TK. Posted July 1, 2014 Report Share Posted July 1, 2014 It's great to see an alternative core module for the STM32F4 which is optimized for this usecase! :) Best Regards, Thorsten. Quote Link to comment Share on other sites More sharing options...
HUROLURA Posted July 1, 2014 Report Share Posted July 1, 2014 Great news. By the way, I happen to already have some STM32F4Discovery boards ... Quote Link to comment Share on other sites More sharing options...
novski Posted July 1, 2014 Report Share Posted July 1, 2014 (edited) Hi Benoit<br /> I am doing the same at the moment but with a different approach. As Thorsten implemented a way to connect many OLEDs in a Serial chain i designed a way to make that connection easy available over a 10pin DIL Header. So maybe it wold be possible for you to implement that at your PCB so i don't have to do same again...?<br /> I just uploaded my Eagle files to github: <a href='https://github.com/novski/Midibox.git'>https://github.com/novski/Midibox.git</a><br /> There is a PDF with two marks that show the connection. I will attach it here.<br /> MBHP_CORE_STM32F4v1.1.2.pdf<br /> <br /> I had no time yet to Layout it. But if you can use my outlines (Dimensions) it wold fit with all my other Layouts to be stacked easy...<br /> It took me quite a while to find a matching form for all Midibox Parts... :-)<br /> its stored here: <a href='https://github.com/novski/Midibox/tree/master/VLR-Eagle%20Library'>https://github.com/novski/Midibox/tree/master/VLR-Eagle%20Library</a><br /> <br /> What do you think?<br /> <br /> br, novski Edit: maybe i have to mention what it is for... 😀 I have made a wiki page. Its still in progress but can help understand what it is for... http://www.midibox.org/dokuwiki/doku.php?id=fadercore_-_fadermodule_for_audio_daw Edited July 2, 2014 by novski Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted July 3, 2014 Author Report Share Posted July 3, 2014 Hello Novski, no problem, I can add this header to my board, it's an interesting feature. Just be sure that I do not miss something : the serial control lines RCLK AND SCLK are in the red square on the left, but they do not go in the new header (only D8, D9 and D10 are going there). Is it what you want ? Benoit Quote Link to comment Share on other sites More sharing options...
novski Posted July 3, 2014 Report Share Posted July 3, 2014 Hi Benoit Thats not all... I make a list of "STM32F4Disco Pins" and "Disp-Ext. 10pin Header" pins: Disp-Ext. Pin 1 : GND Disp-Ext. Pin 2 : GND Disp-Ext. Pin 3 : LCD_Voltage_Option its called J15 on the Core Schematic Disp-Ext. Pin 4 : LCD_Voltage_Option its called J15 on the Core Schematic Disp-Ext. Pin 5 : PA8 Core pin Disp-Ext. Pin 6 : PC13 Core pin Disp-Ext. Pin 7 : PC9 Core pin Disp-Ext. Pin 8 : PC14 Core pin Disp-Ext. Pin 9 : PC11 Core pin Disp-Ext. Pin10:PC15 Core pin Is that more clear? Best regards novski Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted July 3, 2014 Author Report Share Posted July 3, 2014 Is that more clear? Best regards novski Yep, perfect !!! It will be on the next schematic release Benoit Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted July 12, 2014 Author Report Share Posted July 12, 2014 I have made some last minute changes (and included the connector requested by Novski :smile: ), and the PCB now receives the KissBox board directly, in the same way we do for the STM32F4 board. The final PCB review was done yesterday and the files are now in the factory for production. For those being interested, the PCB will be available starting from the 29th July. I will publish the updated schematics on my website very soon (probably this weekend). As I explained in the previous post, I have decided to sell the PCB either "naked" or in the form of a kit, including the KissBox OEM module plus all other components (the KissBox board will be optional of course) (And I have already made the first application for this board :rolleyes: ... I will post a message in the "design" section of the forum very soon about it, I need to make first some photos of the proto application) Quote Link to comment Share on other sites More sharing options...
novski Posted July 12, 2014 Report Share Posted July 12, 2014 Wonderful! 😀 Quote Link to comment Share on other sites More sharing options...
Hugo106 Posted July 27, 2014 Report Share Posted July 27, 2014 Hi Benoit. STM32 Cortex M4 good choice and more safe regarding the shortness SRAM on your RTPMIDI OEM board :rolleyes: Like the hardware, the code will be open :question: Quote Link to comment Share on other sites More sharing options...
Shuriken Posted July 27, 2014 Report Share Posted July 27, 2014 Hi Benoit. STM32 Cortex M4 good choice and more safe regarding the shortness SRAM on your RTPMIDI OEM board :rolleyes: Like the hardware, the code will be open :question: I don't think Benoit is using an STM32 M4 for the RTPMIDI board. However the MIOS32 part is running on that MCU. And yes the code for that is open :wink: Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted July 30, 2014 Author Report Share Posted July 30, 2014 (edited) And here we are finally :sweat: The PCB for the RTPMIDI_CORE_STM32F4 arrived yesterday evening, and I assembled it as fast as possible to test it. Here it is, doing the first test run, and already equipped with the RTP-MIDI OEM module Of course, I discovered a small design glitch: I added a extra connector (J16), moved J1 in turn... and completely forgot about the mini USB connector on the STM32F4. And of course, the 3M connectors are a little bit smaller than the ones I used on my prototype, so it's a little bit difficult to insert the USB connector when the flatcable to the RTP-MIDI board is installed (I have to remove a little bit of plastic on the connector, and slightly lifted the PCB from the connector if I want to use the ST-LINK for debugging). I will correct that in the next version of the PCB. If you do not use the ST-LINK (for example with MIOS studio), or if you do not want to install the RTP-MIDI option, there is no "collision" problem. By the way, it's also possible to use J16 to connect to the RTP-MIDI OEM (and then get more room), but the cabling is not direct between the board and J16 (not very funny to twist connectors like that) Apart this little problem, no smoke and both CPU are running :wink: So the PCB is now available for those interested Edited July 30, 2014 by BEBDigitalAudio Quote Link to comment Share on other sites More sharing options...
Shuriken Posted July 30, 2014 Report Share Posted July 30, 2014 Looking good Benoit. And i see an XMOS xCore there. Quote Link to comment Share on other sites More sharing options...
Hugo106 Posted July 30, 2014 Report Share Posted July 30, 2014 Yeah, really impressive board, X-core model XS1-L16 :santa: STM32 M4 processor code is open, but X-Core :question: IDE is free :rolleyes: What's the price for both cards :question: Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted July 30, 2014 Author Report Share Posted July 30, 2014 Looking good Benoit. And i see an XMOS xCore there. Thanks Shuriken, yes, we use an XMOS chip on the V3 CPU because of its incredible flexibility and speed (more than 800MIPS, 16 realtime threads with a 10ns accurate scheduler...), a lot of memory even for our most complex applications (the complete RTP-MIDI engine uses only 60% of the RAM if I remember well, including Bonjour services, configuration tool support, etc... :smile: ) Historically, the XMOS chip was the reason I took contact with Thorsten, because we were discussing of making a MBHP board based on this chip (just imagine that XMOS has a four-core model - the G4, providing 1600MIPS and 256K of SRAM... The only problem with this chip is to find an application able to use the huge processing power :tongue: ) But finally, we agreed to use our OEM board as a "ready to use" co-processor, handling all the communication job and freeing the Cortex for other tasks (a little bit like with the Ploytec chip) Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted August 1, 2014 Author Report Share Posted August 1, 2014 (edited) As I got some questions from people who have received (or about to receive) their OEM board, I would like to give here some details and explanations about this board : - the most important : in the small package containing the CPU module, you will find a sticker with a serial number and the MAC address. Do not stick it on the RTP-MIDI OEM or the CORE board, this sticker has a metallic support for better resistance, and it may create short circuits on a PCB. This sticker was designed to be put outside a device, not inside - you can see on the package and on the PCB : "V3.1 - 03/2012". This is not the production date, it's just the date at which the PCB files were created - the date code (for the production date) is visible on the right side of the PCB, under the processor. The boards delivered now indicate 0814, which means that PCB was manufactured in week 8/2014 (the components are generally soldered in the next 3 or 4 weeks) - the package indicates "firmware : --". Do not worry, I install OEM firmware V1.0 before packing back the CPU module. So the CPU are delivered with firmware already installed (this is the same firmware as the one Thorsten used to validate the code on his side for the MIOS) - I am currently finishing the OEM firmware V2.0, which some new functions, which will require Editor V11 (Editor V11 will tell you that firmware is obsolete if you use it with OEM firmware V1.0, so just use Editor V10 for now). I will post a message here as soon as OEM firmware V2.0 becomes available (normally, in the middle of August) Benoit Edited August 1, 2014 by BEBDigitalAudio Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted August 9, 2014 Author Report Share Posted August 9, 2014 (edited) Just back from holidays, and playing already with the RTPMIDI core board... and doing some magic tricks with it : no USB cable on the STM32F4DISCOVERY, no USB cable on the USB socket of the motherboard... but everything is running :tongue: (Nothing magic : I just installed a Power Over Ethernet module on the OEM board, it provides more than enough current for the complete motherboard... and it makes a nice picture) Edited August 9, 2014 by BEBDigitalAudio Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted August 13, 2014 Author Report Share Posted August 13, 2014 During the tests with the RTPMIDI MBHP, I found a very strange problem : the communication between the OEM module and the STM32F4 was sometimes becoming garbled and some messages were lost (you can easily see that with Thorsten's test application and MIOS Studio console) In Thorsten source code, the speed is set to 10Mbps (after Thorsten and me made a lot of tests on our development prototypes). My first proto and the two first RTPMIDI MBHP are working flawlessly at this speed, but the third module I made is failing. I finally found what was wrong: on one board, the 74HC125 used to select the MISO line between SD Card and RTP-MIDI module reacts quite strangely when it is powered by 3V. I discovered that the signal becomes out of phase (because of the lower speed of the chip) and this is enough to garble some bits in the message. Honnestly, I do not have a real explanation for that, since the 74HC125 is supposed to work between 2.7V and 5V, and it's supposed to accept more than 25MHz signals under 3V. But when I replace the 74HC125 on this board or when I power it under 3.3V, the problem disappears. I found a workaround for now if you encounter this problem: just reduce the SPI CLK on SPI0. When it is lowered to 5MHz or 2.6MHz, no more problems are occuring (and even 2.6Mbps is 83 times the speed of MIDI, so you still plenty of bandwidth even with 16 sessions in parallel) To change the SPI speed, you can do one of the two following things: - in MIOS32_SPI_MIDI.H file, locate the definition of MIO32_SPI_MIDI_SPI_PRESCALER. Change it to another prescaler value (MIOS_SPI_PRESCALER_32 gives 2.6Mbps, MIOS_SPI_PRESCALER_16 gives 5.2Mbps) - define MIOS_SPI_MIDI_SPI_PRESCALER in your own project (the code from Thorsten will detect your own definition and it will override the default one). This solution is the best one from my point of view, since it does not affect the source code from Thorsten (but I have asked to Thorsten if he could use a lower value by default) Quote Link to comment Share on other sites More sharing options...
ilmenator Posted August 13, 2014 Report Share Posted August 13, 2014 I wonder why you would rather "cripple" the software if a hardware fix has been found already? Apparently, there are HC125s around that conform to the specs, so why not use those? Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted August 13, 2014 Author Report Share Posted August 13, 2014 I wonder why you would rather "cripple" the software if a hardware fix has been found already? Apparently, there are HC125s around that conform to the specs, so why not use those? I agree with you, but since I can not explain why some HC125 are acting in this way, I prefer to provide an easy workaround for those who are already working (or will work soon) with the RTP-MIDI module. Not sure that everybody has a full lot of HC125 from different suppliers to swap and find the "good" ones (moreover I already found two chips from two suppliers giving this behaviour... so it's not related to a given manufacturer) I have to say that Thorsten also saw a degradation of performances on his setup when the speed was too high, and we think that it comes from the cabling, generating parasitic oscillations. I think that most members here do not have a fast oscilloscope (maybe I am wrong about that...) to check the signal integrity, so I prefer to provide an easy fix which makes sure that a working solution can be found even under degraded conditions due to cabling Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted August 13, 2014 Author Report Share Posted August 13, 2014 By the way, as promised, the firmware V2.0 is now available for those who want to upgrade their OEM module. Note that this version requires the KissBox Editor to be upgraded to last version too (V11) to get access to new functionnalities The file is available on this page http://www.kissbox.nl/downloads.html Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted September 1, 2014 Author Report Share Posted September 1, 2014 Important messages to all MB members who are using the RTP-MIDI OEM : I found a stupid bug in the firmware 2.0, which impacts the SYSEX messages (a message is being duplicated erroneously) I am working on the bug right now, I will try to deliver a corrected firmware as soon as possible (when I will have finished the tests of course). Sorry for this problem. Benoit Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted September 11, 2014 Author Report Share Posted September 11, 2014 The new OEM firmware is now available on the download page of KissBox website (V2.1) It solves a bug which occured when the module was receiving some SYSEX messages. Contact me if you need any help to install the new firmware Benoit Quote Link to comment Share on other sites More sharing options...
borfo Posted September 17, 2014 Report Share Posted September 17, 2014 I've just skimmed through this thread and saw a lot of discussion about price, but I must have missed the post(s) that actually set out the price for the RTP-Midi board... Are these still available? how much are they? Are you selling that custom Core board? Quote Link to comment Share on other sites More sharing options...
Hugo106 Posted September 17, 2014 Report Share Posted September 17, 2014 Yeah, what's the price and availability? It's private or commercial or DIY project? it's confuse, can you precise, thx Quote Link to comment Share on other sites More sharing options...
BEBDigitalAudio Posted September 17, 2014 Author Report Share Posted September 17, 2014 I've just skimmed through this thread and saw a lot of discussion about price, but I must have missed the post(s) that actually set out the price for the RTP-Midi board... Are these still available? how much are they? Are you selling that custom Core board? Hello Borfo, yes, the RTP-MIDI boards are available. The price depends if you buy the OEM module alone or as a complete kit with the MBHP. If you want only the RTP-MIDI board, please contact my colleague Sierk directly at KissBox (via the contact page of www.kissbox.nl). We proceed like this to keep the price of the OEM board as low as possible when it's alone If you want the complete kit (the motherboard PCB + the OEM module), the price is 115 euros for the complete package (not including the postage fees). You can find the whole documentation on my own website : beb.digitalaudio.free.fr Benoit Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.