setting up for LINE and GTR inputs on a box.

[loss1234]

I want to be able to have Line level and guitar level inputs on my box. (as  I play synths as much as guitar)

I know/think that for guitar I want to have a high input impedance (to ground?)

most boxes seems to have a small cap (104 to 474) followed by a resistor to ground and then into an opamp.


fine.

BUT for Line, I do now want that resistor to ground and most likely want a different cap choice.

Obviously guitar needs a good amount of gain and Line needs little to none.

So my question is whether people think this is best done with a switch (to change the parts in or out) OR a separate in jack for LINE.


are there any good examples out there (schematics) that are aimed at BOTH line inputs and guitar?


any help highly appreciated!

thanks so much

[amptramp]

The issue with a separate jack is that you still have the other components connected in parallel unless you have completely separate buffer stages and combine them after buffering.  In this case, the most important function of the buffer would be to unilateralize the signal so that a signal at one input cannot appear in some form at the other input.  But combining the signals without switching off the unused source merely serves to amplify more noise, so I would insist on a switch if only one input were in use at the same time.

[loss1234]

so if i use a switch, I would need to be switching in and out, the resistor to ground and the cap? and also the feedback resistor?

is this correct?

I suppose I could use  three pole switch.

[loss1234]

OR can i get  away WITHOUT the resistor to ground for the GTR? I worry about loading issues.

also...on the output, do i once again need a resistor to ground or JUST a low a impedance?

The cap differences between a high level LINE and a guitar, seems to be important as it can cut out frequencies in an undesirable way on the synth.

there are a lot of conflicting pieces of information about this so I have never been quite sure.

Craig anderton (in device) warns against using too high of an input impedance on a single pedal as he warns that for people that chain together 5 pedals, the total impedance can cause problems.

(i may have read it wrong though...it is in his article about standards)


is there is a good article on solving all of these questions?
thanks

[amptramp]

I will try to answer these in order:

1. If you are using an op amp and you couple the guitar directly into it, you will go outside the linear operating range of the op amp because the input may swing below ground.  In this case, it helps to have a positive and negative supply for the op amp.  These supplies do not have to be equal voltage.  Once you do that, the guitar input will not need a resistor or capacitor (although I would put a 1 megohm resistor from input to ground just to avoid any possible damage to the op amp due to stray static charges when the input is disconnected).  This assumes a non-inverting amplifier stage.

2. The op amp will already have a low output impedance but it will be necessary to isolate the op amp output from the cable capacitance with a resistance in series with the cable (not connected to ground).  This can be in the range of 150 to 1000 ohms.  Also, it usually pays to have a small capacitor from the op amp output to the inverting input of the op amp to improve stability.  Stability benefits from the output isolating resistor and feedback capacitor.

3. If you are attempting to match the impedance of the line input, a capacitor must be much larger than it would be for a guitar.  But there is no reason to match an incoming line.  You can use a high impedance for both.  The requirement for switching only applies if you have both instruments connected at the same time and want to switch between them.

4. If you chain pedals in series, the input impedance of a pedal is the load for the previous pedal in the chain, so this is not an issue.  There is an issue with pedals that have low-impedance inputs but do not have true bypass operation that the low impedance will cut some of the high frequency signal.  True bypass disconnects the input and output when the pedal is not active.

[loss1234]

hey thanks

this is some great info!

i always power my opamps from +/- 12volts...so when you say pos and neg voltage, i assume you mean giving it a neg rail and pos rail into the power pins...

in this case, there should not be any need for a bias, correct?

what would be a good cap value that will work for both guitar and LINE that wont color things like drums or synths too much

i really appreciate the help!!

[amptramp]

If you have both positive and negative supplies, a resistor to ground is all that is required.  If your guitar does not have an on-board amplifier, I would not use a capacitor on the input - everything is referred to ground and all the capacitor would do is drop the low-frequency response.  For the synthesizer, I would use a lower frequency turnover than you would need for a guitar since the incoming frequencies can be lower.  If you have a 1 megohm resistor to ground, a 0.1 uF capacitor would give you a time constant of 0.1 second or an w of 10 radians/second.  this translates to a cutoff of 1.59 Hz where the signal is at -3db.  The use of 1 megohm requires a low bias current in the op amp, so a FET-input amp would usually be best.

[ashcat_lt]

I think most of it's been covered, but maybe not quite explicitly.

1)  1M input impedance is fine for both guitar and line level input.  The important point is that it be at least 10 x the expected impedance of the driving device.  There will be some small amount of extra noise from the large resistor which could be eliminated for the line input, but I bet it won't be enough to bug anybody.

2)  The cap value is generally independent of the source impedance.  You calculate the cutoff frequency using the input impedance.  So it should cutoff the same for line as well as guitar, assuming you're not changing the in-Z.

3)  Most bass amps have both Hi nd Lo inputs to accomodate both passive and active basses.  Since an active bass output is pretty close to line level as is, it might be worth your while to look at some schematics.  Most of the time the difference between these is a simple voltage divider for the "Lo Gain" input.  Doesn't actually change the amount of gain, but attenuates the stronger active (line level) signal before the gain stage.  This could be accomplished with a switch about as easily as using a seperate input.

[Pablo1234]

10 x the impedance? I know the higher the resistance values in a gain stage on an op amp will drive the S/N ratio down , lets say for example is a tlo 72 in inverting configuration. My single coil pickups have on avg a 7k impedance (generically). so your saying for single coil pickups that I should be roughly over 70k input impedance to be safe? I have always went with 220k just because i have allot of them lol, but I think I may be able to reduce some noise if I lower it. But if I can go as low as 50k or so what would be the optimal impedance match? How much will going from a 500k input I to a 75k input I reduce noise and improve the quality of the signal and at what rate will it become diminished returns?

Currently my compressor/dist board is tuned really well for my double coil pups with 1 tone stack selected (a midrage boost), and a 1.7kHz 3 Db dip for my neck and mid pickups. I also have a high pass filter after the distortion to control how much fuzziness I get with a cuttoff at around 1kHz and variable to about -6 db to +9 db. I want to and a bandpass / notch filter between the ranges of 800Hz-3k Hz for a bit more fine tuning though.

I have a few fenders with single and humbucker pickups that drive at different levels and have vast changes in impedance and I find some pedals to work better with certain guitars due to the vastly different tone stacks in my peddles and impedance of the guitars. I may just have to try out a few different input I's for each guitar to see if maybe I may be able to eliminate some tone stacks I have made. Probably should have started at that point insted of making so many lol. My compressor/distortion is starting to look like a switch board from me adding tone circuits for each guitar at this point.

[ashcat_lt]

10 x the impedance?

This is a good question.  I know for a fact that 70K in-Z will cause significant high frequency attenuation (or reduction of the resonant peak)  when connected directly to a 7K passive pickup.  I'll attempt to explain this from my limited base of knowledge:

You pickup's 7K "impedance" isnt.  This is the DC resistance (R) of the very long, very thin wire in the coil.  This number ignores the inductance (L - considered to be in series with the resistance) and the inter-winding capacitance (C - in parallel with the resistance/inductance series structure).  Impedance is a complex (aparently involves imaginary numbers) combination of the interaction of these factors.  

Also keep in mind that most guitars have the resistance of at least one Volume and one Tone control in parallel with each other and the input to whatever device it is connected.

10x is the standard answer.  I think it works in typical active electronics because they tend to lack the L.

The point I was trying to make with my post was that impedance is not an issue in this case.  A line level (or lo-Z in general) source will not experience any negative effects from driving a higher Z input.  The Out-Z of the source does not significantly affect the corner frequency of the capacitors in the circuit.  The only concern is that the input V does not cause unwanted distortion.  For this, you may need a voltage divider.

[loss1234]

so let me ask...is the importance of the REAMP then mainly to make sure that what is coming OUT of the mixing board not TOO HIGH a level for the stompbbox? or is it to clear up impedance issues?

these are things I want to be clear on as I am trying to design things that i KNOW appeal to me when I put things like drums, synths, etc through them, but i want to make sure they also sound good with a guitar and amp

also....is the cap on the input ONLY to block DC or is it for filtering?

if i can use a high impedance device with a line input then i THINK  all i need to worry about is putting in a voltage divider for high level signals and a different cap depending on LINE or GTR...correct?

if I go with 1m impedance...I am a still fuzzy on this...this 1m would be on the input to the non-inverting amp OR would it be going to ground on the input of the non inverting amp

thanks thanks thanks

[amptramp]

I have assumed using a non-inverting amplifier because the input does not feed a virtual ground (that is, a point in the circuit held at ground by feedback).  Gain can be varied without affecting the input impedance.  The 1 megohm resistor goes from the non-inverting input of the op amp to ground.  Gain is determined by two other resistors, but the gain in a non-inverting stage cannot be less than 1 unless you have a voltage divider at the input (but for noise reasons, it would be better at the output, if necessary at all).

The input resistor in an inverting amplifier is the input impedance due to the virtual ground and may be forced to be quite low in order to get adequate gain without too much DC offset due to bias current.  There is a noise issue with an inverting amplifier - the higher the input resistor, the higher the noise.  This is not as much of an effect in a non-inverting amplifier because the input resistance is effectively in parallel with the source resistance, which should be much lower.  There is also resistance on the inverting input of the non-inverting amplifier, but it can be reduced to minimize noise and offset (at the expense of extra current drain).  The gain on either an inverting or non-inverting stage is set by the feedback resistance and this resistance is what should be changed if two different gains are required.  And if a switch is used, it should be connected at the output side of the resistor rather than the input to minimize input capacitance.

The cap on the input is to block DC, but it will have a filtering effect.  It is preferable to have the filtering after amplification since a filter right at the front end isolates the input from the noise reduction that is obtained by paralleling the output impedance of the instrument with the bias resistor at the amplifier input.

There are application notes in various places on the web that show how to design with op amps.  The key is to understand that you are juggling gain, bandwidth, slew rate, noise, offset, current drain and maybe a few other criteria in the design.  I used to design analog circuitry for military programs and you are expected to justify every decision at a design review, so I recommend getting into the habit of determining what you want the circuit to do before you rush into a design.