Compact desktop line mixer?

[OneRoomStudio]

>> Here is what your LED-bargraphs look like "BEFORE" and "AFTER" by having one of the slots enlarged by 0.040" or 0.020" on each side of the LED-bargraph itself. On the left is how you currently have things fitting together by using a small hammer to force the LED-bargraph into the panel-slot. On the right, this shows a 0.020" tolerance gap all around the LED-bargraph, allowing it to easily fit through the slot and end up becoming flush with the top of the panel. For reference.....0.020" is about the standard thickness of 3 sheets of standard notebook paper pressed together, with each sheet being approximately 0.007" in thickness.

View attachment 145270

>> This image shows the LED-bargraph corner-edges and the tolerance gap clearance to the panel slot radii in each corner:
View attachment 145271

/

Thank you! I wouldn't have caught that. I'll update my file.

BTW, were you ever able to able to convert the enclosure to a 3D STL? I'd love to mess around with trying to figure out how I could fit things on the backplane PCB, so if you could share that, I'd really appreciate it.

 

[MidnightArrakis]

Thank you! I wouldn't have caught that. I'll update my file.

BTW, were you ever able to able to convert the enclosure to a 3D STL? I'd love to mess around with trying to figure out how I could fit things on the backplane PCB, so if you could share that, I'd really appreciate it.

[were you ever able to able to convert the enclosure to a 3D STL?] -- As a single enclosure file? Or.....as individual pieces???.....

/

 

[rh001]

Rather bizarre post tbh. There is still a choice of standard desktop mixers. The original question was about dedicated line input mixers. The reason lies in the evolution of audio tech.
Not bizarrely due to "population growth".
fwiw Population growth may be required to support an increasingly older non-wealth creating demographic.

It's ok to disagree but it's also ok to be polite. Why not just pretend you're Japanese. They say national hero and figure skating champ Yuzuru Hanyu is exceedingly polite...it works for him. Age: 30yrs Estimated networth $30 to $50 million.

 

[OneRoomStudio]

[were you ever able to able to convert the enclosure to a 3D STL?] -- As a single enclosure file? Or.....as individual pieces???.....

/

Ideally, as an explodable/ungroupable group of pieces (not sure if either of those are words).

 

[MidnightArrakis]

One interesting aspect of Schaeffer's front panel service is that they can use customer provided material so, rather than buy a complete enclosure like the Takachi and have to replace some of its panels with custom ones you can send the blank panels to Schaeffer for them to use on your design.

Cheers

Ian

[One interesting aspect of Schaeffer's front panel service] -- And.....speaking of "Front-Panel Services", check-out this website:

https://www.lcsc.com/front-panel

>> @OneRoomStudio may also thoroughly enjoy this website as well!!!.....

https://www.lcsc.com/front-panel/about-products

>> HAVE FUN!!! BROWSING AROUND THE WEBPAGES ON THIS WEBSITE!!!.....

/

 

[OneRoomStudio]

Ok, back to the channel circuit. I made some updates. I decided to use @JohnRoberts bicolor LED signal/peak light circuit (hopefully I implemented it correctly). I still need to determine the right mixing resistors and finalize the panning circuits.

 

[JohnRoberts]

I started to respond to this last night but saw enough differences between the circuit I designed 3+ decades ago to pause.


Here's the schematic I shared from the AMR Production series years ago.

I used a two lead bicolor LED connected between APK+ and APK-. For modest low audio signals the green LED direction starts conducting. This green direction is modulated brighter by increasing signal level until reaching the peak threshold established by resistor divider feeding the + input. When the peak direction snaps positive a simple peak hold circuit keeps the red LED illuminated long enough to easily see narrow transients.

SMPLA is the common input for multiple sampling diodes. Sampling diodes can detect for overload at multiple nodes in the channel strip. Preferably at least one sample from each polarity to detect asymmetrical waveforms. These sample diodes have their cathodes facing the SMPLA node. These sample diodes need to be DC connected to low impedance op amp outputs.

Not through two capacitors in series. I would use at least one sample diode from the input THAT chip, and another from the post fader 5534 output. Note: the 5534 is stable for gains >10 dB so the 22pF compensation cap is not necessary with that much gain. Most post fader gain stages use +10dB (3x).

JR

PS: Back when accepted consulting gigs I charged $100 hr... Now I would be more expensive and I am not looking for work.

 

[OneRoomStudio]

I started to respond to this last night but saw enough differences between the circuit I designed 3+ decades ago to pause.

View attachment 145407
Here's the schematic I shared from the AMR Production series years ago.

I used a two lead bicolor LED connected between APK+ and APK-. For modest low audio signals the green LED direction starts conducting. This green direction is modulated brighter by increasing signal level until reaching the peak threshold established by resistor divider feeding the + input. When the peak direction snaps positive a simple peak hold circuit keeps the red LED illuminated long enough to easily see narrow transients.

SMPLA is the common input for multiple sampling diodes. Sampling diodes can detect for overload at multiple nodes in the channel strip. Preferably at least one sample from each polarity to detect asymmetrical waveforms. These sample diodes have their cathodes facing the SMPLA node. These sample diodes need to be DC connected to low impedance op amp outputs.

Not through two capacitors in series. I would use at least one sample diode from the input THAT chip, and another from the post fader 5534 output. Note: the 5534 is stable for gains >10 dB so the 22pF compensation cap is not necessary with that much gain. Most post fader gain stages use +10dB (3x).

JR

PS: Back when accepted consulting gigs I charged $100 hr... Now I would be more expensive and I am not looking for work.

Thank you for that thorough explanation, John! I clearly misunderstood the input to the circuit, and this helps a lot. I will rework it. I also appreciate the free advice, and I hope that by having a discussion of the circuit here, it will help others too. Your time and knowledge are very valuable. The same can be said for the tips and advice from @Brian Roth, @ruffrecords, and @MidnightArrakis!

The magic of this place is being able to get insight into DIY projects from experienced and knowledgeable folks.

 

[JohnRoberts]

Here's another TMI anecdote that explains why I sample for clipping at multiple points inside each channel strip.

Back when I was in the trenches competing with Mackie mixers (who actually advertised their SKUs). I had one of my sales reps share an encounter he experienced when visiting one of his dealers. The dealer had a popular 4 bus Mackie sound reinforcement mixer set up with its output feeding Peavey amps and speakers on demo. When my rep walked into the store he heard obvious clipping distortion. The dealer said that the Mackie's output was so strong it was blowing away the input of the Peavey amps that couldn't handle it. Obvious BS that reveals the power of advertising to influences customer expectation bias. Apparently even that Peavey dealer was drinking the Mackie koolaid.

It turns out that that particular Mackie mixer was not sampling at multiple points of each channel strip so it was possible to adjust the channel gain structure such that it could clip internally without firing the peak LED. IIRC they only sampled for clipping after the post fader gain stage, so it was possible to pull down the fader gain low enough while clipping the mic preamp front end without ever triggering the peak LED. I actually started demonstrating that example at Peavey dealer seminars to inform and amuse Peavey dealers and salespeople in attendance. Clipping is most audible on bass, so while speaking into a handheld mic I could easily make an impressive clipping demo, without waking up the peak LED just by talking.

Ironically (perhaps?) Mackie advertised about how much headroom they had, while not accurately indicating actual clipping. Caveat Emptor.

JR

 

[OneRoomStudio]

Ok, I *think* I got everything fixed up.

@ruffrecords - I referenced your routing and mixing doc heavily for the solo/aux/panning circuits and values, so hopefully I'm on the right track there. I also employed your "ground conductor buffer" suggestion on the ribbon cable.

@JohnRoberts - I think I fixed the issues you pointed out.

This is not a complex circuit, but it's still easy to make dumb mistakes, so I appreciate the additional eyes.

 

[JohnRoberts]

Better than Mackie...

Edit: I might be tempted to suggest making the post fader gain stage differential from the wiper (10k in series with C8) to improve fader kill, while this is a bigger concern in large consoles with very picky customers. /edit

JR

 

[OneRoomStudio]

Hmm...small hiccup. 50K pots are a lot harder to find (in matching single-audio/dual-linear formats). 10K pots with 22K mix resistors could work, but maybe that's too low a load?

 

[ruffrecords]

You definitely need to look at the loading of the NE5534. It will drive 600 ohms but it makes sense to raise this as much as you can (from the points of view of power consumption, dirty power lines and distortion). The input of the signal/peak detector is no problem but you have quite a few 50K pots and 10K slug resistors to account for. It is definitely worthwhile calculating the worst case load for that chip.

More serious is the stereo AUX where you have the pan circuit driven from a 50K pot. The pan circuit will seriously load the 50K pot or at least badly distort its law and possibly that of the pan too. I agree with JR that, at least in this position, a 10K pot would be better. You could probably change all three AUX send pots to 10K and keep the two pans at 50K.

Cheers

Ian

 

[MidnightArrakis]

Ok, I *think* I got everything fixed up.

This is not a complex circuit, but it's still easy to make dumb mistakes, so I appreciate the additional eyes.

[I appreciate the additional eyes] -- Merely a "soft" suggestion here, OK??? ..... In my assuming that you are using the KiCAD software for your schematics, how about instead of creating a -- PNG - image-file -- that you do a "Plot To PDF" of your schematics. This way, your schematics will be clearer and be more easily read, while still being able to be enlarged as someone may find to be necessary depending upon the size of their monitor. Maybe something like the schematic PDF file I have attached, as an example.

Again.....just a mild suggestion.....

By the way.....I just received a notice from KiCAD of their newest "Release 8.0.8" for your downloading and enjoyment.

/

 

[MidnightArrakis]

Hmm...small hiccup. 50K pots are a lot harder to find (in matching single-audio/dual-linear formats). 10K pots with 22K mix resistors could work, but maybe that's too low a load?

[50K pots are a lot harder to find (in matching single-audio/dual-linear formats)] -- See if there is anything included in the link below that will work for you concerning this project:

https://futurlec.com/PotSliding.shtml

>> 50K Stereo Sliding Logarithmic Taper Potentiometer


>> 50K Ohm Stereo Linear Sliding Potentiometer

Even if there isn't anything in their website that you can use concerning their linear-faders, you should still browse around and take a look at their pricing for resistors, capacitors, switches and IC's, etc. Rather enticing, if I don't say so myself!!!

/

 

[ruffrecords]

[50K pots are a lot harder to find (in matching single-audio/dual-linear formats)] -- See if there is anything included in the link below that will work for you concerning this project:

I think JR was referring to rotary pots.

50K pots are common. Dual 50K pots less so.

Cheers

Ian

 

[MidnightArrakis]

I think JR was referring to rotary pots.
50K pots are common. Dual 50K pots less so.

Cheers

Ian

[Dual 50K pots less so] -- This distributor seems to have -- DUAL 50K-Ohm -- pots by the barrels-full in both "Linear" and in "Logarithmic" tapers!!!






>> General "Rotary Potentiometer" webpage link: https://futurlec.com/PotRot.shtml

>> How did the "3D Fader-Knob" printing exercise go?

/

 

[OneRoomStudio]

[Dual 50K pots less so] -- This distributor seems to have -- DUAL 50K-Ohm -- pots by the barrels-full in both "Linear" and in "Logarithmic" tapers!!!

View attachment 145458View attachment 145459

View attachment 145460View attachment 145461


>> General "Rotary Potentiometer" webpage link: https://futurlec.com/PotRot.shtml

>> How did the "3D Fader-Knob" printing exercise go?

/

The trouble is that I need a dual 50k Linear pot with center detent, as well as a single 10k audio (log) taper without detent from the same series (ideally).

The best I can find at the moment is this Bourns dual pot:


And this Bourns single pot (same shaft length, despite the picture):


Unfortunately, the single one is "Non Stocked" at Mouser.

 

[MidnightArrakis]

The trouble is that I need a dual 50k Linear pot with center detent, as well as a single 10k audio (log) taper without detent from the same series (ideally).

The best I can find at the moment is this Bourns dual pot:
View attachment 145493

And this Bourns single pot (same shaft length, despite the picture):
View attachment 145494

Unfortunately, the single one is "Non Stocked" at Mouser.

[The trouble is that I need a dual 50k Linear pot with center detent] -- Within the past few responses of mine here in your thread, I have provided you with the website link to a distributor that -- HAS -- everything that you're looking for, but apparently you can't be bothered to go check it out!!! I'm literally handing you the parts you want on a "silver-platter" and you can't be bothered to even go look!!!.....

>> YOU WANT A -- DUAL 50K LINEAR POT WITH CENTER-DETENT??? -- Well.....



[a single 10k audio (log) taper without detent] -- Which are also just a few clicks away from you just by using your mouse!!!




>> Not only this, but..... -- UH!!! Houston!!!..........There are some front-panel problems!!!.....

While I was going over your "Mini-Mixer Front-Panel" design, I came across some additional -- mechanical dimensional -- issues in your layout. However, at the moment right now, I need to take-off and have dinner with my lady-friend. So, I won't be able to detail what these issues are to you until sometime much later on this evening. But, in essence, your "patterned hole cutouts" across the panel are NOT as uniformly-spaced as you think you made them. In addition, similar to the LED-bargraph slot issue dimensions, the hole-cutouts you have made for the switches on your panel are -- THE EXACT SAME SIZE -- as the "switch-knobs" are. So, again.....you'll need a flat-blade screwdriver and a small hammer in order to get your switches to fit through your panel!!!

Anyway.....I will address your front-panel issues sometime much later this evening.

[the single one is "Non Stocked" at Mouser] -- You do realize that there are -- other -- electronic components distributors, right???

/

 

[OneRoomStudio]

[The trouble is that I need a dual 50k Linear pot with center detent] -- Within the past few responses of mine here in your thread, I have provided you with the website link to a distributor that -- HAS -- everything that you're looking for, but apparently you can't be bothered to go check it out!!! I'm literally handing you the parts you want on a "silver-platter" and you can't be bothered to even go look!!!.....

>> YOU WANT A -- DUAL 50K LINEAR POT WITH CENTER-DETENT??? -- Well.....

View attachment 145495

[a single 10k audio (log) taper without detent] -- Which are also just a few clicks away from you just by using your mouse!!!

View attachment 145496View attachment 145497


>> Not only this, but..... -- UH!!! Houston!!!..........There are some front-panel problems!!!.....

While I was going over your "Mini-Mixer Front-Panel" design, I came across some additional -- mechanical dimensional -- issues in your layout. However, at the moment right now, I need to take-off and have dinner with my lady-friend. So, I won't be able to detail what these issues are to you until sometime much later on this evening. But, in essence, your "patterned hole cutouts" across the panel are NOT as uniformly-spaced as you think you made them. In addition, similar to the LED-bargraph slot issue dimensions, the hole-cutouts you have made for the switches on your panel are -- THE EXACT SAME SIZE -- as the "switch-knobs" are. So, again.....you'll need a flat-blade screwdriver and a small hammer in order to get your switches to fit through your panel!!!

Anyway.....I will address your front-panel issues sometime much later this evening.

[the single one is "Non Stocked" at Mouser] -- You do realize that there are -- other -- electronic components distributors, right???

/

Thank you for pointing out the clearance issues on the panel. I will have to keep that in mind when I layout the final panel. Things are still subject to change since I haven’t finalized the pots or the PCB layouts yet.

The supplier website you linked to does have a lot of pots, but the selection is strangely limited - for example, while they do have 50K dual gang liner pots with center-detent, they only carry them with a shaft length of 18.5mm, whereas the single log pots they carry only come in 11mm shaft length. Yes, I could cut 7.5mm off of each dual pot…but that’s a real pain.