MBBLM schematics and PCB discussions

[diwib]

Of course it's no good idea to lay bricks before you know what you're gonna build, but that's how i built pretty much everything in the past (with the schematics being the very last thing or even nonexistant). In the end it's me who has to spend another 2 hours to re-route stuff but I don't mind at all. Gives me reason to watch another Episode of Star Trek   :phone:

[diwib]

Good news - I'm bored.

 

The 16x4 and Extra 1 schematics are alright, the other two will take longer.

 

I found a really minor mistake in TKs schematic though.

 

 

 

 

Little edit: I spent 2 hours thoroughly checking the schematics. They're all good.  :shocked: Kinda unexpected, especially since I was like f* this a couple of times while making the bottom row (Extra 2 and 3 boards). Also a little update for better mechanical fit.

 

Now if only I had money to order the boards, lol. :sad: Well, buttons, diodes and LEDs are ordered for like 50€. And they're enough for 3 16x16 BLMs. Nice square ones for the extra fun aligning them ^.^ Thanks to maigre, don't know why I never considered Alibaba/Aliexpress before.

PCBs ordered in batch are gonna be some 50€ aswell; even cheaper if I would panelize them but I don't wanna risk breaking them trying to cut 'em apart - don't currently have access to our laser cutter.

Plus a Core8, ribbon cable, connectors and cheapo components for the Scalar boards, that's maybe another 50€. And personally I will get a "Biegelehre" for the Diodes. Whatever that's called in english...

 

Aaaand another edit: TK I need a little help! Over here you described how to set up the Seq V4 to be used with the virtual BLM. For the V4L I can unfortunately only find info on how to use with Lemur via ethernet. But how would I set up a V4L to be used with an actual BLM (without a display)? Obviously via some MIOS-Commands, but which?

Ha! Nevermind. "set mclk_in USB2 on", "set mclk_out USB2 on", "set blm_port USB2" did the trick. Just writing this here for personal reference :smile: I suppose for a real BLM, it'll be MID2 or so respectively.

I see there are a lot of unused "menus" - shift 3-11. Would it be possible to assign various settings to these so that one could completely omit any interface besides the BLM? Kinda not a MB Seq V4L anymore, but rather a Seq V4BLM with a feature set inbetween a V4 and V4L?

 

Made a Version 2 of my board now, which is a single PCB 250x250mm, perhaps I can get those instead of using my modular version. Now they're nice and square too; the rectangular thing was bugging me. And the buttons don't have a gap between the 8th and 9th step anymore. No Eagle files for this yet, there's too much going on on the board, making Eagle crash everytime I try to load it, but it's DRC tested anyhow. I'll put the Eagle script here though, maybe someone's luckier executing it. Expect some 800 DRC errors if you do, they're all fine. And all the errors concerning .lbr files are fine too. Maybe a licensed Eagle 6 does the job - I only got a licensed version 5.

 

 

Some solder fun.

BLM16x16+x.zip

BLM16x16+x V2.zip

Edited July 15, 2013 by diwib

[diwib]

Ordered at Itead... 200€ for 5 sets (min. batch). Right now that hurts.  :sick: But that's rather cheap compared to other PCB services, where I'd pay that much for 1 BLM board excluding the scalar ones. Or up to 400€ from Targets recommended german manufacturer O.o

 

I didn't want to learn how to use KiCAD so I spent just enough time to figure how to resize the outline of findbhuddas Scalar boards to be a nice 50x50mm so that they're cheaper. It looks really ugly and I hope overlapping the outline with parts of the bottom layer is something Itead can just ignore. Otherwise another OSHPark order it is. I already hate KiCAD. Could someone proficient with KiCAD maybe do this?

 

Should these boards be alright, I'll have 3 BLM boards and 10 Scalar boards to sell (40€ a set). Otherwise hopefully I can fix them with some wire, unless they're fubar, which I doubt... Also got 1 set of spare (square)LEDs, buttons (both 7mm) and diodes for the first buyer :phone: .

 

Hate to tell you this but I just can't manage to export the V2 board into Eagle, which received lotsa updates over the V1. You can find the Gerber files for V2 in this post or if you want minor mods to the silk screen or so just throw me a message and I'll export new gerbers for you. Unless of course you have a Target 3001 license...

Should there be urgent need for an Eagle file I can try my best to split the board into parts again for you to reassemble in a licensed Eagle...

 

Do not(!) order any of these gerber files yet, just look at them xD

BLM Scalar module resized gerber.zip

BLM16x16+x V2 gerber.zip

Target and KiCAD projects.zip

Edited July 19, 2013 by diwib

[findbuddha]

Hey there, hope you saw this 

 

- blm-scalar.* contain an error. Pins I4-I7 are mirrored. This is easily fixed with a connector modification, and the blm-scalar module seems to work as intended. A corrected but currently untested file is separately available below.

 

on my wiki page before ordering  :happy: 

[diwib]

Hey there, hope you saw this 

 

 

on my wiki page before ordering  :happy: 

 

I was thinking it would be corrected in the file called BLM-Scalar-corrected.brd ? Did you call it like that or did I?

 

Ah, just saw it's still a separate file. I made sure to highlight it for myself but thanks for reminding me! Maybe I broke it anyhow by resizing. 

Edited July 19, 2013 by diwib

[findbuddha]

Yeah that's the corrected file. I have never tested that version, but it should be fine. Can always make new connectors using wire if it doesn't work out.

[diwib]

For convenience, would it be possible to simply disable analog events? My Core8 is going nuts and I'd like a nicer solution than to ground everything :p

(What do I comment out?)

Plus, why does the SIL header A0-A7 have a different voltage than the DIL headers A0-A7? Almost fried my 7805 using the DIL headers and I'm not exactly sure why. Built myself a SIL jumper now but the above question remains.

 

Edited September 4, 2013 by diwib

[diwib]

I think this is worth an update:

 

 

Edited September 8, 2013 by diwib

[TK.]

Hurray! :-)

 

Is the communication with MBSEQ V4L working at your side?

 

Best Regards, Thorsten.

[diwib]

Yes, with IN2 as blm_port everything's working fine. Thanks again.

 

I don't currently have access to my schematics, so the +x isn't connected yet and for midi action I'll wait for my new notebook to arrive, then hopefully put a video together.

[Altitude]

Have you considered just attaching the driver boards directly to the control surface PCB via female DIL header (like on Anushri?) would save some wiring..

Edited September 9, 2013 by Altitude

[diwib]

Ironically I got a broken wire or two somewhere...If I figure out how to do this in KiCAD I'll try that, otherwise moving the headers on the blm board is risky business because I could only guarantee it'll work theoretically. I'm not soldering another of these boards :p

That's one of the initial errors I had anyhow, thought I could just use regular ribbon cable which is barely any hassle. The Pin headers are offset by just 1,5mm but ugh...KiCAD...Maybe I'll make the scalar boards from scratch.

Edited September 9, 2013 by diwib

[ilmenator]

Don't blame it on Kicad... I don't fully get what the problem is, but maybe you simply went for the wrong component or module?

[diwib]

I don't wanna learn to use KiCAD, that's all, ain't blaming it. The problem is that for reasons I dont remember (probably routing) I reversed the pin headers on the blm and the spacing between both headers isn't alright either. An easy fix would be to just edit the Scalar boards in KiCAD. Or solder all resistors crossed ...

 

Here's a super professional visualization:

 

Edited September 9, 2013 by diwib

[ilmenator]

I see what you mean now - at first sight it looks like a solder-/component-side confusion, but it is more complicated than that :sad: .

[diwib]

:rofl: Managed to avoid KiCAD! Yet another order to wait for...But those wires I bought myself are really utter crap, I wouldnt wanna hand out my boards with those.

Edited September 13, 2013 by diwib

[Macotronic]

Hi All first congrats for this and just mention that could be cheaper to get parts for repairs like the rubber keypad for apc40 link 

 

http://www.mpcstuff.com/1aru1504315.html think could find parts from other hardwarwe 

 

and also I found this supplier :

 

http://www.top-up.com.tw/front/bin/home.phtml

 

they have the buttons with integrated leds and a lot of stuff (pots-sliders-etc) please check and i want to know if we can implement this parts to this and save space, joints etc.

 

Im in for the PCB and group buy

 

Regards

[Phatline]

looks perfect for my project!

 

I am looking for the correct  Gerber files, or maybe a prototype pcb, for me it is not clear which are the ones that are working for you.

As far I found out, you shrinked Findbuddas Layout (no space between 8 and 9), but whre is your wiki page?

thx for more info

I think this is worth an update:

[voronin75]

Hi all midiboxer) Need your help!

I built blm 16x16+X  on PCBs.
CS consists of two printed circuit boards: the first contains 16+Xx8, 2 scalar modules are installed under it, the second 16+Xx8+X contains 3 SCALAR modules
The device connects to midibox seq v4 normally on port IN3
Firmware PIC4620 blm_scalar_v1_1,
Power from 5V 3A power supply
Test of the first section 16x4 with SCALAR module -ok
The next section test is 16x8 with two SCALAR modules -ok
Test of the next section 16x12 with three SCALAR modules - all the LEDs blink, the buttons transmit the midi event to the midi monitor if you click
Section test 16x16 + extra - the device goes crazy, all the LEDs are lit randomly, the buttons transmit the midi event to the midi monitor

All the same if you change the scalar modules in any configurations, the same if you connect the modules arbitrarily on the board i.e. e.g. starting at line 5, etc.)
Those. Errors on the PCB seem to be eliminated, on the scalar, it seems, too. Remains Core (voltage drop in the last scenario to 4.9V)

Connection length between SCALAR modules about 100mm

jumpers on Core J5 are not installed as stated in the description of the firmware  project_without_ain.hex  (the same result with jamper)

Has anyone come across? Please tell me where to look?

Regards

 

 

 

 

 

Modules are installed under CS and connected by a ribbon cable about 10 cm

Edited February 25, 2020 by voronin75

[latigid on]

Inrush current issue? Though the LEDs seem to be driven with 1k, so not terribly high. Grounding issue between power supplies? Check all 0V are common. SRIO integrity? Try buffering the J8/9 signal and/or terminating the chain with an RC shunt.

 

[voronin75]

This is my version of SCALAR. Mount under main pcb to  dil socket.

 

Edited February 26, 2020 by voronin75

[voronin75]

22 hours ago, latigid on said:

Inrush current issue? Though the LEDs seem to be driven with 1k, so not terribly high. Grounding issue between power supplies? Check all 0V are common. SRIO integrity? Try buffering the J8/9 signal and/or terminating the chain with an RC shunt.

 

latig on thanks for your answer! The grounding of the blm power supply is connected directly to the seq ground.

Can you explain  buffering the J8/9 signal and/or terminating the chain with an RC shunt.? 74HC125 chip  on minicore is buffer, right?

Works fine only two scalar in different combination CS setions. How check SRIO integrity?

Here short video of trouble

 

https://fex.net/uk/s/slzdkay

Edited February 26, 2020 by voronin75

[voronin75]

The problem is solved by connecting the last output in the SC chain to the ground (to the main chassis ground, not the last board gnd) It works correctly and stably. I still do not understand why. Can someone explain? Or i missed something?

I found the answers here. I think this is what latig on said.

On 17.08.2010 at 0:58 AM, TK. said:

After I read the blog and want to thank you for starting this report! It makes the signal integrity issue better understandable and will help us to give newbies some arguments, why SRIO chains are usually limited to 16 shift registers (longer chains could require expert knowledge and technical equipment to get stable transfers)

 

There are two points I would like to add:

 

• we get a more difficult situation once a DIN chain is clocked by the SCLK signal in parallel to the DOUT chain: parallel chains can result into two "ringings" (one from the DOUT, one from the DIN chain - if they have different lengths resp. impedances.
  The situation will become even more difficult if DIN/DOUT registers are connected to SCLK in mixed order, in this case it can happen that the serial data won't be shifted correctly anymore due to setup/hold time violations caused by unbalanced clocks (with different delays) at SR inputs.
  This results into the effect that bits could be missed or added in the DIN and/or DOUT chain at random moments and positions (-> flickering LEDs, random button events) - this effect could vanish if you put your finger on the SCLK line (see also next point)
• by probing a signal with your scope you will add a capacitance to ground of ca. 10..100 pF (so far I remember, the value could be wrong or different for your scope). This has to be considered while comparing waveforms with theoretical calculations.
  This means in other words: currently your waveform looks nice, but once you remove the probe the "untouched signal" could be bad again. You won't see it (as you removed the probe ;))
  Seppoman recommented AC termination to solve this - it works! E.g., at the end of the chain add a 100R +100 pF to ground (so far I remember, meanwhile he even prefers higher capacitances)

 

 

Best Regards, Thorsten.

 

 

 

 

Edited March 1, 2020 by voronin75

[latigid on]

The observations suggest interference on this clock line such as antenna effects (EMI) from sub-optimal trace routing/cabling, or transmission line reflections. Hence a suggestion of an RC (resistor + capacitor) shunt termination (forum search will turn up more info). 

Grounding to the "chassis" seems an odd choice and means there is relatively high resistance between the chassis and system 0V. Otherwise the clock that is wired in parallel to all circuits would be shorted out, no?

[voronin75]

1 hour ago, latigid on said:

The observations suggest interference on this clock line such as antenna effects (EMI) from sub-optimal trace routing/cabling, or transmission line reflections. Hence a suggestion of an RC (resistor + capacitor) shunt termination (forum search will turn up more info). 

Grounding to the "chassis" seems an odd choice and means there is relatively high resistance between the chassis and system 0V. Otherwise the clock that is wired in parallel to all circuits would be shorted out, no?

Grounding on the chassis I meant the direct connection of the last SC output to the ground of the power supply. Without RC shunt termination. Everything works well  with one exception. when you turn it on, you must reconnect the SC-> GND bus) Otherwise  the device does not start correctly. I am an amateur))) I will try.