[Rowethren]

If your speaker is a full-range, single driver.   Add multiple drivers in and crossover networks and things get more interesting.

I don't think that matters too much otherwise multi BA IEMs would work differently on headphone amps compared to single driver IEMs for example. All the amp sees is the load it is driving not how it is allocated to drivers through the crossover so as long as the amp has enough power to work the loads within set tolerances I don't see why it would make a difference.

 

[Ableza]

  Meh, who cares about the topology?  Care about how well it works.
 
 
 
If your speaker is a full-range, single driver.   Add multiple drivers in and crossover networks and things get more interesting.

For amplifier design?  No, not really.

 

[Fr8dog]

I realize this is from a while back, but I just started reading this a few days ago.
 
It appears you may have solved my biggest complaint with the addition of the new preamps. I'll get to try different types with just one. I usually try one and then sell it to buy a different one. Need the cash to upgrade/change and also don't like having equipment just sit. I still have the SYS. Not using it, but I'm almost embarrased to tell people what preamps that put to shame.
 
Dan

 

[FrivolsListener]

 
For amplifier design?  No, not really.

 
You don't get any weird reactive loads with "exotic" speaker topologies?  (Serious question.)  That's good, then.

 

[Ableza]

   
You don't get any weird reactive loads with "exotic" speaker topologies?  (Serious question.)  That's good, then.

Weird reactive loads?  All loudspeakers are weird reactive loads since all audio signals are AC and loudspeakers work via Faraday's law.  Take a look at the typical impedance curve of any loudspeaker, single driver or multiple element.  As long as the loading is sufficient to prevent output stage overload (loudspeaker offers high enough net impedance) then the amplifier really doesn't care what it's connected to, it does what it does. 

 

[FrivolsListener]

  Weird reactive loads?  All loudspeakers are weird reactive loads since all audio signals are AC and loudspeakers work via Faraday's law.

 
Of course.  I was just surprised that various ribbon-whatevers and 4 driven elements with a 4 way crossover with its inductors and caps doesn't get even more weird.

 

[Ableza]

   
Of course.  I was just surprised that various ribbon-whatevers and 4 driven elements with a 4 way crossover with its inductors and caps doesn't get even more weird.

​It's all quite predictable and easily modeled.  And by the way ribbon drivers are really just esoteric dynamic drivers with very similar impedance curves. 
 
Amplifier designers really don't change how an amp is built based on what kind of load they think will be placed on it.  They are much more universal than that,- designers just test into standard loads and report the results, and recommend certain minimum net loads for best performance.

 

[Armaegis]

  Weird reactive loads?  All loudspeakers are weird reactive loads since all audio signals are AC and loudspeakers work via Faraday's law.  Take a look at the typical impedance curve of any loudspeaker, single driver or multiple element.  As long as the loading is sufficient to prevent output stage overload (loudspeaker offers high enough net impedance) then the amplifier really doesn't care what it's connected to, it does what it does. 

 
Or we can design an outboard crossover and actively bi/tri/whatever-amp a speaker and keep all the drivers in the happy non-wonky zones. Give all the amplifiers easy loads that don't have to deal with impedance spikes and current surges and assorted voodoo. Of course, then you have to monkey with crossovers, but that's another story.

 

[US Blues]

  Just curious: one selling point of Jotunheim is the entirely new Pivot Point topology. Is Pivot Point suitable for 2-channel? No knock on Vidar: anecdotal reports of Vidar + Freya from RMAF are positive, and Vidar sounds like a really robust amp. But exotic topology is exotic!

I asked Jason about this at RMAF, and he said that Vidar is an adaptation or modification of the Pivot Point design, and I'll let him say more about the topology when the time is right.

 

[FrivolsListener]

   
Or we can design an outboard crossover and actively bi/tri/whatever-amp a speaker and keep all the drivers in the happy non-wonky zones. Give all the amplifiers easy loads that don't have to deal with impedance spikes and current surges and assorted voodoo. Of course, then you have to monkey with crossovers, but that's another story.

 
 
That was kinda where I was interested in.  Not where things behave according to the math, but the edge cases.

 

[Ableza]

   
 
That was kinda where I was interested in.  Not where things behave according to the math, but the edge cases.

Not trying to be obstinate, but "things" always behave according to the math.  They have no choice: it's physics.    Can you describe the sort of situation you are talking about and why you think it's different?

 

[FrivolsListener]

  Not trying to be obstinate, but "things" always behave according to the math.  They have no choice: it's physics.    Can you describe the sort of situation you are talking about and why you think it's different?

 
 
Models do not account for all physical behavior, unless they have a lot of money behind some models.
 
What happens if the amplifier oscillates at 40 kHz for a half-second?
 
What happens at the resonant frequency of the driven element(s)
 
Even though it's low-Q, there's impedance peaks and valleys in speaker systems.  What happens to the amp when the impedance of the speaker system at 10 Hz is closer to two ohms than to four?

 

[JoeKickass]

  Meh, who cares about the topology?

 
I'm glad Schiit has a purist philosophy, and strives for no-compromises in the pursuit of quality audio.
 
There's a reason why a lot of the best audio gear uses vacuum tubes even though they aren't the most modern and efficient design.
 
Same thing for class A and AB, no matter how good class D gets it won't be the same:
 

 

[Armaegis]

While I love class-A goodness, as a guy who sets up PA's for people in ratty old venues with no elevators, hauling up several flight cases with a 100+ lbs of rack gear is a nightmare. I switched out some to cheap class D amps and yeah it doesn't sound as nice, but the entire case weighs less than 20 lbs and my poor back get to live another day. I could even cram them into a smaller case because I no longer needed to leave gaps for ventilation. For similar reasons that's why I have Behringer subs for gigs. They didn't sound as nice as the JBL's that I auditioned at the same time, but the key deciding factor was that the Behringers fit neatly in my car.
 
That all said, class D isn't necessarily terrible. Old/cheap class D is meh, newer stuff is much better (but also heavier as they beef up all the supporting circuitry) and quite frankly most people would be very hard pressed to tell the difference as long as you're not comparing crap. Every tech/topology has it's place.

 

[Ableza]

   
 
Models do not account for all physical behavior, unless they have a lot of money behind some models.
 
What happens if the amplifier oscillates at 40 kHz for a half-second?
 
What happens at the resonant frequency of the driven element(s)
 
Even though it's low-Q, there's impedance peaks and valleys in speaker systems.  What happens to the amp when the impedance of the speaker system at 10 Hz is closer to two ohms than to four?

OK,I'll play.  In the interest of full disclosure I'm an electrical engineer and I've designed loudspeaker passive crossovers (professionally in a former life) for more than 40 years:
 
1) Even simple loudspeaker models account for all physical behavior that matters.  It's electricity and acoustics and physics, it's not magic.
2) No audio amplifier should EVER go into oscillation, unless it is defective or there is an error in the design.  This has absolutely noting to do with the loudspeaker load.  Unless something has broken or you have designed an oscillator, an audio amp will never oscillate.  Now if you are talking about parasitic oscillation because of a high capacitance output in an amplifier, then what will happen is distortion and excess heat.  But again, a properly designed amplifier should never do this.  And at 40KHz you will never hear it.
3)  At Fs (resonant frequency) the driver can move most freely, meaning it is easy to push to Xmax.  Electrically the impedance of the driver spikes and efficiency drops very low, so the signal strength at fs is low.  This helps to protect the driver from over excursion.  A properly designed Xover will prevent drivers from operating at or near Fs; except for the woofer, which can experience fs although the impact is generally limited by a combination of system design (Fc > Fs) and designed high-excursion capabilities.  Some drivers (specifically ELF subwoofer systems like those from Bag End or Sun fire) operate exclusively below Fs, which means they must have high excursion capabilities and they consume large amounts of power to achieve usable output levels.
4)  Again, a properly designed Xover will minimize the effects of the natural impedance swings across frequency of the overall system, no matter what might be happening in any specific driver.  Unless you are listening to large pipe organ music or some EDM, you will never experience a signal at 10Hz, but if you do and if a loudspeaker system hit 2-ohms at 10Hz (unlikely) than the only thing that will happen is this: if the amplifier can handle the load, the power output will increase over what it was putting out at higher impedances.  If the amplifier cannot handle the load, it will go into protection.