you could do that with a piece of wire on the PCB, too (just as you said - jumper on the PCB, but don't connect the two open pins. Connect them to your your two switch pins, respectively)! Many greets, Peter

If you are into pattern-based composition, you'll love the MBSEQ! :-) Many greets and enjoy! Peter

That is correct, the displays need 4-wire serial protocol (SPI) support, on ebay some are sold as I2C only, or the SPI pins are not available for (easy) soldering. Driving 24 displays is somewhat demanding: a) you need a separate 3V3 regulator b) if the signal wires are too long, you might need to buffer the serial communications to avoid display garbage Recommendation: use the new STM32F4 core! Here is a project, that also uses 24 of these displays in a MBNG environment. So to conclude, it works, and you can do it! The 128x64px displays are also really tiny, you could use these, too and have higher resolution - they look awesome! :-) Many greets and enjoy building! :-) Peter

As your SPST switch probably stays in place after the switch operation (like off -> on), this is not practical, as MBSEQ functions are triggered by a short "off->on->off" pulse, when you press (and release) the tactical switch. Wilba's frontpanel is able to accept multiple switch footprint sizes, though, no explicit need to use the original recommended ones. Many greets, Peter

Ok, just saw your other thread regarding the resistance between +5V and GND where I asked, if a switcher is in place :-). This explains it! :-) You'll need to trace it through... the +15VDC pin from your power connector must reach the input of the Recom 5V switcher (multimeter: 0 resistance). The output pin of the Recom 5V switcher must connect to the +5V MB6582 baseboard pin (upper-left center of the above picture, also with 0 resistance). If the +5V MB6582 baseboard pin is directly connected to the +15V input pin, there is something wrong :-). Other than that: some switchers require loads to operate correctly. High voltage readings, if no loads at all are connected, are common, but +15V is very high and should not happen, so I think some wrong wiring is involved. If you want to add a temporary test load, you could use a temporarily inserted resistor of 1kOhm or greater across the +5V and GND board pins. This load should only consume 0,025 watts @5V and stay cold, but should drop the voltage of the switcher, when measuring again, at least a bit. Pull out the resistor instantly, if it gets hot after powering up:). Many greets, Peter

Hola, I would not worry, there is no short, the measured resistance is caused by some components already in place. (Is that a switching regulator right to your probes? Which "MB6582 Baseboard Power Option" type are you building?) Recommendation: measure again without ICs installed, when the PSU is attached and you turned on the unit. Then, there should be around +5V across the PIC power pins... Many greets, Peter

Hola! Which LCD model is it? Do you have a link to the datasheet? Imho, it might be an accidentally swapped data wire... you could try rebuilding a new cable, following these instructions: (they are not better than Jin Henrys, just a different approach, maybe it will work like this?) Many greets! Peter

Awesome, that you represent MIDIbox over there! Have lots of fun (and make a video)! :-) Many greets, Peter

it is still existing, but 8 members only and the last posts are from 2011: https://www.facebook.com/groups/6735096007/ so, I would not call it "active"! :-). Many greets, Peter