• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Antelope Ships Reference-Grade A/D-D/A Converter - with negative impedance

dkinric

Addicted to Fun and Learning
Joined
Nov 21, 2018
Messages
666
Likes
1,446
Location
Virginia, USA
As we have seen before, lots of promised performance on a $3k USD DAC, but this caught my eye:

"Two headphone outputs are provided, both with an output impedance that’s adjustable in 17 steps from —4.6 ohms to 85.3 ohms. Antelope calls the ability to generate a negative impedance is a “groundbreaking achievement [that] provides a headphone membrane weight compensation for a listening experience like never before.”

The ability to generate negative impedance? ...with atomic clock synchronization? To the test bench I say!
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,690
Likes
6,013
Location
Berlin, Germany
Negative output impedance is nothing special and easily established by a specific feedback network which applies partial positive feedback. It has been used with speaker drivers for decades (see eg http://www.freepatentsonline.com/4118600.html).
It can lead to lower distortion with a driver that has a very linear force factor ("BL") as it pushes the driver more into velocity-controlled mode. When Rout = -Rdc (that is, effective voice coil impedance=0) the driver is fully velocity-controlled (effective VC voltage is proportional to diaphragm velocity, that is, the velocity is impressed by the amp voltage) whereas when Rout=infinity (current drive) the driver operates in force-controlled mode. With an Rout between this extremes, for example Rout=0, the mode of operation is a blend in between. Both modes have benefits and drawbacks depending on the actual driber so there is no blanket advantage from going to the extremes. Frequency response changes of course and must be EQ'd out.

I would doubt it has much benefit with headphone drivers.

PS: Did I mention Amir efforts to rank HP amp output impedances is basically moot? Once output impedance (absolute value) is below 1/20th or so of the can's impedance it's effect is completely neglegible. On, say, a 50ohm'ish headphone +5Ohm or -5Ohm doesn't make a difference, maybe a hair of frequency response change, that's it.
 
Last edited:

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,370
Likes
234,401
Location
Seattle Area
On, say, a 50ohm'ish headphone +5Ohm or -5Ohm doesn't make a difference, maybe a hair of frequency response change, that's it.
Just recently I measured Woo Audio at 70 ohm from what I recall. I actually just took it out of the graph because it makes the rest too small. I also routinely measure above 5 ohm. Would be nice not to have to do this anymore but it is necessary as it is often not documented.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
The OP is referring to the Amari.
https://en.antelopeaudio.com/products/amari/

Antelope likes being cagey with specs. I don't think they use AES 17 (AES standards for measuring ADC and DAC devices) guidelines for them.

Like that impressive dynamic range of 138 db on the DAC. Uh, I'd say probably not true.

I've got an Antelope audio device which is a good device and performs well. But it lists I think 129 db dynamic range on the DAC. Well that only makes sense if you compare the noise floor at minimum gain vs maximum output at maximum gain. That is a useless number since you can't gain ride playback with the gain knob. It achieves a perfectly respectable 112 or 113 db (about 117 db A wtd) dynamic range on the AES17 version of the test.

Antelope (and some others) do make rubidium based atomic clocks for studio use. They output a 10 mhz signal to let multiple ADC/DAC in a studio all run off the same master clock. There are of course non-atomic based master clocks as well. While the clock is highly accurate and stable, and maybe they do have lower jitter, it doesn't improve the free running oscillators inside the audio gear. It simply maintains locked clocks across multiple devices.

https://www.soundonsound.com/techniques/does-your-studio-need-digital-master-clock

They tested a few clocks and external clocks do not reduce jitter except in extraordinary situations.
 

Ron Texas

Master Contributor
Joined
Jun 10, 2018
Messages
6,078
Likes
8,914
next up, negative mass and negative sound pressure...
 

bravomail

Addicted to Fun and Learning
Joined
Oct 19, 2018
Messages
817
Likes
461
Recently read a book on developing first radar rocket anti-air systems S-25. First design dealt with multiple quartzs and their sync. 2nd version had 1 quartz and delay lines. They saved 200 quartz gens on just 1 device, and god knows how much electronics.
 

KSTR

Major Contributor
Joined
Sep 6, 2018
Messages
2,690
Likes
6,013
Location
Berlin, Germany
Just recently I measured Woo Audio at 70 ohm from what I recall. I actually just took it out of the graph because it makes the rest too small. I also routinely measure above 5 ohm. Would be nice not to have to do this anymore but it is necessary as it is often not documented.
I didn't mean to dismiss your efforts, quite the contrary, as your mentioned example shows. And because phones have such a great span of impedances it really can matter.
It is more addressed to people that might be tempted to think that amp output impedance == zero is sort of ideal and say, 0.25Ohms is "twice as good" as 0.5ohm.
I would certainly agree low single digit output impedance must a achieved by any hifi amp that claims fidelity, no question.

Sidenote, I just designed a monitor phones output for a synthesizer keyboard with 50Ohms on purpose, where sound quality, in terms of varying frequency response vs output impedance is less important than a more consistant SPL output vs volume control setting regardless of phones impedance. It also helps simplifying the circuit, low noise and ruggedness more easily achieved ;-)
 

m8o

Senior Member
Forum Donor
Joined
Jan 28, 2019
Messages
348
Likes
224
Just recently I measured Woo Audio at 70 ohm from what I recall. I actually just took it out of the graph because it makes the rest too small. I also routinely measure above 5 ohm. Would be nice not to have to do this anymore but it is necessary as it is often not documented.
Log scale on the Y-axis, I do suggest. ;)
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
Negative impedance circuits are nothing new, as has been said, but using them in a headphone amp to create a negative output impedance probably is new. Can't really see the benefit in this application but obviously others do; if you have reactance in your 'phones (i.e. nearly all headphones I've ever seen) I'd be very careful the thing didn't oscillate with a negative-impedance output.

The are all sorts of negative impedance circuits and devices. I've designed a bunch of filters (though not recently) using GIC (generalized impedance converters) and gyrator circuits that include negative impedances in them. And once upon a time designed a tunnel-diode ADC for a 50 GS/s application (not audio, a bit above what most tweeters will handle, and probably above what most audiophiles would claim to be able to hear). The final of one of my long-ago grad classes included an op-amp at the transistor level with an external circuit wrapped around it. After figuring out a bunch of op-amp parameters for the circuit (remember this was at the transistor level, not just a block) we had to figure out the net transfer function of the overall circuit. Sure enough, it created a negative resistance. Prof thought it was clever; I (and most of my classmates) were ready to kill by the end of the final... Nobody figured -kR could be right. I turned in my paper showing -kR as I had seen similar circuits before and got it right, mainly by luck. I think 2 or 3 of us (out of twenty or so) got the right answer, and suspect many of the rest had the right answer but figured they must have made a mistake somewhere and ran out of time trying to find it.
 
Last edited:
  • Like
Reactions: m8o
Top Bottom