TK.

Du programmierst den LCD Treiber in C und nicht in Assembler? Die CPU scheint wohl ansonsten nicht ausgelastet zu sein... ;-) Gruss, Thorsten.

Hi Dave, yes, the MBFM application doesn't output the 1kHz tone, instead you should be able to trigger an epiano sound, which is located in internal EEPROM (so, it should even work w/o BankStick). Hope that you've used "Smart" mode during the application upload - just to ensure that it was really uploaded. could be related to the voltage regulator - is the 7812 getting hot? yes, it is Best Regards, Thorsten.

A new release candidate is available: o multi engine: improved approach for routing LFOs to filters (only LFO of last assigned voice will be used for more predictable results) o reduced intensity of LFO pitch modulation for multi and bassline engine (Lead engine: intensity can be reduced by decreasing the depth of the modulation path) o 8 external switches can now be connected to - DOUT shift register (-> DEFAULT_EXT_SWITCH_DOUT != 0) - port J5 of the core mode (-> DEFAULT_J5_FUNCTION == 3) The switch states are stored in the patch structure, and can be accessed via Control Surface (EXT menu) or via Parameter Change (NRPNs/Knobs/Wavetables) o support for dumping ensembles via SysEx o the SysEx structures are now completely documented under doc/mbsidv2_sysex_implementation.txt [/code] Note that the file structure has been changed to improve the oversight. Include files are now located under src/, documents under doc/ Best Regards, Thorsten.

Your PSU is not strong enough, this causes the issue. During Power-On, the PIC requires most of the power. For these 1..2 seconds the voltage probably sinks below 4.6V, and unexpected things will happen - or nothing will happen... I would propose to try it without LCD backlight first, and maybe only with a single PIC (remove the slave PICs from the sockets) just to check if my assumption is correct. Best Regards, Thorsten.

No, this isn't normal - the LCD output should be stable. Best Regards, Thorsten.

Hi Paul, welcome to the forum! :) Note that a 2x20 LCD is recommented for MBSID, with 2x16 some messages are not completely visible. I don't have so much experiences with supplying a MIDIbox from batteries, and also cannot give you a proven schematic for a low-drop regulator. Yes, it should already work with 7809 and 7805, the only danger I see is, that on a battery brown out something unexpected could happen - e.g., if you try to write a patch into the BankStick while there is no power anymore to finish the procedure, invalid data could be written, or the whole BankStick content could be corrupted (thats only an assumption - you wanted to be informed about possible issues...) Best Regards, Thorsten.

Hi Dave, the voltages at IC5 outputs seem to be ok, a positive offset between 2V..4V has to be expected. I would propose to check this first with removed IC3. You could connect the audio input of your mixer/soundcard to the IC5 outputs w/o much risks - you should hear a silent, but noticable sinewave (with the testtone app). If you don't hear it, we have to search for another debugging approach. Best Regards, Thorsten.

Yes, it should be a 1:1 connection. Best Regards, Thorsten.

I guess that DragonMaster means an integrated crystal oscillator, but it won't really help, since clocking the PIC with 40 MHz from external is out-of-spec, and using the PLL to quadruple the frequency will lead to asynchronous effects - so, there is no benefit at all. The frequency stability can be tested the following way: assign all SID cores to the same MIDI channel, select the same BPM rate, select the same drum or bassline patch, and start the sequence with a MIDI note. Last time I checked this, all cores were still running in-synch after ca. 30 minutes, thereafter I stopped the experiment. Best Regards, Thorsten.

The bottom connector is normaly connected to shield, but it will also work without. It would even work if no ground is connected to the middle pin, as MIDI is a current loop. When you are writing "got my leds to light up" - does this also mean, that the Tx LED is flashing when a MIDI event is tranmitted over the IIC port? Receiving MIDI data: did you connect the RI_N output of the IIC_MIDI module with J6:SI of the core module? Best Regards, Thorsten.

It could work somehow, since the control surface already routes most of the parameters through the CC layer (-> CS_M_HLP_TrkCC_Set function), so that the extension only needs to be added to a central function. The mute functions are not routed through the CC layer, but this could be changed... But I already see some issues: how do you want to re-initialize the changes when a song is restarted? Do you want to record all initial CC parameters? Thats a lot of events which will cause a huge delay. To which MIDI port should the events be routed? I think that you don't want to record them over the same port like the events which are directed to synths? Does Sonar allow you to filter incoming events, before they will be recorded? Best Regards, Thorsten.

You are doing everything correctly, my assumption is, that there is anything wrong with the shift registers. You are able to test each seperate pin with the dout_enc_v1_3 test, which is available at the MIOS download page. When selected, the pin should output 5V, when not selected, 0V Does it work? To answer your other questions quickly: I cannot say if the display is broken, but if it flashes during power on, it seems to be ok. This effect tells you at least, that the 74HC595 are powered correctly. Yes, dot LED doesn't need to be connected. Yes, the digits should display the BPM once the MIDIbox SEQ startup screen is visible after power-on. No, you don't need to go into the BPM page to activate the display, it's a hardware option. Display connections: always a confusing topic, it's easier to try out the LEDs directly. Desolder the display, or remove the 74HC595 from the sockets and to the test by connecting cables to the empty socket pins. Connect ground over a 220 Ohm resistor to one of the cathodes, and +5V to a common anode line. The selected segment should lit. Don't forget the resistor - without you will fry the LEDs! Note that it is also possible to do this test by connecting a resistor to the anode instead of cathode - it doesn't matter when the resistor is located, it's only important that you don't drive the LED pins directly with 5V! Best Regards, Thorsten.

Yes, I would prefer the audio solution for delaying the metronome signal. Best Regards, Thorsten.

Thats a strange behaviour, which EEPROM are you using exactly? E.g., I would expect that for the case you are using a 24LC256 (32k) type, the message "<only 32k BS>" would appear on LCD when a patch >64 is selected. For BankStick #1 (Bank A) it's important that all chip select lines (pin 1, 2 and 3) are connected to ground. The write protection (pin #7) has to be connected to ground as well. Best Regards, Thorsten.

Could it be, that the socket for the C64 PSU plug is broken? I would expect such an effect if one of the 9v AC pins has no connection. Best Regards, Thorsten.

You are clever! ;-) Best Regards, Thorsten.

No, it's plug&play Best Regards, Thorsten.

The reason is simple: since I don't own Cubase or Nuendo (I'm a Logic user...) I used a demo version to test the protocol handling. Last check was in June 2003! I don't really know if it is still working with current versions, especially because Steinberg added some SW workarounds into Cubase in order to overcome imperfections of the original hardware. Best Regards, Thorsten.

You forgot to copy&paste the three lines before this code block: ;; duplicate (just because the remaining pins are not used for any other purpose) swapf WREG, W iorwf MIOS_PARAMETER1, F [/code] it's a hidden feature (but please don't tell it anybody else before they blame me that it cannot be found in the schematics ;-)) - each common select signal is available on two pins. In other words: the D0 signal is also available at D4 Best Regards, Thorsten.

No, it should be sufficient to swap the pins at one side for this check (so in your case: MBHP_IIC module). There is a iic_midi_sw_loopback test available at the MIOS Download page, did you already try this? It's very useful for debugging the hardware. Best Regards, Thorsten.

Hi Doug, I will get a MB-808 kit as well, and I'm planning to create a branch of the MBSEQ V3 firmware which will be optimized for drum sequencing. On the other hand especially this version will *not* support a LCD, since most people want to left it out, and accordingly the whole user interface shouldn't rely on it. In addition it would create an imense extra effort at my side, if I would have to maintain a version which has a totally different workflow - therefore I prefer a clear seperation between MBSEQ V3 and the MB-808 variant. So, what I want to say: it's great that you want to work on an alternative firmware, and you could benefit from my variant, but I'm sure that it won't be available before I got the kit and built the hardware by myself. Best Regards, Thorsten.

Hi Matteo, unfortunately there is no forwards-quantisation available. I tried to implement this, but the result was extremely unuseful, so that I gave up. :-( So, there is only a backward-quantisation, which means: the last note which has been played at the time the step was active, will be stored. The current pattern can be easily deleted by pressing F1+Clr Item. Thereafter press the button below the Rec item to go back to the record page. There is also a Move function in the utility page which allows you to shift single steps after recording. The user manual describes how it works. And the Scroll function allows you to shift the whole, or parts of the recorded track. Best Regards, Thorsten.

Hi, no you don't need to connect Vss or Vdd to any LED digit pins - in distance: if you would do this, you would fry the LEDs (-> dead) and maybe also the shift registers (-> dead) You wrote that you've removed the resistors. Thats good - but did you also solder the bridges? In other words: did you replace the resistors by bridges, so that the common pins are directly connected to the 74HC595 outputs? Another possibilty would be a failing .hex build, so that the old (unchanged) firmware has been uploaded - doublecheck that MPASMWIN doesn't print an error message. If you are unsure, you could also check this by opening main.inc, and replacing following message: #define MBSEQ_VERSION_STR "MBSEQ V3.2a " by: #define MBSEQ_VERSION_STR "Echopraxia " With this change, you should see your name after the MIOS bootscreen. Best Regards, Thorsten.

Hi, yes, BankStick support was only temporary disabled, and yes it is possible to read out the BankStick, see http://www.ucapps.de/mios/mios_backup.txt (no special support from MIOS Studio). However, I think that the issue is related to the MIOS Device ID of your core. Because the "CS not enabled!" message (thanks for listing this important detail!) indicates, that the firmware is running on a slave (-> device ID != 0x00) Is this correct? Only the master reacts on program changes, and thereafter sends the patch to the slaves which are assigned to the appr. MIDI channel. In other words: connecting a BankStick to a slave doesn't make much sense... Best Regards, Thorsten.

Since MIOS has been pre-installed, the bootloader will only send a single upload request message during power-on before MIOS is started. This one is hard to notify. No, the Tx signal doesn't go through the optocoupler - with the Core MIDI Out -> PC MIDI In connection it will be connected to the optocoupler of your PC MIDI Interface. If the Tx line outputs 5V, it's very likely that the PIC is running, otherwise Tx would be around 0V So, it looks like an issue with the MIDI In Port, with the MIDI connections to your PC, or with the driver of your MIDI interface. Yes, it is sufficient to connect only the LCD for this check, buttons and encoders can be connected later. Note that for MBSID V2 a SHIFT/UP/DOWN button should be added as well (see MBSID V2 Manual, Hardware chapter). Best Regards, Thorsten.