• Welcome to ASR. 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!

Topping DX9 AKM single ended vs L30 II power

Tatalor51

Member
Joined
Nov 20, 2019
Messages
49
Likes
23
Hello,

Just bought the DX9 AKM (mostly for the looks, yes :) ) and, on single ended, it seemed a bit quiet for my taste on low volume masters like Steely Dan and after applying some negative pre-amp for EQ. I didn't have this issue with L30 II, my old amp.

This made me curious and did a fairly unscientific experiment which confirmed what I was hearing. Basically, I plugged the L30 II in DX9 RCA outs, set high gain max volume for both. Then I measured average SPL with a phone db meter over a few one minute fragments from some songs. During measuring the phone was unmoved and right next to a headphone cup (which was also not moved, just plugged in alternatively in the L30 II and DX9 6.35 outs. Yes, I am aware the absolute SPL measured this way is worth nothing but what I was interested in was the average SPL difference between L30 II and DX9. And this one turned out to be remarkably similar across all headphones used. L30 was always 3.8 to 4 db louder.

The difference is perfectly logical for Sennheiser HD800: more than double the watts at 300ohms in L30 II favor, the math is almost perfect at 3.8 dbs.

Where it made less sense was for Focal Clear and Hifiman HE1000, which are 55 and 32ohms impedance, respectively. The difference should have been 1db or around that, not 4.

And even less sense for Audeze LCD X and Dan Clark Aeon X Closed. 3500 vs 3300 watts at 16ohms would mean the SPL should have been nearly identical. But no, still 3.8 to 4 db difference.

As I said, I am aware it's not entirely but the result was always in the range 3.8-4 db averaged, always in favor of L30 II across couple dozens experimental pairs using the headphones mentioned above. Also, I measured the DX9 balanced output the same way and it was 5.9 to 6 db louder than SE output so the difference checks, as it should be 6 db over single end. Only have Focal and Hifiman balanced cable though so these were the only ones measured on balanced,

Audio theory is not my strongest suit so it's perfectly possible I am missing something but does someone have an explanation from seeing the same db difference for LCD X and HE1000 compared to HD800? For the first 2, if the specs are correct, the difference should be minimal to none.

Thanks in advance!

Edit: now I see the correct thread for this would have been Amplifiers, Phono preamp, and Analog Audio Review, if someone can move the post there (or wherever this kind of post belong, I would be grateful)
 
Last edited:
Watts and headphone impedance are mostly irrelevant for high crest factor, negative pre gain, AVG SPL measurements where the Amp is unlikely to clip.

What you need then is Gain.

DX9 SE Headphone out (High gain): 0dBFS = 8.4Vrms

DX9 RCA -> L30II (High gain): 0dBFS = 16.7Vrms (will clip at 13V but 0dBFS is unlikely here)

So the DX9 RCA -> L30II chain has 20*log10(16.7/8.4)= 6dB more effective gain from digital to analog, irrespective of the load attached.
 
Last edited:
Can you explain in simpler terms, please? :)

Also, I repeated the measurements with an 1000hz, 0dbfs sine wave on LCD-X. The results were 99.7dbs on DX9 vs 104.7dbs on L30. So this seems to confirm your theory.

Thank you!
 
Can you explain in simpler terms, please? :)

Also, I repeated the measurements with an 1000hz, 0dbfs sine wave on LCD-X. The results were 99.7dbs on DX9 vs 104.7dbs on L30. So this seems to confirm your theory.

Thank you!
Output power ratings and headphone impedances only come into play when you're playing loud enough to clip or distort the Amp.

When measuring average SPL however, which is routinely 15-25dB down from Peak, this does not come anywhere near to maxing out the Amp.

That means you can ignore the power ratings and headphone load impedance, as the only relevant metric is gain:

For a given digital input signal, how much analog output voltage can your signal chain produce?
 
For a given digital input signal, how much analog output voltage can your signal chain produce?
I am afraid I don't understand, sorry. Not great on audio theory and calculations :)
Output power ratings and headphone impedances only come into play when you're playing loud enough to clip or distort the Amp.

When measuring average SPL however, which is routinely 15-25dB down from Peak, this does not come anywhere near to maxing out the Amp.
The SPL of the sine wave was constant at 99.7dbs. Question: since Windown volume, gain and DX9 volume were all maxed out, how can I play something loud enough to clip or distort?

And thanks again for the patience and info!
 
Question: since Windown volume, gain and DX9 volume were all maxed out, how can I play something loud enough to clip or distort?
Increase the digital signal amplitude (measured in dB Full Scale = dBFS), until you reach the DX9's voltage or current limit.
 
Another question: I assume the 3300mW at 16ohms can only be reached in lab conditions or, as you said, incresing digital signal amplitude. This is doable with a sine wave but I assume not in real life usage? As in, listening to music. So in real life usage is more like 1100mW or so? Or i am over simplifying/not understanding completely?
 
Another question: I assume the 3300mW at 16ohms can only be reached in lab conditions or, as you said, incresing digital signal amplitude. This is doable with a sine wave but I assume not in real life usage? As in, listening to music. So in real life usage is more like 1100mW or so? Or i am over simplifying/not understanding completely?
3300mW output power at 16Ω is equal to 7.3Vrms amplitude.

With single-ended output in High gain, full volume the DX9 will (attempt to) output 8.4Vrms when fed 0dBFS from your PC.

So the DX9's clipping point at 16Ω is 20*log10(7.3/8.4)= -1.2dB down from full scale.

Therefore, with every volume slider and gain setting maxed out, whenever your music reaches an instantaneous signal level of -1.2dBFS, in that moment the DX9 will output 3300mW per channel to your 16Ω headphones.
 
Thank you! Between you and ChatGPT (which was actually going around in circles before I fed him your info), I now have a clearer understanding about things.

So, for practical purposes, the power is rarely the issue, gain is. Power just offers more headroom. And for 'normal' listening there aren't really other headphone amps that can get much louder than with L30 II or DX9 balanced, because gain limitations.

Just ordered HE6se. Tested it from a friend some time ago and, with -6db preamp, for some music it was somewhat on the quiet side even with L30 II on max gain and volume. I should expect the same for DX9 balanced or maybe just a tiny bit louder. A70 Pro would be around there too, correct? Can't think of any practical solution to increase the loudness on normal listening, while keeping the negative preamp. Are there any? Now I kinda understand the whole speaker amp for HE6 thing :)
 
Last edited:
Can't think of any practical solution to increase the loudness on normal listening, while keeping the negative preamp. Are there any?
First, make sure that the tracks that you're playing get close to 0dBFS Peak amplitude.

Replay Gain is purpose-built for that.

Any headroom your tracks leave to 0dBFS, limits your playback volume with zero upsides.

Next, there are plenty ways to get more gain. Here's a list:

ChainMax D->A Gain [Vrms/FS]Perceived gain-limited loudness relative to DX9 SE
DX1 II BAL18.5dB1x
DX9 SE18.5dB1x
DX5 II SE18.5dB1x
DX9 RCA + L30 II23dB1.4x
DX9 BAL24.5dB1.5x
DX5 II BAL24.5dB1.5x
DX9 BAL + A50 III BAL30dB2.2x
DX9 BAL + L70 BAL 32dB2.5x
DX9 BAL + A70 Pro BAL32dB2.5x
E50 II BAL + L70 BAL41dB4.8x
 
Also, I repeated the measurements with an 1000hz, 0dbfs sine wave on LCD-X. The results were 99.7dbs on DX9 vs 104.7dbs on L30. So this seems to confirm your theory.
DX9 balanced sits around 106db here so a bit over L30 II

So if I connect an L70 or A70 Pro to my DX9 and repeat my test, what would the results be, versus the ones above?
 
So if I connect an L70 or A70 Pro to my DX9 and repeat my test, what would the results be, versus the ones above?
See the chart above :)

The relative gain differences listed there, should correlate exactly to measured dB SPL differences, so long as you remain gain-limited and not current- or voltage-limited.
 
So if my measured db difference between SE and BAL is 6db, A70 Pro or L70 would be 10db over DX9 SE?
 
So A70 Pro or L70 would be 10db over DX9 SE?
Screenshot_20260630-153219_Chrome.png


What is 32 minus 18.5?

Edit:
Or are you asking about DX9 SE vs DX9 + A70 Pro SE?
 
No, about DX9 SE vs DX9 BAL + A70 Pro BAL. Just that I still didn't get how the dbs posted translate in perceived listening level, as I tested.

Also for DX9 BAL vs DX9 BAL + A70 Pro BAL, same thing.

More precisely, I can roughly measure 6dbs and clearly hear a big loudness difference when normally listen to music, between DX9 SE and BAL.,
Or a 4-5db difference between DX9 SE and L30 II. How would this difference translate this exact way if I add an A70 Pro BAL to DX9 BAL?
 
Last edited:
No, about DX9 SE vs DX9 BAL + A70 Pro BAL. Just that I still didn't get how the dbs posted translate in perceived listening level, as I tested.

Also for DX9 BAL vs DX9 BAL + A70 Pro BAL, same thing.
If you're trying to get quiet content to play as loud as you can, therefore maxing out every volume setting, then the config that has 10dB higher max D->A Gain than the other, will produce 10dB higher SPL.

Case in point:
ConfigMax D->A Gain [Vrms/FS]measured SPL by @Tatalor51
DX9 SE18.5dB99.7dB
DX9 + L30 II23.0dB (+4.5)104.7dB (+5)
DX9 BAL24.5dB (+6)106dB (+6.3)

See how the differences in theoretical gain track the differences in measured SPL?

(Slight deviations are expected with the way you measured SPL)
 
Last edited:
If you're trying to get quiet content to play as loud as you can, therefore maxing out every volume setting, then the config that has 10dB higher max D->A Gain than the other, will produce 10dB higher SPL.
So a DX9 BAL + L70/A70 Pro BAL would go 7.5dB higher than DX9 BAL? And I would be able to hear the difference in all scenarios just as I am able to hear it between DX9 SE and DX9 BAL?

Also, I repeated the measurements with an 1000hz, 0dbfs sine wave on LCD-X. The results were 99.7dbs on DX9 vs 104.7dbs on L30.

So if I test this with DX9 BAL + L70/A70 Pro BAL will get around 114dbs? (DX9 SE 99.7 + 14.5 ~ 114?)
 
So a DX9 BAL + L70/A70 Pro BAL would go 7.5dB higher than DX9 BAL?
Correct, in a gain-limited scenario.

And I would be able to hear the difference in all scenarios just as I am able to hear it between DX9 SE and DX9 BAL?
So long as the input signal is low enough that you're not clipping the Amp, yes you'll hear the increased gain in full.
 
What is 32 minus 18.5?
I meant 13.5 :) Which is also 7.5 + 6 :)

So for these scenarios (HE6, negative preamp gain and low volume masters), adding a balanced amp to the equation matters a lot. Because under these circumstances it won't clip that easily so most of the db difference is very useful.

Saw the correlation a little too late, I guess I was not expecting such a large difference. So a balanced DAC + A70 Pro/L70 should beat my DX9 by a pretty large margin, 7.5dbs is a lot!
 
Back
Top Bottom