VxF vantanalogue RGB crossfader

Hi Aladan,
how’s the noise on the ribbon cables working for you?
I tested the same idea and got almost always some noise. At the end I’m going for 90 degree headers, that’s the funny use I was referring to. Sorry Robert, no patch bay :confused:
The idea is ‘prepatchability’, so some inputs and/or outputs are already connected in the back of the modules. Plus a matrix mixer can be extended ‘indefinitely’ in the number of inputs.

1 Like

I put a ground on every second pin so it seems to be fine. I can very occasionally see noise in a solid screen of around half brightness (0.5V) of a single colour only, but I believe that’s more a factor of my power supply or maybe even my cadet encoder, because it happens even with other modules.

The only issue that would be nice to fix up is very slight colour bleed/crosstalk if one colour channel is black/0V and the one next to it is a comparatively high voltage (e.g. >5V). But given video voltages are only meaningful in the 0V-1V range it’s usually avoidable by winding back the input attenuator a bit.

I’ve using 90 degree headers too, but the builder is free to do whatever they want and the non-90 degree ones will work just fine. Soldering individual shielded cables would be an option too.

1 Like

I had some time to work on it and its almost there. I’ll order pcbs and test.
The changes have been:
4 instead of 3 ‘vaders’
vaders can be 2quadrant or 4quadrant (mix of cadet fader and cadet multiplier)
Top inputs (top of the vader), normalised to 1v and cascaded
Bottom inputs (bottom of the fader) normalised to ground and cascaded
mix output, inverted mix output and luma of v1, v2 and v3 as output.

Usage scenarios:

  1. solid color:
    connecting outputs 1,2&3 to rgb, you can get a precise color using v1,v2&v3
  2. Attenuator x4:
    connecting your signal to ‘top’ inputs, you get an attenuated output using the vaders
  3. Mixer:
    connect your signals to top0 to top3 and mix with the vaders to get the output at mix and inverted at -mix
  4. Multiplier x4:
    multiply cv and topInput with switch in 4q (then get the mix out)
  5. Fader x4:
    fade between top and bottom inputs with switch in 2q either with the vaders or with a cv signal (then get the mix or luma out)
  6. kind of ‘soft key’ x4
    Turn the gain all the way up and the signal at cv will overdrive, making ‘solid shapes’. dirtier then any other solution for this
  7. RGB offset:
    Connect two rgb sets to top1,2&3 and bottom1,2&3 and fade individual colours with the vaders, or all at the same time (with offsets) thanks to the cascaded inputs.

of course you can mix and match those since they are 4 independent circuits. :slight_smile:
the xxx output is to be decided, right now is a XOR between out0 and luma, what do you think?

8 Likes

I’ll do another topic for ordering,
but for now, please like this message if you’d consider to get one. It’s just for me to estimate how many pcbs should I order.

9 Likes

What do the “direction” knobs do?

1 Like

Thats what’s called offset in the Cadet Fader: left is -1 of cv, center 0, right +1. Gain will stretch those margins beyond ±1 of the input.

Direction is the pot on the top of the simulation, v is bottom left and gain is in the center next to the 330ohm resistor.

2 Likes

I would be interested in a pcb/panel combo when they’re ready

3 Likes

This looks super hot. Count me in for one! if you can set me up with a kit or pcb/panel set I’d be stoked!

Side note if you print the panel in the same finish as the Orion/Castle modules I’d be ecstatically happy.

3 Likes

What does the Luma output mean? Is it intended to be a weighted mix of RGB like all the LZX modules that use Y to indicate that or is it an arbitrary mix?

1 Like

Hey @VanTa let me know if you want a second pair of eyes on the schematic just as a sanity check before you send away for the hard copies. What CAD are you using? I can maybe share my custom DRC file.

2 Likes

Yes, it is an averaged percepted Luma from the outputs 1-3, not including 0. Similar to Y.
Rx0.3, Gx0.55 and Bx0.15 , its my own approximation and what I use in shaders since a while.

Awesome @Fox ! Thats fantastic.
It is of course based on the Cadet Fader, plus a switch to turn it into a Multiplier.
I added a potentiometer in series on the amplifier buffer (min is 2.5x and goes up to ‘don’t know’) VR5 and VR6.
Thats where my main question is. Not sure if that’s ‘correct’.

Here’s the simulation: https://tinyurl.com/yzt2tlwh

In the Inputs, I just added protection for the opamps via double diodes, taken from the ‘LZX Interface examples’ by @creatorlars
The mixer and luma section, I did from memory, so probably something is missing there. And the luma resistor values are quite odd… room for improvement.

On the power side, nothing fancy.
I think 7805 and 7905 are a correct method to get ±5V, if anybody has a better idea…

For my first electronics project (beyond micro controller based stuff), I’m quite happy. :slight_smile:

1 Like

I have a few suggestions for you, Vanta, so I’ll just send them over via PM.

1 Like

any updates vanta? been super stoked about this idea

Some updates have been made.
Thanks a lot to @Fox and @Zifor for helping me out figuring out the most basic stuff :relaxed:
The main change is a dual gang pot for the ‘direction’ allowing me to attenuate clockwise and invert+offset cow, although not linear. Expect weird madness there :wink:
This also means a ‘exotic’ part to source, for example right now not stocked at thonk. Maybe only kits and no panel+pcb.
Thanks to @Fox the luma output should be correct now and the last output is now an inverted Luma on the 0-1V range.
Here’s the schematics:

If nobody sees a major problem in there, I’m going to route it and order a test pcb. :slight_smile:

6 Likes

Trivial feedback, instead of “top” and “bottom” perhaps just label each channel A and B
So then the inputs would be A0, A1, A2, A3, B0, B1, B2, B3, for smaller neater labels

3 Likes

Make sure to get someone to review your placement and traces , in video it is equally important they are done well

1 Like

I have to agree here as well on this. A/B numerical labels will make more sense.

2 Likes