LED's on MIDI in/out on the CORE32

[orange_hand]

Hi,

I am wondering what I need to do in order to enable the LED's for the MIDI in's and out's on the CORE32, like it is realized on the IIC board. When I read the documentation a couple of weeks ago I found somewhere an explanation, but I can't retrieve it. The only thing I remember is that it has to be in parallel to the MIDI connectors ...

Can anyone help me out with a small schematic ?

Cheers

orange

Edited July 17, 2012 by orange_hand

[TK.]

MIDI IN/OUT LEDs are no MIOS32 feature, they have to be supported by the firmware.

MBSEQ doesn't support such LEDs - but an integrated MIDI monitor screen, which is very helpful for debugging MIDI connections: MENU+MIDI, select "Misc." -> GP16

Best Regards, Thorsten.

[julienvoirin]

you have to do something similar to the GM5X5X5 board designed by nILS, using 74HCxx chip. this is NOT included in the Core32 PCB.

another trick is to put a LED in the current loop of the MIDI connector :

http://ogloton.free.fr/realisations_minutes/test_midi.html

[TK.]

The GM5x5x5 LED circuit isn't an ideal design, the LEDs are flickering too fast.

And putting a LED into the current loop means that there is no current for the optocoupler at the MIDI target anymore ;-)

Have a look into the http://www.ucapps.de/mbhp_ltc.html schematic for a perfect HW based MIDI activity indicator.

Best Regards, Thorsten.

[orange_hand]

Thanks guys for the explanation !

Cheers

orange

[gslug]

I am also looking to add activity LEDs. I have come up with a schematic based on the LTC module, but with what I assume are unnecessary parts removed. Could someone please confirm that this is all I need.

Thanks,

Kevin

Edited July 18, 2012 by gslug

[lylehaze]

I am also looking to add activity LEDs. I have come up with a schematic based on the LTC module, but with what I assume are unnecessary parts removed. Could someone please confirm that this is all I need.

Thanks,

Kevin

As drawn, this circuit will drive negative spikes to the inputs of the "upper" NAND gates.The built-in protection diodes will help, but it's hard to say how long they will last before the HC00 chip just dies,

[lylehaze]

I have attached a screen grab from the java circuit simulator to help show why.The sim is at falstad.com.. under analog circuit simulator, 555 timer, square wave generator if you want to have a play.

To show the problem, I attached a capacitor in series to the square wave output, then a resistor to ground. this is the same arrangement you have between the lower and upper NAND gates.The scope trace at the bottom shows the problem. Every time the square wave falls from 5v to 0v, you get a negative voltage spike of over 4 volts (actually, that would be UNDER -4 volts).

The datasheets of most chips require all voltages to stay between V+ and V-, or in this case 5V and Ground.

There _ARE_ built-in diodes at each input to the chip to help protect the chip from damage, but these are last-ditch protection devices, and it's not a good idea to design a circuit that requires this protection to avoid frying the chip. The particular HC00 datasheet that I looked up (no link, sorry) even puts a statement near the top of the page specifying that you should never depend on the protection diodes to save you.

Most chip datasheets offer two sets of limits.. "operating" conditions and "absolute maximum" conditions. The difference is simple.. the chip may fail outside the operating conditions, the chip may suffer permanent damage outside the "absolute maximum" conditions. As you might guess, the requirement to keep all inputs between VCC and VSS is an absolute maximum condition.

I hope this has helped to explain my previous comments.

LyleHaze

[gslug]

Thanks lylehaze. I based the schematic on the relevant part of the MIDIbox LTC module: http://www.ucapps.de/mbhp_ltc.html, so figured that it would be OK. Do you have any suggestion as to how to protect the inputs of the gates?

Thanks,

Kevin

[lylehaze]

Thanks lylehaze. I based the schematic on the relevant part of the MIDIbox LTC module: http://www.ucapps.de/mbhp_ltc.html, so figured that it would be OK. Do you have any suggestion as to how to protect the inputs of the gates?

Thanks,

Kevin

I'd try to avoid any negative voltages in the first place. The attached drawing is untested, PLEASE build a test circuit and experiment with the R and C values before committing anything.Without any timing delays, you get such a brief flicker of light that it's hard to see. So the question becomes, how long should a minimum flash length be? You get a longer LED ON timeby increasing resistor ohms OR increasing capacitor uf. Lesser values for either will make a shorter flash.I drew this using inverters, you may use NAND gates instead if you want, just gang both inputs together.Bonus Design goal: CMOS chips WILL operate in the linear region (as they do in this circuit), but they draw less power when run with hard ON or OFF signals.. You can get that by using this same circuit but choosing inverters with "schmitt trigger" inputs. Basically these are inputs that "snap" high or low when the input voltage is wandering around in between. What you get in return is a bit less power draw and LED's that are ON or OFF instead of some (minor) "fading" during the delay portion.

Please experiment before deciding what values of R and C to use.. I just threw some random numbers into the simulator.

Have Fun.

[gslug]

Thanks again lylehaze. I like the idea of using inverters with schmitt triggers. I'll get some parts and have a play on a breadboard.

Kevin