latigid on
-
Posts
2,516 -
Joined
-
Last visited
-
Days Won
147
Content Type
Profiles
Forums
Blogs
Gallery
Posts posted by latigid on
-
-
So I should be able just to use the out on the Midi i/o correct? so J11->midi I0-> Quad IIC BLM connector
Yes, it should work from J11E. Or if you connect a MIDI IO board, you could for instance take MIDI IO4 from J2.
-
Thanks for the kind words! It's very satisfying to see everything come together after countless hours of design, component sourcing and dealing with many different companies.
-
Here some amphetamine for those who haven't built their case yet! :-)
Note that I added some enhancements & fixes to the MBSEQ firmware while working on this demo (demos are always a good possibility for me to doublecheck that everything is working as expected!). The changes are already in the SVN repository and will be released soon.
Best Regards, Thorsten.
TK.: very nice! It looks like the slider implementation is very good. I guess you kept the longer sliders in? This will be a personal preference I suppose. An idea I had was to make or buy some silicone covers to use as slider caps. This might make them a bit more ergonomic and also act as diffusers for the LEDs which can be quite bright.
I take it that there's no issues with the case? Stability (i.e. no PCB flex) is good with the extra standoffs? Are all M2 grub screws installed? Do the bottom feet work?
Next up will be to connect up extra expression pedals, CV inputs etc. :) (another DIN connector?)
So just to confirm: quad IIC blm port connects to J5B on the STM32F4 core?
The BLM/quad IIC schematic always gives me a headache :). My approach was to harvest old PC cables (with single or dual female pin sockets, could also crimp new ones) that connected all of the front switches, LEDs etc. to the motherboard, and join them with an IDC10. I'm still using the F1 Core, but luckily I have a breakout board which also separates MIDI IO 3.
If you have an older Core8 you can always swap out this for the miniCore during testing, just use temporary clips on the MIDI outs. You need to supply power to the BLM, and please don't send data without the main board powered up.
-
Hmm,
I still think there's some merit in this, for example with a single PCB and 4 displays (e.g. MBCV) you could get away with 4 standoffs and simpler wiring/less POR circuitry etc.
Did you consider FPC/FFC sockets? A quick search revealed that all of these are SMT but the metal pins are more amenable to solder than the flat cable. 0.5mm pitch correct? Tricky but doable.
-
The correct side is the rear, otherwise you will also need to mirror the cable or its connector. You may find the easiest way to remove it is to cut or slide off the plastic part of the header, then remove the pins one by one. (Technically you could also melt the solder and push the pins through, but the result will probably be quite messy.)
-
Keep in mind there are two main types of LEDs: older yellow, green and yellowy-green which each have a forward voltage of about 2V and newer diodes (especially green at 525nm, blue, white etc.) with Vf= to 3-4V. Check the data sheet and compare both the forward voltages and brightnesses in mcd. Brightness is somewhat non-linear with respect to the current passing through the junction but it can be tuned somewhat.
-
Okay, please check the build guide thread for drawings and graphics files of the side.
-
Here are a few images of the side in case you want to make wooden end cheeks. Dimensions are in inches, please check them against your case to be sure.
EDIT: suppose you could make some metal flange rack ears too! Oh, these screw threads are 6-32 and all screws are stainless steel so the aluminium parts should last for a very long time.
-
I don't have my unit here to test, it's at a friends awaiting a few modifications (my case had fewer standoffs so needs mechanical strengthening). TK. should be able to give you an idea of current draw with all LEDs lit, probably he has a nice bench supply running :).
Perhaps a small caution to check the board for errors before starting. They have been electrically tested but there's a remote possibility of unmasked ground plane which could give a short if covered with a lot of solder. I have a few spares here which I will go over carefully when the light is better, will let you know if I find anything.
-
Glad you're enjoying the cases! I was equally impressed with the quality and maybe it's a road to go down for later MIDIbox enclosures.
They had to do a few minor modifications during production, so I'll just get an update and provide the side dimensions for chunky wooden add-ons :).
-
For the older 32 bit Cores, yes you can chain an ETH and SD off the same cable because they are pinned to accept different RC lines if I recall properly. For the F4 Core, you have SD on board, so why the need for another?
-
Cases have shipped, please let me know if there are any problems (e.g. defects etc.) as soon as possible.
-
Okay, well I'm on holiday at the moment, so I won't be able to remote debug very easily. Even with a short on this resistor it would only kill that particular section and not the whole 5V line.
5V short could be anywhere on the board, not necessarily at that IC. Check the vias and mounting holes too.
Bonne chance!
-
Interesting approach, what is the intermediate board between the Core and IDC6 connectors?
-
Thanks for the tip, originally I had 5 extra RNs in.
For the transistor/resistor pairs, what short do you mean? Short to ground? You might have to upload photos. Note that the bottom right transistor pin should be connected to ground. All 74HC165 inputs have a pullup to 5V. Please check carefully around this pin, is the solder mask scratched? If you try to reflow the solder it can help.
-
Good news is that these WS2812B 5050 LEDs will probably fit (just!) underneath the clear shaft encoders I have, everything on the top PCB layer. No cutouts will also make the PCB routing easier, this was a challenge with the BLM.
One caveat that I read about these is that the sample clock must be very accurate (to within 150 ns). Seems like the F4 should handle this though?
Now to look for a transparent tact switch :)
-
A new BLM has Been Bugged!
have got a short beetwen 0v and 5V...
Can happen. Start with the R1 bridge, you may have a small short if you scratched some soldermask off. Check every cap/ pins for 5V and pin 16 of the 10k resistor networks.
-
Interesting solution! But we have to consider how the illumination will work with the hardware too. There's one option using a clear shaft encoder but it only has a 4mm hole. Maybe through hole addressable LEDs bent over a PCB hole could do.
-
Thanks for the tests! Can I ask the specs/datasheet of the LEDs you are using i.e. luminous intensity in mcd? And when you say "directly to DOUT pins" were you using current limiting resistors? I think these will be needed to smooth out differences in brightness one often sees with RGB LEDs, not to mention the lower forward voltage for the Red die.
Without matrix connections the PCB will also be much simpler to route :).
-
I see we are coming from different points of view: menu-based vs. knob-per-function. I know that I prefer the latter approach, for instance the sammichFM is a great compact synth but editing parameters on a 2x20 LCD and one encoder is quite unergonomic for me. I can see how having only depth controls makes panel labelling easier, but I prefer full access using the encoders (albeit requiring ninja skills to learn where each parameter is positioned). Selecting each channel one at a time and the configuration of its LFOs, ENVs, SEQ (also amounts from MIDI data such as velocity, AT, CC) etc. is also my preferred workflow.
For example, I would see that Channel 5 is active from its illuminated button, so I would know the encoder banks adjust the modulators' depths for this Channel. To change parameters for ENV5B (using my nomenclature) I press the left hand buttons "B+2" which shifts the encoder bank to ENV control. LFO5A = "B+3." These could be indicated with multicolour LEDs e.g. Channel Selected = green, "Bank" Selected (in my concept one of CONFIG, ENV, LFO etc.) = blue, and if the Channel and Bank numbers happen to match you will get cyan.
So you do have some multipurpose buttons, but thinking on the idea for a while this is the best "full control" solution IMO.
I think that at the end we might be able to use the same frontpanel, but labels need to be customized based on personal preferences.
I agree, there could be one PCB/PCB set with different faceplates depending on your preference :). Buttons/encoders could also be left out for instance if you didn't want a dedicated tuning control.
-
I fear that I've to create an alternative frontpanel design for my own usecase... :-/
If we can agree on the hardware there's no reason why two different front panels couldn't be created. My approach places editable parameters for each modulator on its own encoder bank, on a per channel basis. It also needs a paging key for a second set of controls on the 4x4 encoders. Note your proposed encoder assignments do not include rate controls for the LFOs, A/D/S/R for the envelopes etc. -- how would you adjust these parameters? -
Don't know what happened, so I'll put the pictures back in here:
-
1
-
-
Please note: LFOA, LFOB, ENVA, ENVB, MODA..D are actually called LFO1, LFO2, ENV1, ENV2, MOD1..4, please change the graphics accordingly to avoid confusion.
I think it will be simpler to identify modulators this way. I.e. Channel 1 LFOA = LFO1A, Channel 5 MOD C = MOD5C. Then the MOD is easy to assign across Channels.
From my point of view, the encoder assignments are too much channel-related.
Can you explain what you mean by this? I'm thinking about the following structure hierarchy, sorry if the pictures weren't clear:
Choose a channel. Here on one page you can adjust all 8 modulators, activate the sequencer and apply velocity/aftertouch.
Click "Config" and you have a second page of options.
Bank+channel number will select each ENV, LFO, MOD and SEQ, some have two pages switched by Config such as the 16 steps of ENV2/B and SEQ. To me this is the most intuitive way to access all parameters of a modulator.I would prefer the configuration 2(VCO, VCA, VCF, MOD; Gate, Trig, Retrig, /Trig) with encoder assignments which change the most important parameters of an "instrument", such as
I think a more flexible approach would allow 2, 4 or 8 voices. Especially in a modular setup the concept of 8 CVs is very powerful, going beyond just a MIDI to CV converter.
And the default encoder assignments could be:
ENC1: CV1 Octave
ENC2: CV1 Semitone
ENC3: CV1 Finetune
ENC4: CV1 Portamento
ENC5: CV2 LFO1 Depth
ENC6: CV2 LFO2 Depth
ENC7: CV2 ENV1 Depth
ENC8: CV2 ENV2 Depth
ENC9: CV3 LFO1 Depth
ENC10: CV3 LFO2 Depth
ENC11: CV3 ENV1 Depth
ENC12: CV3 ENV2 Depth
ENC13: CV3 Value (Changes Octave/Semitone/Finetune)
ENC14: CV4 Value (Changes Octave/Semitone/Finetune)
ENC15: CV4 MOD1 Depth
ENC16: CV4 MOD2 DepthENC1-3: not required as there is a dedicated octave/fine encoder + a chromatic keyboard for semitone transpose
ENC5-16: for the most flexibility I suggest all modulators of a channel are accessible on one "bank"
Same for the ARP and SEQ configuration.
Is it not better to have all 16 steps on their own encoders?
-
Im sure you guys have already thought about this...... But, will this design be eurorack friendly? i.e. Fit in a 3U rack?
I'm aiming for 3U, although I don't have definite PCB dimensions yet. Not quite sure if it will go between a pair of rails. The next question is: should it? The older idea with lots of LED rings consumes about 700mA if I remember correctly. And it would be around a whole block of 86 HP. But the IO is flexible. Either you can use the line drivers and connect up a breakout panel in the rack or else put your jacks/bananas/whatever on the box itself. As I've mentioned before my plan is to enclose a SEQ and CV in some sort of 19" case with a breakout panel and a decent PSU.
BLM 16x16+X build guide
in MIDIbox BLM
Posted · Edited by latigid on
Okay, thanks for the file, let's see how yours goes and I can add it to the OP. I didn't have issues with mine (corners are properly rounded) and I think TK.s was also good? In any case, the spacer holes are a few mm larger than the buttons.