Can someone help me out with signal voltage?

[DIY_Guy]

So I recently discovered the joy of Spice modeling, and I've been using it to help me learn more about how preamps function. One thing that I realized is that I know nothing about signal voltage.

Messing around with some different preamp designs brought up a few things I'd like to know.

1. What max voltage should a LTP phase inverter be seeing at its input? In other words what max voltage should the preamp be putting out with a standard master volume at full clockwise rotation?

2. What max voltage should my fx send be showing? I think it should be line level(1.7v max right?)

This interests me because I can't find anything explaining it in any of my books. I went ahead and loaded up some known working preamps in LTspice. When I send a 1.5v sine wave through them, the voltages I see before the PI are not what I'd expect them to be. So any help on this would be appreciated.

[ThermionicScott]

Have you read "The Fender Bassman 5F6-A" by Richard Kuehnel? He goes through a neat analysis of the headroom (maximum input signal) of each stage in the amplifier.

In the case of the Bassman, it was +/- 2.61 V.

- Scott

[MWJB]

It will vary from amp to amp & depend on the signal voltage that you are applying, position of pots etc...at Fender.com they have schematics with AC signal voltages. The 5F6A shows 4.2V peak to peak @ 56W RMS at the PI grid, with a 3.85mV signal at normal #1, all controls fully CW.

I'd start at the output, ensure you are getting proper rated RMS voltage, then back track through the amp, noting AC voltages.

To build a meaningful comparison, pick a method & stick to it.

[d95err]

1. What max voltage should a LTP phase inverter be seeing at its input? In other words what max voltage should the preamp be putting out with a standard master volume at full clockwise rotation?

2. What max voltage should my fx send be showing? I think it should be line level(1.7v max right?)

What voltage the PI should be seeing depends on what you want the output stage to do. If you want it to run clean, then you need to find the maximum signal level before clipping. If you want to crank it and get poweramp distortion, then fire away with as much amplitude as you want!

The FX send level depends on what kind of equipment you want to use in the loop. +4 dBu (1.7V peak) is typical rack-gear level, while consumer audio products (and stompboxes) typically use -10 dBU (0.45V peak).

Here's some info on line levels:
Line level - Wikipedia, the free encyclopedia

[DIY_Guy]

Alright thatspretty much what I thought. So if I'm going to throw a channel volume(or fx send control for that matter), I'd want it to be able to adjust from -10 to +4 to accomodate all kinds of gear?

[BackwardsBoB]

You can tell that something's bogus when the "nominal" level is given to four digits of precision. The controlling specs are for "nominal". What, you may ask, is "nominal"? Nobody knows for sure. It's a nebulous concept, sort of like "average", but less precise, that bears no correlation to the real world.

Am I being cute? Not really. The Red Hot Chili Peppers mixed one of their CDs with so much compression (to boost the nominal level and give it more presence, i.e. make it louder) that when radio stations played the same trick, the listener was treated to false harmonics of false harmonics, and even the RHCP didn't sell. Stompboxes "typically use -10 dBU (0.45V peak)" right? You just got it from d95err and Wikipedia. but you can easily get 2V out of the attack with passive humbuckers.

So, yeah, there are "nominal" numbers, but all you can do if you're trying to pass a clean analog signal is make sure that any optimization is at the nominal level, and support lots and lots of headroom.

Those of you with amp "modeling" know what I mean. You have to turn your levels way down to keep from clipping on chord strikes compared to the sustain portion of single notes.

[Steve Conner]

You can tell that something's bogus when the "nominal" level is given to four digits of precision.

In Windows NT, XP, and so on, timestamps on files are in the form of a 64-bit number that represents the number of 100-nanosecond units since January 1st, 1601. (Or the year 0 After Gates as it's also known.)

MSDN also quotes an accuracy of +/- 1 hour for the timestamps, which is a great example of spurious precision.

Re the Chili Peppers thing, if you compress a turd, it's still a turd.