OTB Mixer Using Current Summation
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
I'll have to find more time to think in depth about this, but off hand the #1 things that keeps me away from the standard summing box is lack of linear-style long throw faders and pan. I really pooh-pooh all over the idea that panning and submixing in a DAW and routing to an analog box lacking pan is somehow better than staying ITB. If I want OTB, I want OTB!! That means pan and faders I can grab and drive for max ergonomics and efficiency. The lack of ergonomics with a DAW is my primary complaint. Everything is 400% slower than driving a traditional control surface. Make a traditional control surface that addresses the minimal needs of DAW track output mixing and we have something. I'd find mutes of use, and a few aux sends would be nice. Mute can be done in the DAW, but I like the idea of a hard analog mute. Solo can be covered fine in the DAW usually. Sends could even be bussed up (dirty! bad!) in the DAW and output directly to reverbs, etc. More later....
Best,
Doug Williams
Electromagnetic Radiation Recorders
Doug Williams
Electromagnetic Radiation Recorders
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
Yes the concept of voltage control for all variables, makes it very interface friendly (and smaller). I see all the boys and girls who already tried to make a DIY OTB mixer, as target customers, to upgrade the insides of their existing mixer.
Re: real pans, yes, from my research last weekend, one of the complaints about ITB was the ergonomics. We can add mutes pretty simply by clamping the VCA gain at -xx dB. Might have to add a cap to prevent clicks. Though I still want to stay with KISS. I don't see this as a full control surface, just a very high performance 2 bus.
But I'm not the one melting solder over this so keep the cards and letters coming.
JR
Re: real pans, yes, from my research last weekend, one of the complaints about ITB was the ergonomics. We can add mutes pretty simply by clamping the VCA gain at -xx dB. Might have to add a cap to prevent clicks. Though I still want to stay with KISS. I don't see this as a full control surface, just a very high performance 2 bus.
But I'm not the one melting solder over this so keep the cards and letters coming.
JR
Cancel the "cancel culture", do not support mob hatred.
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
Ok, after wasting too much time and brain cells on the PAN law inspection, I have come to the conclusion there is no arguably "correct" pan law. The constant power law (which they don't really follow) doesn't hold water for coherent sources, since the power is not in fact constant when SPL or voltages are recombined, while I prefer the feel of -3dB law, versus -6dB which is arguably more correct for voltage summing (mono'ing the master will make this apparent).
The THAT circuit is symmetrical, in the region where both L & R signals are significant, so while not true constant power (which isn't symmetrical between 0dB and -3 dB), as good as any other estimate. I am reluctant to throw more complexity at this (more log conversions) for a pan law who's primary function is to save having to retrim the fader lever after panning a signal. Since I suspect most people trim fader moves after major pan changes anyhow, this doesn't warrant reinventing a rounder wheel. I very much like the 0-5V control standard.
My vote is to adapt this application note, for perhaps newer opamps, and maybe scale values for improved noise/DC offset sensitivity. Only the very last stage needs to be low noise, while all will benefit from having low offset voltages.
I would propose also accommodating the simpler switched pan control in the layout, and add mute function capability (switch closure to ground). The mute may need some bench tweaking to make sure it doesn't click. and un-mutes gracefully (real mutes click on signal, but this doesn't have to).
My engineer side is wanting to put cheapo GP opamps in side chain path where they don't matter, but I suspect folks building these DIY would prefer better opamps So I'm inclined to design for generic duals which will accept almost any substitutes.
JR
The THAT circuit is symmetrical, in the region where both L & R signals are significant, so while not true constant power (which isn't symmetrical between 0dB and -3 dB), as good as any other estimate. I am reluctant to throw more complexity at this (more log conversions) for a pan law who's primary function is to save having to retrim the fader lever after panning a signal. Since I suspect most people trim fader moves after major pan changes anyhow, this doesn't warrant reinventing a rounder wheel. I very much like the 0-5V control standard.
My vote is to adapt this application note, for perhaps newer opamps, and maybe scale values for improved noise/DC offset sensitivity. Only the very last stage needs to be low noise, while all will benefit from having low offset voltages.
I would propose also accommodating the simpler switched pan control in the layout, and add mute function capability (switch closure to ground). The mute may need some bench tweaking to make sure it doesn't click. and un-mutes gracefully (real mutes click on signal, but this doesn't have to).
My engineer side is wanting to put cheapo GP opamps in side chain path where they don't matter, but I suspect folks building these DIY would prefer better opamps So I'm inclined to design for generic duals which will accept almost any substitutes.
JR
Cancel the "cancel culture", do not support mob hatred.
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
Yes and no. In a perfect acoustic space a mono sum would be +6dB and in a terrible room it will be +3dB. Many fudge it with +4.5dB.JR. wrote:I have come to the conclusion there is no arguably "correct" pan law. The constant power law (which they don't really follow) doesn't hold water for coherent sources, since the power is not in fact constant when SPL or voltages are recombined, while I prefer the feel of -3dB law, versus -6dB which is arguably more correct for voltage summing (mono'ing the master will make this apparent).
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
OK, single opamps, at least for critical audio path.
Yes I'm familiar with general practice in console design. My point about the pan, is irrespective of the target pan law, few (any?) are very strict about following even that target. In principle one could make an accurate derived -6dB pan law by subtracting one channel from unity to get the other, and that would always sum to one, but for constant power you need the signal squared to sum to unity, (not so simple). For a 4.5 dB law, it doesn't even have a theoretical basis other than being none of the above.
I suspect a continuous pan with optional laws (-3 dB, -4.5 dB, -6dB) should cover the meat of the user base. With pan switch option for KISS crowd. That said if we want to buffer the VCA control voltage, even when using switch for pans, perhaps just use the voltage input into pan law circuit and just create a simple voltage divider with the pan switch.
OK, back to one version.
JR
Yes I'm familiar with general practice in console design. My point about the pan, is irrespective of the target pan law, few (any?) are very strict about following even that target. In principle one could make an accurate derived -6dB pan law by subtracting one channel from unity to get the other, and that would always sum to one, but for constant power you need the signal squared to sum to unity, (not so simple). For a 4.5 dB law, it doesn't even have a theoretical basis other than being none of the above.
I suspect a continuous pan with optional laws (-3 dB, -4.5 dB, -6dB) should cover the meat of the user base. With pan switch option for KISS crowd. That said if we want to buffer the VCA control voltage, even when using switch for pans, perhaps just use the voltage input into pan law circuit and just create a simple voltage divider with the pan switch.
OK, back to one version.
JR
Cancel the "cancel culture", do not support mob hatred.
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
I think the only useful metric of pan law is spl, and that is a moving target. Hence the fudge. I have a feeling I don't quite get your point though.
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
Sorry if I'm not clear...Gold wrote:I think the only useful metric of pan law is spl, and that is a moving target. Hence the fudge. I have a feeling I don't quite get your point though.
Pan law is for the convenience of the mix engineer, so they don't have to drastically readjust the fader when altering the pan.
#1 there is no perfect law that delivers this, for all situations.
#2 the popular pan implementations don't even accurately follow their target laws.
My point is that I will not waste any more time trying to perfect a circuit that can't be perfected, and is only an estimate to start with. Life is short and I'm not getting younger.
JR
Cancel the "cancel culture", do not support mob hatred.
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
Change the input opamp U5B from grounded inverting, to simple differential. The resistor from +5 to + input needs to be 49.2k (49.3k 1% sounds close enough).
Just lose U3A and U3B output opamps. Connect the VCA outputs (8) directly to L bus and R bus.
JR
PS- I will do mute later- to keep with voltage controlled concept I am leaning toward 0v= on, and 5v = mute. Will take two steering diodes and two resistors to sent mute current to two different opamps. May want to add RC to slow down edge rate of mute control signal.
Last edited by JR. on Wed Dec 10, 2008 1:30 pm, edited 2 times in total.
Cancel the "cancel culture", do not support mob hatred.
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
This should be tuned for several hundred kHz but needs to be bread boarded to confirm it doesn't scream...
Bus should be cap coupled or servo'd if X VCAs introduce DC error. As feedback VCA is turned down, any DC error would be result in high DC gain.
!00k (could be as low as 20k) drains off any DC current to keep VCA in good DC operating point. Ground 100K at master opamp +.
To test you can just build one opamp with no input VCAs.. more VCAs will probably make it more stable.
Sum bus opamp should be decent quality, low noise, but will be operating at low noise gain so not working very hard.
JR
Last edited by JR. on Wed Dec 10, 2008 1:32 pm, edited 1 time in total.
Cancel the "cancel culture", do not support mob hatred.
Re: Pease - What's All This Adjustable Slew Rate Stuff Anyhow?
That was my point too!JR. wrote: My point is that I will not waste any more time trying to perfect a circuit that can't be perfected,