Question about on-resistance in CMOS switches, and guitar vs line signal

[tysonlt]

Hi!

I am looking to build a CMOS-based looper, nothing unusual.

I want to send guitar into a box, switch in a few loops, go out to amp, then have the fx loop come back in and switch a few more loops.

1) On-resistance... I am scared! What is ok here? Is 125 Ohm ok? What does resistance actually do to audio signal... make it quieter? (Pardon my ignorance This is the chip I am looking at: http://www.taydaelectronics.com/servlet/the-2115/CD4053-4053-CMOS-MULTIPLEX-fdsh-DEMULTIPLEXER/Detail. It's tex instruments... are they any good? It's so cheap!

2) Will I have any problems with guitar signal vs line signal? I want to put a buffer first... will that bring the guitar signal up to a level that it is close enough to the fx loop signal anyway? I am using fender hot rod deluxe, I am not sure what the signal is like.

3) What part of the guitar signal determines its volume and tone? ie, does more volts mean louder? What makes it lose highs?

Was thinking of asking who was the third gunman on the grassy knoll, but decided against it.

Thanks legends. I intensely approve of this forum and its members.

[Pablo1234]

Resistance Creates a Voltage Drop or also Called an IR Drop. It limits current.
125 Ohms is ok.

yes buffers are almost always a good Idea.

Volume is almost always looked at in Voltage(dB). But if the thing that's driving your load doesn't have the current to maintain the voltage, it will drop and volume will go down. Hence IR is Volume. IR drop is Current x Resistance. Buffers specifically do not normally change the Voltage. They provide current for the voltage to be maintained across low Reactance Loads http://en.wikipedia.org/wiki/Electrical_reactance.

Tone is specifically Reactance, its the change in Voltage over Different Frequencies. If you reduce the voltage at 600 hz by 6dB with a notch filter you change the tone.Reactance can be created in many ways but it is all just a form of Reactance. Examples of intended tone change are Tone Stacks, Graphic EQ's, Parametric EQ's. and then their is Time domain Effects that do the same thing like Phase shifters, Flangers, Reverb, Pitch Shifters, Ring Modulators(PLL), Chorus and so on. they all create peaks and dips in the frequency domain.

Unintended Tone changes are from Impedance Miss-matches. Guitar Cables have  Resistance, Inductance and Capacitance. The resistance magnifies the Reactance and the induction and capacitance set the frequency and direction of loss. this happens in every component in the entire system.

You could defiantly use this chip for what your talking about, but you need to do some research first.
1. Get a resistor like 10K and put it into the system you are trying to test, Measure the voltage across it to find out what the amp is pumping out for voltage. This will give you a starting point to creating what you want.

[PRR]

> On-resistance... I am scared! What is ok here? Is 125 Ohm ok? What does resistance actually do to audio signal... make it quieter?

Yes, in proportion to the ratio of switch and load resistance.

Round the switch to 100 ohms. Most "loads" in guitar-cable chains are 100K or more. 100K is 100,000 ohms. The actual computation is 100,000/(100+100,000). The output is 0.99999 of input, or 99.9999%, or -0.01db. "Less", but so close that you can not hear any difference.

(Contrast with low load impedances. Use 100 ohm switch to connect 10 ohm speaker. The output is 10/(100+10) or 0.0909 or 9% or -21db... a BIG drop. {Also, CMOS won't handle the voltage and current typically put in a speaker.})

The 4053 switch is good general-purpose audio switch.

> vs line signal

"Line" covers many things. Even buffered guitar can't be over 9 Volts peak-to-peak 3V RMS, and is usually much less. Mixing consoles may run over 10V RMS, though they should accept "line input" below 1V and boost as desired. Back when radio transmitters connected to studio on telephone wires, we sometimes ran many Volts signal (though on shared cable the levels were limited to a few volts).

If you are working on the "Pro interface" side of a mixer, the 4053 will distort or die on very strong signals. CMOS switching is done internally at lower level then boosted to interface level.

Guitar and hi-fi "line" is usually fine in CMOS, especially at +/-5V or +/-7V CMOS rails.

> who was the third gunman on the grassy knoll

His name is Bob and he's now a retired banker in Wichita.

[Gurner]

Was thinking of asking who was the third gunman on the grassy knoll, but decided against it.

I have to say that the only time I've ever heard the phrase grassy knoll, is wrt JFK - does America invent specific unique location terminolgy for dead presidents or do you lot use it all the time eg ...

Q: "Hey mate, where's the nearest bar, I'm gagging for a beer?",

A: "Walk down two blocks, you'll come to a grassy knoll, it's then just to your left")

[tysonlt]

http://mcadams.posc.mu.edu/gk_name.htm

Looks like it was a special case, and people are fighting over who said it first:

"...Careful study of the tape shows him forming his lips to make the "mmm" sound, not "nnn." "

Wow. And I chose to become a history teacher

Aaaaaanyway

So, if I understand:

1) 100 Ohm is nothing, so just relax, and;

2) Buffered signal from pickups and the signal coming out of an amp's fx loop will both be fine in CMOS, so refer solution to 1).

I have a 12V adapter and was thinking of running the CMOS at a different voltage to the PIC. The pic of course will get a 7805 regulator, I was just going to use another regulator (12V->12V) to send to the CMOS. I believe this will give it more headroom to prevent distortion?

If it is true that the chips won't care if it is buffered guitar or line level (as many pedals don't seem to care, I suppose), then all I have to do is tell the switcher which loops are pre-amp, and which are effects loop. That seems like a simple switching puzzle that even moi could work out.

Thanks all, and please bark at me if I have the wrong assumptions.

[PRR]

> does America invent specific unique location terminolgy for dead presidents

Nobody remembers where Garfield or McKinley were shot. (Train station, Pan-Am Expo)

Lincoln was at Ford's Theater, long before some other Ford was making cars.

Garfield, McKinley, Lincoln, and Kennedy actually died elsewhere. (JFK was probably dead on the spot, but they went to the hospital to be sure.)

FDR was at Warm Springs when he got his last headache.

Many others from Washington to Nixon were at home or a nearby medical facility.

I can not think of any other dead-president place-name after Ford's and the Knoll; and a stupid joke about Grant's Tomb.

"Grassy Knoll" may be a Texas Special. Much of Texas is bare and flat. Grassy knoll is something special. It is also a handy place-name in a fairly dull public plaza.

I was in school at the time JFK was shot. By the time I got to a TV, it was all "book depository". I don't recall mention of Grassy Knoll or Mall until later, when doubters were combing the first reports.