Super Simple theory on speaker cables

This seems perfectly plausible. If you're correct, it's likely that there are other subtle mechanisms that are also involved here.

Well explained Erik. I will keep that in mind and curb my skepticism. 

What was the configuration of the resistors (series, parallel, combo)?
What were the values?
Did you measure frequency response with & without?

Hey Rocknss
If you are truly interested in the details of the Focal, read here:

https://pqltd.blogspot.com/2016/01/focal-profile-918-ultimate-upgrade-guide.html

This is not my only data point however. I’ve also seen beginner DIY’ers alter the bass crossover by changing the second pole cap, which changed the ESR but only slightly, and found it sounded significantly different. Increasing the series resistor by maybe 0.5 Ohms fixed the issue.


This is not proof, but it's interesting. Like the doctor who first noticed smokers and lung cancer seemed related.


Best,

E

@Erik Interesting, thanks for sharing.
I’ll have to check my cables characteristics! Empirically, I find that slightly capacitive ICs seem to perform slightly better — by which I mean a better sense of clarity & air, i.e. clearer stronger upper mids (& frequencies over 13kHz).

Unless of course, in the case of amps, it comes down to slew arte?

I meant slew rate, of course

You're trying to describe some complicated things and seemingly understand them in simple terms. 

I'm not at all surprised that a Focal crossover is doing weird things to get the amplifier to push more or less power. That's how they mitigate aberrations in the performance of the drivers. My assumption finding something peculiar in a Focal crossover would be that it's trying to control a particular trait of the speaker. 

Damping factors tend to be a little more complicated than the number except in a very simple amplifier. Virtually all solid state amps degenerate the output devices with resistors. That tends to lower the damping factor. Feedback loops tend to fight against the degeneration by making the the voltage gain stage, itself often degenerated, compensate for the truncated gain characteristics of the output devices and raise the damping factor. The result is more complicated output impedance characteristics than a damping factor rating might imply. 

Since most amps have feedback, and that's where they get a significant amount of their damping factor, it's reasonable to assume anything presented as a load to the output will influence the momentary damping capability of the amp. However, wisely implemented feedback will do what feedback should do; simply control the load. 

Looking to the measured properties of a variety of cables seems to suggest that those building them have some understanding that adding impedance and capacitance does effects the sound. Typically they go for the higher capacitance to ameliorate inductance since an amplifier's damping capabilities are more robust at lower frequencies. 

That's my line of thinking at least. 

The midrange resistor on my Magnepan 20.7 seems to adjust the tone of the mids rather than the loudness. Adding the 1.0 ohm resistor (furnished) warms the mids and aligns them better fit to the bass panel. So the fact this is happening to the Magnepans just tends to confirm the op here theory a bit. 
Thought the treble resistor does seem to be to reduce the loudness. as well as change the tone. Just to mention.

@Kosst_amojan
I did a complete analysis of the speaker in XSim. Gathered acoustical and impedance plots. The only thing that circuit did was lower the impedance.

You can remove one cap and resistor section with nearly zero effect to the voltage transfer function. With a "normal" low pass filter, you can raise the impedance, with the identical transfer function, and eliminate the resistors.

I’m confident in my conclusion: The crossover was deliberately designed to have a low impedance point.

Best,
E

Hi @elizabeth,
Yeah, I put that in a different category. Several high end manufacturers have some sort of level setting which DOES show up in simulations as altering the output. I think those circuits are legit, whether dials or resistor changes.

What I'm talking about are how simulations miss how changes to the impedance curve alters the sound, even when the simulation says it does not.

This is kind of like how we assume tube amps are more sensitive than SS amps to hard to drive speakers.

We think, oh, it's a SS drive, what can 1 ohm change in impedance to? Actually, a lot.

Best,
E

@erik_squires 

I don't see what the problem is with that. That's very characteristic of Focal. That's how they get their unusually small woofers to create the bass performance that they do. 

@kosst_amojan


You are referring to using a lower driver impedance. When you do that you gain sensitivity at the cost of efficiency. The lower driver impedance causes more current to flow through the coil and act against the magnet.


That is unfortunately not what I was talking about in this particular crossover. This is an impedance miss-correction circuit which lowered impedance without gaining efficiency or sensitivity. There is no increase current flow in the driver. Only via the resistors. So it is wasted heat.


Often, impedance correction circuits are used to flatten out the impedance making them a more consistent load for tube amps. Keeping the impedance between 6 and 8 ohms for instance. That can really help tubes perform.


As I mentioned, I thoroughly analyzed the drivers individually for impedance and amplitude. I produced an alternative crossover which raised the impedance curve significantly without altering the frequency response, and improving efficiency.

Best,

E

erik_squires
What I'm talking about are how simulations miss how changes to the impedance curve alters the sound, even when the simulation says it does not.

You would need to include some model of the amplifier in your simulation to possibly see this.

Also, don't know on your specific example, there may be some rational to there design choice. Some issues that were displayed during bench testing. Possibly reducing variation over component tolerances, have you tried running a Monte Carlo analysis utilizing the component tolerances?