when presented with a load the total output can not exceed the calibrated voltage on the next stage
Umm... If you mean "can not" literally, you're wrong. If you mean it
can but then it'll be clipped by the downstream gain stage, then yes.
While most high output pickups average from 90-110mV when plugged into a guitar amp, this can change greatly depending on the load that it is represented with. Under a lighter load, the pickup is able to produce a higher voltage and the same goes for any active gain stage's output depending on the input impedance of the next stage. Considering that the output impedance of the average guitar pickup ranges from 8k-16k and that a guitar amp averages 1M, its easy enough to say that loading a guitar pickup into a 1K load typical of mic pres and A/D converters (lower than the pickup which is very bad), you will easily have peak voltages that are beyond what 1.74v.
Actually, an active gain stage with very low output impedance will function as a near-perfect voltage source. The load impedance affects the amount of current drawn (I=V/R), but the voltage appearing at the output remains the same -- unless the current demanded by the load exceeds the ability of the output device to supply. The line out from a DAW will be like this, and should be able to drive a 10k load according to the standards.
OTOH, a very high-impedance output resembles a current source, in which case more voltage will develop across the load as its impedance increases, in accordance with Ohm's Law V=IR.
The lower the load impedance, the lower the voltage that can be developed across it for a given current, and the more current it will require to achieve a given voltage.
Loaded with only 1k, the voltage output from a passive pup will be lower than if it were loaded with 1Meg.
The other problem with mis-loading a
passive pup is that the reactances in the coil form a tuned circuit, which is not independent of the load resistance. The resonant frequency drops, and the Q factor drops, as the load impedance is reduced.
Normally the biggest influence on this is the pots in the guitar. 500k pots (à la Gibson) give hotter output with more edge on the treble than 250k pots (à la Fender). But if you switch from a stomp/amp with like 2M2 input to one with 470k, a bit of the tonal difference you'd hear comes from retuning the pup resonance. Drop the load to 10k line-in or a 2k mic-pre and you've pretty much buttfucked the tone before you even consider the dramatic increase in noise.
Active pups don't have this effect because the preamp in the pup buffers the coil from the load. If EMG hadn't put the series 10k on the output of the EMG-81 you could plug it directly into a 10k unbalanced line input without needing a DI.
You'd get reduced battery life due to the greater current draw, that's all.
What makes you think that a transistor (or vacuum tube) buffer isn't clean ... There is nothing about a transistor that will alter the sound nor the circuit if it was designed correctly.
All additional circuitry will increase noise. This includes passive components like transformers, and active devices also.
Both passive and active components alos increase distortion. Active devices create harmonic distortion, transformers have hysteresis effects as well as odd-order distortions created in the core.
And an active gain stage will require a power supply, which can be fun and games in itself.
All I'm saying: if you can do it without a transformer, great. If you need a transformer for hum/noise rejection, fine. But provided your line out can swing as much voltage as the guitar would output, you don't need a gain stage, and if you don't need it it's better not to bother.
If you used a 1:1 tranny in a passive DI the only thing you would be doing is matching an unbalanced to balanced signal.
Yeah, I recommended a 1:1 tranny for the reamp box, not the DI. A 1:1 DI for guitar would suck, only useful for synths, drum machines etc.
I once made the connection that matched impedances weren't indeed matched, but purposely "mismatched", and I had someone else jump my ass for it.
If he was asserting that matched connections are appropriate for line-level audio signals I hope you jumped his ass right back.
(In a non-gay way, I mean.)
In electrical-engineer-speak "matched" impedances specifically means the input impedance is the same as the output impedance. This is not used for line-level audio, but is used for data transmission.
Usually the transmission line connecting the output and input has the same matched impedance also, to prevent ringing and oscillations. That's why RCA cables for digital signals have a specified impedance, and why normal RCA audio cables are not recommended for S/PDIF connections.
...the general practice at least ten times greater, if not one hundred times greater.
In electrical-engineer-speak this is called a bridged connection.
Bridged connections give greatest transfer of voltage, so are used for audio line-level signals.
Matched connections give greatest transfer of power, not usually required for audio except in the output of tube power amplifiers. On the subject of which...
Its safe to load more than not enough...
For line-outs and low-Z power outputs, this is correct. Not so for tube poweramps.
I know that's not what you were refeerring to, I'm not contradicting you here, just digressing.
Check out the General Electric datasheet for 6L6GC:
http://www.mif.pg.gda.pl/homepages/frank/sheets/093/6/6L6GC.pdf
Scroll down to the examples for "Push-pull Class AB1 Amplifier values for two tubes". See the first two examples have the same voltage on the plates (360V) and screens (270V) and the same bias voltage (-22.5 V). The only difference is in the load impedance.
Dropping the load from 6k6 to 3k8 results in a reduction of power by nearly a third. Maximum screen current also drops by a similar degree, and plate current is increased by only about 6 percent. The power tubes are a near-constant current source.
Loading a tube amp with greater-than spec impedance actually increases the danger to the amplifier compared to connecting lower-than-spec load impedance.
This is the opposite of a solid-state hi-fi amplifier or PA amp. Which is drifting OT, but kinda worth mentioning...