AINSER deadband configuration?

[FantomXR]

I was at my local electronic-shop today and got a LM3940 which makes 3.3V out of 5V. And here is something strange:

Lets take the original schematic from the AINSER64 which is basically what I use. If I do not set the jumper J5 at all, I still am able to read out the potentiometer with nearly the full range (starts at 1 instead of 0 and goes up to 125 instead of 127). I'd expect that I either get jittering values or I'm not able to read out the potentiometer in total. 

I tried to connect the 3.3V to the Vref of the MCP3208. Now I'd expect that I can read out the faders but wouldn't reach 127, because I have not configured them in the NGC via pinrange for 3.3V. But nothing changes here... I get the fullrange of the potentiometer. But jittering continues as soon as I change the colors on the LEDs....

//edit: I measured the voltage when connecting 3.3V to VREF. It seems to be right. At the faders I measure 3.3V and not 5V anymore. I'll do another test with a separate power supply for the LEDs again...

Edited September 12, 2017 by FantomXR

[Zam]

Hi Chris

 

53 minutes ago, FantomXR said:

Lets take the original schematic from the AINSER64 which is basically what I use. If I do not set the jumper J5 at all, I still am able to read out the potentiometer with nearly the full range (starts at 1 instead of 0 and goes up to 125 instead of 127). I'd expect that I either get jittering values or I'm not able to read out the potentiometer in total. 

That's strange...without any Vref it should not work...I don't have a look at 3208 datasheet recently but don't remember any "sens" that use VDD is case Vref is not here ???

 

56 minutes ago, FantomXR said:

I tried to connect the 3.3V to the Vref of the MCP3208. Now I'd expect that I can read out the faders but wouldn't reach 127, because I have not configured them in the NGC via pinrange for 3.3V. But nothing changes here... I get the fullrange of the potentiometer. But jittering continues as soon as I change the colors on the LEDs....

You get full range because it is given by  Vref ! If Vref=3.3V you get max value when AD input is at 3.3V

Definitely thinking your led draw enough current to cheat your 0V.

Is the jitter only at the instant you switch led or change color, or it  jitter stay as soon as you have led ON ?

Again, check for strong as possible GND path from LED pcb to PSU, as AINSER pcb to psu.

 

Do you have pcb design of your new layout?

 

Best

Zam

 

 

[FantomXR]

On 12.9.2017 at 1:53 PM, Zam said:

Definitely thinking your led draw enough current to cheat your 0V.

Might be! But today I tried to add two 0.5mm wires for GND and VRef.... without any success.

I checked again the NGC-File. I was wrong. I checked the faders at 7bit, which works okay and not with 11bit like I stated above. The jitter only happens when switching colors. With 11bit it doesn't even matter if the LEDs are connected. The fader jitter all the time. 

Anyway: I really do not need 11bit on the faders. 7bit is totally sufficient. 11bit I do only need for the pitchbend-wheel. All other controllers (also foot controller) will work fine with 7bit. So if it would be possible to set the resolution on a per-pin-base, it would be the easiest solution. Maybe I could hardcode the pitchwheel.... I think I need to dig deeper into the code.

Maybe @TK. can help here?

[FantomXR]

@Zam: Just a quick note on this: I tested several things on my PCBs but I was not able to get rid off these value-jumping-issues when changing the colors on my WS2812B. After researching I found a little chip called MCP1541 from microchip. This generates a super stable reference voltage for the MCP3208 out of 5V. Even out of the noisy USB-5V. And this solved all problems. Such a great IC. Also very low component count. MCP1541 + 2 caps and you are done. 

 

Edited February 20, 2018 by FantomXR