Matrix mixer expanded edition

a) ok, I was uncertain about that. …soo many opamps :slight_smile:
b) when I try that in the simulation, I can’t get it to work properly . There has to be some amplification else the signal is too weak (300 to 500mA)
c) do you mean, 3x jacks straight into the clipping stage, bypassing the attenuverters?

1 Like

update

It’s great to see the progress on your matrix mixer @reverselandfill & especially to see the skill sharing dialogue between you @transistorcat :+1:t3:
Regarding the pcb layout, my only suggestion would be to use the Analog Devices ADA4851-4 IC, its 4 op amps for about $2.50 and would substantially reduce the cost of a complete DIY through-hole version when compared to the price of the LM6172. The ADA4851-4 is an smt component but maybe if can be mounted/fitted at the factory where the PCB will be produced. Just a suggestion, nothing more.

Here’s a link to the datasheet.

@reverselandfill, the resolution is quite low on the screenshot so I can’t read the fine print. I count 19 op amps, are they all LM6172s?

there are 10 opamps in the latest version (dual LM6172’s)
I’ll check the ADA4851. They are 1,89 euro when you buy +100 at Mouser.
That is pretty cheap. runs on ±5v . ±12v? not sure is this is possible/ or needed.
What do you think @transistorcat?

SMD might be necessary, because of the amount of parts . I was already considering this.

1 Like

Looks like JLC has both 6172’s and 4851’s in stock for some auto-populated boards. Thats who I use for smt boards.

1 Like

yes me too. might be an option then. Although this would be the first time that I would use pre-populated boards… exiting :slight_smile:

2 Likes

I can help you out if you need. For auto populated boards, you need to create a BOM and Centroid file pertaining to their format and file extensions.

1 Like

Thanks for the offer! We’ll see when the moment get there.
I’m unsure as of now if I go that route. (SMD and pre-populated boards)
First things first: getting all the features packed into it :slight_smile:

1 Like

For A) and B):

A) This should work, the output is still looking into a strong driver at the same level through the same 499R resistor.

B) This should work, as long as R1 and R2 are at least as large (ideally larger) than Rm.
My guess is that R2 should be ~2x R1 or slightly smaller than that, so having them be e.g.12k and 24k should give ok gain and a linear enough response.

For C) i was thinking like the THRU inputs on the processors, after both the limiter and the master gain.
When i suggested it i was thinking to just sum it into the post-limiter amp (assuming the pot is moved), but that would leave it inverted, i guess.
Pre limit like you suggest might be the best, and that should be possible without requiring any extra amps.
The idea is that if you have a full RGB background, there is currently no way to get it into the mixer without using up all three inputs

wrt. smd i can easily solder anything that doesn’t require reflow (like BGA and QFN) so for me this makes little difference. The lower cost should be weighted against the extra cost/work for the 5V regulators and the required protection diodes on the input, as well as the DIY-friendliness for others around here

1 Like

If you’re going to be using the ADA4851, the input buffers from LZX Patchable Standard Reference Design Schematics is a very good starting point

1 Like

this looks ok. if I make r2 82k and r1 12k, it comes to 1v output.
The 3x gain switch takes it to 1.2v and has a rounded triangle shape.
if the input voltage is 3v , it becomes a weird sine shape of about 1.7v
The max voltage can be lowered with r2, so I can adjust that value when I get the prototype.

I’ll work on the ext inputs. if I use the SMD opamps, maybe I have some to spare to invert that signal before that last opamp (so that becomes non-inverted)

2 Likes

I’m reviving this project.
I have to refresh my memory a bit, as this project got quite expansive !
Maybe I’ll do a TH prototype version first (with 10 lm6172’s) to see if it all works
I don’t yet understand how to replace the LM6172 with the ADA4851 in this whole project.

6 Likes

very excited to see this moving along again!

1 Like

What’s the current list of new features? :slight_smile:

1 Like

3x3 matrix with attenuverting pots per channel
3x gain control per row
3x switches to set the gain to 1x or 3x with diode clipping
3 inputs
3 outputs
1v normalled to RGB channels for matte color mixing

I might add extra inputs and outputs, but I have to see what type of config I will use for that.

8 Likes