I thought your DAC architecture had an essentially constant draw on the power supply, independent of the frequency or phase of the audio signal, because you’re only switching the direction of flow through the audio transformer. Constant voltage, constant current, just a different polarity downstream of the opamps. Is that incorrect? If it’s correct, how could any audio frequency characteristics be seen way back at the fuse in order to interact with the non-linear resistance?
Whether or not non-linearity is key to the effect it’s also not clear to me how the directionality of a fuse could be relevant, because it’s in the AC part of the circuit and swapping the direction of the fuse would just shift any distortion by 180° of your 50/60Hz mains supply then get rectified… seems it would be indistinguishable from the other fuse orientation. (EM emissions, on the other hand, can have asymmetric radiation patterns.)
Regardless, every time I hear about some analogue component that can impact the sound like this I want to figure out how to avoid using it. This is why I refuse to use vacuum tubes anywhere in my system – I hate the variability of it all. So how come we don’t have 1A circuit breakers instead of non-linear-resistance fuses? Would that be a sensible mod to consider, or do those little electromagnets cause more harm than they’re worth?
Just did some digging on X/Y caps. My draft idea for the MKII:
place two X2 polypropylene caps (Vishay series BFC23384 with values 0.1uF and 0.68uF) line-to-line in series; to be placed right after the IEC connector
place two Y2 ceramic caps (Vishay series VY2 with the values 0.0010uF and 0.0047uF) in series line-to-ground; to be placed right after the X2 caps
I have learned that Y2 capacitance values should be really small, agree with this idea?
Ted I have brought this up in my past posts on the beta forums, I’m glad to see you talking about this. Is this not a 60/50hz fuse? Voltage drop across that fuse would be an issue. If the voltage drop during the most demanding periods is 10x lower than the demand it’s most likely ok, right. I’m trying to wrap my head around a fuse making that much of a difference.
The voltage drops probably don’t matter a lot for low current devices. If the fuse acts in a linear manner, then the result is that the AC voltage wave shape is the same, but a little lower. If it’s non-linear, then it might interact with other things downstream.
With constant current draw any diode like non-linear effects of a fuse will still affect the two polarities of the AC differently leading to a 60Hz ripple downstream. Even so if the only non-linear effects of a fuse were at 60Hz/120Hz then I don’t think they would make a lot of difference downstream in the Mk II.
Still my overall impression is that nothing is perfect and that people’s hearing is good enough that those non-perfections leave holes for tweaking. As your system gets cleaner, blacker, … the more these very small effects become audible.
So… run two half-sized fuses in parallel and in opposite orientation?
No comment re circuit breaker as an alternative?
If I had the gear I’d do some analysis of the AC power circuit between a P15 or P20 and a DS using different fuses and different polarity, with a variety of audio test content including silence. Probably also replace the fuse with straight wire and analyse that as a control. I’m very curious about the nature of the problem.
But I’m just musing. I so love the way the DS and esp MkII architecture side-steps or neutralises issues like input signal jitter, and non-linearity caused by variable demand on the analog PSU (by making the demand constant). The idea that fuses – and their polarity – might matter is jarring.
Sorry. When I was looking at them it was a question of reliability while in a fault condition. Fuses are a lot safer than anything that needs current to work especially at 120VAC (or more). In the past I’ve looked at circuit breakers multiple times, but other solutions always seemed better.
In some sense that seems more probable than power cords making a difference but the differences in power cords have always been more apparent to me than the differences when changing of any other cables in a system.
Fuses are designed as current restrictors and (ignoring them) much of the time simply having a more direct connection to power seems to be better.
OK. My view on power cables is that the only significant difference between them for low current devices is shielding. In a system, better shielding for a given component’s AC cord might help by keeping EM radiation out, keeping EM radiation in, or a mixture of both.
Oh yes, that’s the strange thing. I just once tried with a solid piece temporarily inserted in a fuse holder, which is suboptimal, but others worked harder. I guess it should not be thicker than a normal wire.
But–if I may–“brainstorming” with a brain, smaller than the brain what resides in the skull of our hero, Ted, the Tedster, Tedaroonie, DigitalTed will just lead to no good, in my small-brained opinion.
But hey, that’s just me.
