As someone truly excited to experiment with removing any preamp/processing from my audio pipeline, I was discouraged to see the data plotted on the Roll-Off graphs posted at http://www.tortugaaudio.com/what-is-a-passive-preamplifier/; considering the length of the cable runs to my mains (35’).
I don’t currently have a passive pre, but my source has digital volume control in its dac’s, so I was experimenting with a source direct set up (Oppo 105 --> Pas Labs XA200.5 mono’s).
Being that I am not an engineer, can anyone enlighten me with regards to whether I am likely to encounter roll-off in the audible range, considering the spec of my current system?
What I’m running:
Source peak output level: 2.1Vrms
Source output impedance: 100Ω
Balanced XLR Length: 35 ft.
Balanced XLR capacitance conductor to conductor: 39.372(pFm) Note: Meter, not ft.
Balanced XLR capacitance Cond. to other conductor and shield : 85.306(pFm)
Power Amplifiers input impedance: 30,000 Ω
Any other pointers that might help guide me in weigh the benefit/deficit of losing the pre/pro?
On an aside, can anyone share the formula used to plot the graphs I referenced?
When you distill is all down, most passives represent a variable resistor between your source and your amp. Add in the capacitance of the downstream cable etc. and you have a classic RC low pass filter. "low pass" means the lower frequency stuff gets "passed" through and the higher frequency stuff gets filtered out. Depending on the quotient of R*C, it's possible to roll-off some of the signal below 20khz not to mention the phase shift that comes along with this.
The equation to calculate the roll-off frequency where the signal is down 3dB due to this effect is f = 2*pi*R*C; where R is in ohms, C is in farads, and pi = 3.1416.
Let's take your 35 foot cable. At 85 pf/meter that's roughly 28.4 pf/ft (i divided by 3, yes I know that's not exact, roll with me here, I didn't want to look it up). Not the best but not the worst by way of cable capacitance. 35*28.4 = 9.95E-10 (a very small number indeed).
Now let's look at the passive. Say it's running around 2k ohm at normal listening levels. Run the calc and you get a roll-off freq of ~80kHz. Way above 20kHZ. No worries.
Now let's turn down the volume a fair amount and the series resistance of the passive goes up a lot. Let's say it goes to 50k. Run the calc again and you're at 3.2kHz. Woah! Where did all the highs go? Flat sound indeed.
Next let's cut your loooonnnnng cable down to 6 feet. Rerun the calc and now f = 19.6kHZ. Much better and you probably won't even notice it. Cut it in half again - 3 feet - and presto! f = 37.3 kHZ. No worries again. Or alternatively, rather than shortening your cable, the same effect can be achieved by throwing out your cable and buying a better one with half the capacitance. If it were me, I'd just cut the cable but that's me.
In my view, realistically, you're not going to spend much time listening at attenuation levels where the R of the passive is high enough to cause a problem.
I spent quite a bit of time recently with a reviewer who was working with our LDR1. His cables were 20+ feet in length and I cautioned him on this issue. After extensive critical listening, he was convinced that there was no discernible roll-off of the highs notwithstanding his long cables. I think that's because at the attenuation levels his was at, the R of the passive was too low to matter.