Console signal flow, input cards, general discussion on API styled DIY mixer

[Fuccimain]

i'm a little lost honestly, I know the 325 has more gain in it than I should need if wired for the full 40db, as an input channel,

not exactly sure how to go about making a level diagram, I did find that the 325 when running at full 49db gain has amplifier noise of -125dbm at 20khzyou saying i should configure my inputs for less gain..?

trying to learn.. ]

also, about increasing the fader buffer by 6db, not going to ask for the correct feedback resistor value to make this happen, but perhaps the equation? so I can learn how to calculate this..

thanks

 

[Harpo]

also, about increasing the fader buffer by 6db, not going to ask for the correct feedback resistor value to make this happen, but perhaps the equation? so I can learn how to calculate this..

+6dB (or more exactly +6.02dB) is voltage gain of 2.
Voltage gain of your 2520 DOA is set by 1+R3/(R2+GainTrimRes) with -from original schematic- feedback resistor R3=47K and min.value shunt arm resistor R2=47R.
You get max.gain with GainTrimRes dialed in for zero ohm and -ignoring unity gain config by leaving out this GainTrimRes- you get usable min.gain with GainTrimRes same value as feedback resistor. For a more linear dB increase when dialing from its CCW to its CW end, this GainTrimRes rheostat will have a rev.log taper law.
From the 1+... you know, voltage gain won't drop below 1 for this non-inverting gain stage.
To get a voltage gain of 2, just make feedback resistor same value as shunt arm resistor, IE GainTrimRes=47K.
(For voltage gain of 2, the 47R will vanish in 47K parts tolerances from previous example values).
Double check dB readout by LOG(voltage gain)*20 = LOG(1+47K/(47R+47K))*20 = x dB.

This calc is ignoring the AC gain calculation for the HPF and LPF setting C3 and C6 for now.

 

[Fuccimain]

hry thanks Harpo..

I follow you if working off the orig 325 schematic.. 47R is the 470 ohm resistor?,.. but I guess i'm unsure which resistor is the R2 on the fader buffer schematic I'm working off of... it would be the 20k thats tied to ground via the 47 uf cap?

LOG(voltage gain)*20 = LOG(1+47K/(47R+47K))*20 = x dB
LOG(1+47K/(47,470))*20 = x dB
LOG(1+.99009901)*20 = x dB
LOG(1.99009901)*20 = x dB
LOG(1.99009901)*20 = 39.801980 dB

im doing something wrong

 

[Fuccimain]

I haven't updated the drawing yet, but as Abby road said a few posts ago, I'm planning to change feedback and shunt resistors to 10k to lighten the load on the 2520

 

[Harpo]

I follow you if working off the orig 325 schematic.. 47R is the 470 ohm resistor?,..

oops typo. yes 470R.

but I guess i'm unsure which resistor is the R2 on the fader buffer schematic I'm working off of... it would be the 20k thats tied to ground via the 47 uf cap?

yes. 10K from your last post.

LOG(voltage gain)*20 = LOG(1+47K/(470R+47K))*20 = x dB
LOG(1+47K/(47,470))*20 = x dB
LOG(1+.99009901)*20 = x dB
LOG(1.99009901)*20 = x dB
LOG(1.99009901)*20 = 39.801980 5.977 dB

or with your updated schematic LOG(1+10000/10000)*20=6.02 dB
The 68pF (not polarized) cap across the now 10K feedback resistor set LPF to 234kHz, your DOA hopefully can handle without running out of GBW.
The DC blocking 47uF in series to the now 10K shunt arm resistor set HPF to 0.34Hz.

 

[Fuccimain]

Guess I thought the astrik was multiply but that would be x, what is the astrik doing in the equation?

I'm gonna play some more with these numbers.

Thanks again.

 

[Harpo]

Guess I thought the astrik was multiply but that would be x, what is the astrik doing in the equation?

multiply.
Voltage gain is 1+10000/10000 = 2
LOG(1+10000/10000) = LOG(2) = 0.301
LOG(2)*20 = 0.301*20 = 6.02dB

LOG(x) is the logarythm of a positive real number x with base 10. Some ancient calculators use the function LOG10(x), giving the same result.
No idea how your calculator comes up with 39.801980 dB for previous values, that would be LOG(1.99009901)*133.173 or LOG(97.74)*20
Edit: Found your fault. You wrongly calculated 1.99009901 * 20 = 39.801980.

 

[Fuccimain]

Oh also, if I do change the feedback resistor to 10k, it's going to increase the gain of the 2520? Is that how it optimizes noise?

If you change both resistors, gain will be unchanged. However the impedance seen by the negative (inverting) input will be halved, which is good for noise performance.

formula makes sense now,.. but as quoted, ... the formula says changing the feedback and shunt arm resistors to 10k will raise the gain 6.02 ddb,.. but abby road said gain would be unchanged... so I'm curious about the discrepancy.... did I miss something?

 

[Fuccimain]

Unless its been running at 6db the whole time...

 

[Harpo]

Oh also, if I do change the feedback resistor to 10k, it's going to increase the gain of the 2520? Is that how it optimizes noise?

If you change both resistors, gain will be unchanged. However the impedance seen by the negative (inverting) input will be halved, which is good for noise performance.

formula makes sense now,.. but as quoted, ... the formula says changing the feedback and shunt arm resistors to 10k will raise the gain 6.02 ddb,.. but abby road said gain would be unchanged... so I'm curious about the discrepancy.... did I miss something?

With feedback resistor same value as shunt arm resistor, voltage gain in a non-inverting opamp gain stage is always 2 or +6.02dB in its passband, no matter if these resistor pairs are 10K, 20K, 47K or another same value pair, so gain is unchanged. What changes/decreases operating at same temperature is the resistive noise figure from the lower value resistors.

 

[Fuccimain]

Gotcha. What does GBW stand for? YOu think I should change the values of those caps? Curious how to determine LPF and HPF, is it better to be closer to the audio range of 20-20? I'm gonna try to figure out what value resistors using that formula to increase another 6db.

Thanks!

 

[Fuccimain]

So to increase 12db.. Any advantage using  20k Rf with 6k6 Rs or 10k Rf with 3k3 Rs? Both are a gain of 12db.... Curious...

 

[Fuccimain]

I guess using 10k feedback resistor will lower the load on the input of the 2520, curious if there are any other issues or draw backs using one vs the other from last post.

 

[Harpo]

What does GBW stand for?

GainBandwidth. For your +6 or +12dB wanted gain probably not your concern. If you change your mind for +40dB from 1st.post, things might change.

YOu think I should change the values of those caps? Curious how to determine LPF and HPF, is it better to be closer to the audio range of 20-20?

At cutoff frequency signal is already -3dB down. For a more linear response in your assumed 20Hz...20kHz audio range you might set these filters a decade below and a decade above these numbers, IE 2Hz...200kHz. These are 1st.order filters with a -6dB/oct. or -20dB/dec. slope for -3dB cutoff frequency in Hz = 1/(2*PI()*R*C) with R in ohm and C in Farad. At some point caps size will matter or your opamp/DOA runs out of of its safe operating area.

So to increase 12db.. Any advantage using  20k Rf with 6k6 Rs or 10k Rf with 3k3 Rs? Both are a gain of 12db.... Curious...

Resistive/Johnson noise will be -3dB down with your half value parts config. IE the 20K resistor introduces -111.8dB, the 10K -114.8dB resistive noise.
[Johnson noise in dB = LOG(SQRT(4* 1.38E-23 * 300 * 10000 * (20000-20))*20] with Boltzmann constant 1.38E-23 Joules/Kelvin, assumed room/box temperature 300 Kelvin, 10000 your tested resistor value in ohm and bandwidth 19,980Hz for your 20...20kHz audible range of interest.

I guess using 10k feedback resistor will lower the load on the input of the 2520, curious if there are any other issues or draw backs using one vs the other from last post.

A lower value resistor increases the load. Your opamp/DOA doesn't come with infinite drive ability, so by further decreasing your 10K/3K3 down to 10R/3R3 as an extreme example for same +12.1dB gain, you probably trade decreased resistive noise for increased smoke.  ;D

 

[joaquins]

A lower value resistor increases the load. Your opamp/DOA doesn't come with infinite drive ability, so by further decreasing your 10K/3K3 down to 10R/3R3 as an extreme example for same +12.1dB gain, you probably trade decreased resistive noise for increased smoke.  ;D

And distortion, smoke could be a nice touch in live applications... 8)

JS

 

[Fuccimain]

What does GBW stand for?

GainBandwidth. For your +6 or +12dB wanted gain probably not your concern. If you change your mind for +40dB from 1st.post, things might change.

yeah, most of what we have been talking about has been regarding the fader booster circuit,.. the 40db was a reference to the input side of the channel, which I haven't discussed all that much.. but yeah. if 40db seems like too much gain for an input channel i'll need to figure out what to do.. I'd like to be able to add or remove gain to the line level inputs (via API 325 circuit) then go to the fader buffer circuit, then to pan and mix buss.

thanks for the frequency formulas,.. I'm going to play with some numbers and get to a decade" above and below the 20-20 range and see if I can do the math correctly.

 

[Fuccimain]

Oh also, if I do change the feedback resistor to 10k, it's going to increase the gain of the 2520? Is that how it optimizes noise?

If you change both resistors, gain will be unchanged. However the impedance seen by the negative (inverting) input will be halved, which is good for noise performance.

A lower value resistor increases the load. Your opamp/DOA doesn't come with infinite drive ability, so by further decreasing your 10K/3K3 down to 10R/3R3 as an extreme example for same +12.1dB gain, you probably trade decreased resistive noise for increased smoke.  ;D

guess I mean't "halved" instead of lowering the load, sometimes I forget to refer to Ohms in reverse

 

[Fuccimain]

Ok. Well played with the frequency formula a bit.  If I change the 68pf to a 75pf it gets me closer to the 200khz at 212khz, and if I change the 47uf to an 8uf it gets me to 2hz... Any drawbacks to those changes?

Curious, feedback resistor plus parallel cap always determines LpF and shunt arm res in series to cap to ground HPF? For non-inverting opamp only? Or inverting as well?

 

[Fuccimain]

although, from earlier in this thread, the API schematics for the fader buffer have the feedback cap at 68pf and the shunt arm cap at 47uf  ... so if it's good enough for API, do I really need to change the HPF and LPF cutoffs?                                                                      d                                                                                                                                                                                                                                                                                          Just thinking that perhaps they chose those values for a reason that we haven't possibly considered and changing them is going to mess soemthing up... If I change the voltage of the 2520 to 2, that will give me 12db "in hand" for the fader correct?

What to do about the input.. I understand the logic behind the level diagram, just not sure If I have all the info to create one yet... I know that having 40dm for the 325 input channel is perhaps I'll limit myself to +20db and perhaps -20db as well... if i put a t-pad at the input to pad down from unity, and then fine tune the gain setting.. if i increase the gain, i'll also be increasing th noise floor level... perhaps I don't need to do this,.. but it would be good to perhaps pad the input -6db so the channels can run at unity with the +6db increase from the output transformers.

 

[Harpo]

although, from earlier in this thread, the API schematics for the fader buffer have the feedback cap at 68pf and the shunt arm cap at 47uf  ... so if it's good enough for API, do I really need to change the HPF and LPF cutoffs?

Context.
From API schematic, they use a 68pF in parallel to a 20K feedback resistor, giving a LPF at 117kHz when used with a API 2520 (maybe Huntington) DOA to keep it in its safe operating area. From your last plan you have half of this resistive feedback value, so you'd need to double up the cap value (68pF||68pf) for same response. If your DOA (whatever type this might be, comming with whatever OSI) can do better in this context than this ancient API part, you can increase this LPF for a more linear response in the assumed audible range of interest.
The 47uF in series with 20K shunt arm resistor from API schematic set a HPF at 0.17Hz. Your 10K shunt arm resistor would require a doubled up 47uF for same response, else this HPF would rise to 0.34Hz. I wouldn't care, but for investigating seismic figures YMMV.

If I change the voltage of the 2520 to 2, that will give me 12db "in hand" for the fader correct?

If 'voltage' would be voltage gain and connecting to a 1:2 wired transformer for +6dB iron gain, yes.