JDS Atom Amp 2 Headphone Amplifier Review

[Robbo99999]

I love the look of this headphone amp and it measures great whilst being made in the USA which is really quite a feat for the low price! I have just one concern with the measurements and that's the channel balance - Amir said it was perfect for channel balance but there's a significant blip in the channel balance of over 2.5dB difference between channels which occurs between 12-13sec on this graph:

So surely that's not great channel balance performance and particularly not if you were to use it between the 12-13sec point which equates to the -40 to -75dB position on the volume knob. Am I missing something here?

I mean for comparison the original JDS Labs Atom Amp from year 2018 had better channel balance:

 

[Hammeredklavier]

Um, listen to headphones through it perhaps
sorry that was too easy, no harm meant at all.

Indeed, but all my DACs have got high quality headphone amps built in.

 

[restorer-john]

Can you clarify a few things for for me?

You mention the 600R load, but dashboard view is displaying the 200k balanced input on the the AP's dashboard. So I therefore assume you are using the Neurochrome load box for the 600R load.

But the 2V out point on your Neurochrome load box sweep shows a much higher distortion at the same corresponding 600R load point.

In the past, you used the 600R internal load in the AP and it displayed as such on the dashboard view.

JDS Atom Amp 2 Measurements

Let's feed the amp 2 volts and measure 2 volts out into 600 ohm load:

2V out point is about -113dB THD+N, which is around 0.0002%, 6dB worse than the number above.

So how are you leaderboard ranking the headphone amplifiers exactly? On the basis of an unloaded output, the unloaded line output, or a fixed 600R load on the H/P jack?

 

[gvl]

Well, well, well... another must-have device!

A good rule to follow with these things is for one to buy a must have-have device he has to sell a must have-device

 

[Sokel]

Can you clarify a few things for for me?

You mention the 600R load, but dashboard view is displaying the 200k balanced input on the the AP's dashboard. So I therefore assume you are using the Neurochrome load box for the 600R load.

But the 2V out point on your Neurochrome load box sweep shows a much higher distortion at the same corresponding 600R load point.

In the past, you used the 600R internal load in the AP and it displayed as such on the dashboard view.

JDS Atom Amp 2 Measurements


2V out point is about -113dB THD+N, which is around 0.0002%, 6dB worse than the number above.


So how are you leaderboard ranking the headphone amplifiers exactly? On the basis of an unloaded output, the unloaded line output, or a fixed 600R load on the H/P jack?

The two measurements have different gain I think,one says Low and the next high.

 

[Saponetto]

A good rule to follow with these things is for one to buy a must have-have device he has to sell a must have-device

I know, I should embrace this good discipline but it's too heavy for my hoarder-torn shoulders...

 

[staticV3]

The two measurements have different gain I think,one says Low and the next high.

shouldn't make a difference according to JDS:

source

It could just be different FFT settings for the sweep to accelerate data acquisition since you need a lot of points.

 

[Sokel]

shouldn't make a difference according to JDS:
View attachment 328899
source

It could just be different FFT settings for the sweep to accelerate data acquisition since you need a lot of points.

That would be perfect,in other implementations we often see some penalty,@amirm could measure it that way too to confirm their data sheet.

 

[BDWoody]

This looks great. Wasn’t there a concern with the original that there was limited protection for your headphones if something went wrong? (I can’t remember the specifics well enough) Is that resolved in this model?

From the blog post on their site, it looks like they put some thought into it.

"Failsafe DC Protection
With great power comes great responsibility: DC protection is standard in Atom Amp 2.

This is no off-the-shelf protection circuit. Common protection circuits are built for speakers, for which a volt may be considered safe. Protecting headphones requires detecting DC at 1/20th of this level without accidentally tripping during normal listening. Not exactly trivial.

Atom Amp 2’s protection circuit began in simulations. Real world prototypes were more sensitive than predicted, tripping at moderate volumes. Weeks of iterations later, the final sensing circuit relies on a precision rectifier built with a zero-drift opamp. Muting disengages only for safe audio output, and never accidentally mutes. The circuit has already protected us during development, muting output for minor assembly errors."
https://blog.jdslabs.com/2023/11/in...=Failsafe DC Protection,minor assembly errors.

 

[AnalogSteph]

Typical of other JDS products, it comes with the AC transformer which gets a bit warm.

No kidding. I've clocked the 230 V version that came with my Atom Amp+ at 3 W (real power) doing absolutely nothing. This doubles to about 6 W with the amp turned on. Suffice it to say, an off-mode power consumption that high is likely in violation of EU ecodesign regulations. It's also annoying because the 2012 Dell laptop I use with it with its 130 W power supply needs a lot less in standby and off (not a lot more than 1 W / 0.3 W if memory serves).

I suppose that even finding a decent AC/AC wall wart to ship with your product in this day and age is not easy, but still, transformers do not have to be this inefficient. Even if high-efficiency transformer design seems to be a bit of a black art that's not necessarily in high demand and the budget when shipping with a $129 product is likely to be quite limited.

It's probably about time to modernize the power setup (i.e. DC SMPS + inverter)... that still leaves the question of how to deal with the pesky mains leakage though. I mean, it shouldn't be a biggie if you have access to a grounded outlet so mains filtering can dump its leakage current there, but in some regions of the world that's not exactly a given (like SE Asia, but not even within Europe things may vary a fair bit).

 

[Jimster480]

Those b@$+@rd$! What were they thinking?!?

Atom was such a good value / performer but the cheesy feel was kind of a turn-off. Good to see a budget amp crashing the SOTA party. For a company that was supposedly “getting out of the SINAD race”, they sure delivered on that. Wonder if this change of heart will carry over to the next Element?

FWIW the company told me in response to a query about the Atom 1 and the CRINACLE Zero Blue ( a 10 ohm load) that the amp would handle such a low impedance load with no problems, which seemed to be borne out in use. No reason to expect any less here.

Well this version is definitely better despite not being much more expensive or the same price.

 

[NwAvGuy]

Very close to the “Holy Grail.”

 

[Joramun]

Which DAC will be a perfect pairing for this?

 

[staticV3]

Which DAC will be a perfect pairing for this?

Naturally, the JDS Atom DAC 2

 

[amirm]

You mention the 600R load, but dashboard view is displaying the 200k balanced input on the the AP's dashboard. So I therefore assume you are using the Neurochrome load box for the 600R load.

That's correct.

 

[amirm]

2V out point is about -113dB THD+N, which is around 0.0002%, 6dB worse than the number above.

That graph is made in high gain so that people can see the max power they can get. That costs you the 6 dB in noise relative to low gain used in the dashboard.

So how are you leaderboard ranking the headphone amplifiers exactly? On the basis of an unloaded output, the unloaded line output, or a fixed 600R load on the H/P jack?

The ranking comes from the dashboard which is 600 ohm on loadbox, but represented at 200K to AP.

 

[restorer-john]

@amirm Thanks for clarifying.

 

[amirm]

shouldn't make a difference according to JDS:
View attachment 328899
source

It could just be different FFT settings for the sweep to accelerate data acquisition since you need a lot of points.

You can see the difference in my power sweeps:

They may be referring to SINAD at max outputs which equalizes as you see above. But for the range that low gain is functional, it has lower noise floor than high gain.

 

[oscar_dziki]

Good product, but if one is to buy it today it would cost $153.44 with shipping and taxes compared to about $105.32 (again with shipping and taxes) for the L30II from aliexpress with current Black Friday discounts and coupons. With this, in my opinion, massive price difference makes the L30II the better buy.

L30 II is simply better.

 

[oleg87]

There’s more to a product than inaudible increments of performance.