• 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!

Topping A90 Discrete Review (Headphone Amp & Preamp)

Rate this Headphone Amp/Pre-amp

  • 1. Poor (headless panther)

    Votes: 17 4.0%
  • 2. Not terrible (postman panther)

    Votes: 15 3.6%
  • 3. Fine (happy panther)

    Votes: 72 17.1%
  • 4. Great (golfing panther)

    Votes: 316 75.2%

  • Total voters
    420

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
45,596
Likes
252,288
Location
Seattle Area
This is a review and detailed measurements of the new Topping A90 Discrete balanced preamplifier and headphone amplifier. It was sent to me by the company and costs US $599.

Topping A90 Discrete Review Headphone Amp Preamplifier Balanced.jpg

Announcement was supposed to be a few hours from now but one of the distributor already released it so here we go. As you can tell, there is a new design in the form of those pin-hole LEDs which I find quite elegant. Inside there is stepped relay for volume control. Of course as the name indicates, this design uses discrete transistors rather than op-amps. 39 transistors are used to accomplish that task in each module. Will performance be as good as op-amps? And is there an advantage? We will find out in the measurement section. Here is the back panel:

Topping A90 Discrete Review back panel Headphone Amp Preamplifier Balanced.jpg


I was pleasantly surprised to see the same extension connector to connect to Topping EXT90 to provide more inputs. In addition there is 12 volt trigger which many have been asking for. And even a ground lift switch.

In use, I found minor ticks during quiet parts of the music when changing the volume quickly. This is quite common in stepped relay implementations.

Topping A90 Discrete Measurements
In order to save time, all measurements are with balanced XLR input and balanced 4 pin XLR headphone output. Unbalanced has quarter of the power. Here is our dashboard:

Topping A90 Discrete Measurements Headphone Amp Preamplifier Balanced.png


While SINAD of 119 dB is superb and assures transparency in all conditions, it is a hair worse than A90:
best headphone amplifier review 2022.png


This is incredible performance as every other discrete headphone amplifier I have tested has given a lot of performance (in the form of noise and distortion) to get there. Dynamic range is a bit lower which is responsible for that:
Topping A90 Discrete Measurements SNR Headphone Amp Preamplifier Balanced.png


We see that penalty in the 50 mv test:
most quiet headphone amplifier 2022.png


Multitone test shows how low the distortion products are:
Topping A90 Discrete Measurements Multitone Headphone Amp Preamplifier Balanced.png


Frequency response is of course extended and flat:

Topping A90 Discrete Measurements frequency response Headphone Amp Preamplifier Balanced.png


THD vs frequency shows a slight rise at the top end:
Topping A90 Discrete Measurements THD vs Frequency Headphone Amp Preamplifier Balanced.png


I measured power at three impedances: 300, 50 and 32:

Topping A90 Discrete Measurements Power into 300 ohm Headphone Amp Preamplifier Balanced.png


Topping A90 Discrete Measurements Power into 50 ohm Headphone Amp Preamplifier Balanced.png



Topping A90 Discrete Measurements Power into 32 ohm Headphone Amp Preamplifier Balanced.png


The advantage of discrete implementation seems to be in current delivery into low impedances:
Topping A90 Discrete Measurements THD vs power vs load Headphone Amp Preamplifier Balanced.png


Usually I see a pull back at 32 ohm but here, that load doesn't impact the amp at all. It is not until we get to 20 ohm that the output voltage starts to drop a bit. This is a first since I started to make this measurement.

When pushed into protection with loads < 32 ohm, the unit shut off. Powering on everything was fine so it passed this stress test.

The beauty of stepped relay volume control is perfect channel matching which we see here:
Topping A90 Discrete Measurements channel match  Headphone Amp Preamplifier Balanced.png


I am amazed how well this works. Some stepped relay implementations lose their grip at low volumes but there we go past our limits of hearing before the mismatch even registers!

Topping A90 Discrete Listening Tests
I have been using my everyday Dan Clark Stealth headphone as a "it won't drive it but let's see what it does" test for headphone amps. This is the first time I have seen an amplifier have such an upper hand over these headphones. The sound is incredibly dynamic and beautiful. "Loud" listening is at volume level 80 in high gain. I cranked it up to 95 before I got scared for my hearing without any hint of distortion. I had not heard these headphones come to life so well.

Needless to say, driving Sennheiser HD650 and Drop Ether CX was in a walk in the park. I am talking incredible bass, detail and fidelity with volume indicator in 60s.

What a joy....

Conclusions
I must confess that I got a bit worried when Topping told me about the A90 Discrete. I worried that they were making a discrete version just to appease audiophiles but with much loss in noise and distortion metrics. Happily Topping approaches everything with extreme focus on performance and they managed to get the A90 discrete within a split hair of the op-amp version. But then innovated with stepped ladder volume control, trigger support and major step up in class/look of the unit. I happily take these as a trade off for that 1 dB of SINAD.

From design point of view, this is another watershed audio/engineering event brought to us by Topping. Who would have thought that this company would keep breaking new ground instead of all the major/usual names we know in hi-fi.

It is my pleasure to add Topping A90 Discrete headphone amplifier/pre-amplifier to my recommended list. My challenge is to figure out if it goes in my main system or desktop! :)

------------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
 
At first was confused by the USB-C connection labled "FW" - since I read that as "Firewire" not "Firmware" :facepalm: Nice to know they made it upgradeable (or at least patchable) in case there are any issues down the road.

Although considering the performance, it definitely doesn't seem like they left anything off the table at release.
 
Great review!

I have been using my everyday Dan Clark Stealth headphone as a "it won't drive it but let's see what it does" test for headphone amps. This is the first time I have seen an amplifier have such an upper hand over these headphones. The sound is incredibly dynamic and beautiful. "Loud" listening is at volume level 80 in high gain. I cranked it up to 95 before I got scared for my hearing without any hint of distortion. I had not heard these headphones come to life so well.

The Stealth is also my headphone of choice, and I have never had gotten close to maxing out the volume on my A90 with them. Would you say that there is any real sound quality upgrade to the A90 Discrete over the A90 for this use case, or is it just qualify of life advantages?

Thanks
 
The Stealth is also my headphone of choice, and I have never had gotten close to maxing out the volume on my A90 with them. Would you say that there is any real sound quality upgrade to the A90 Discrete over the A90 for this use case, or is it just qualify of life advantages?
I have to do a side-by-side test but for now, I think the Discrete has more headroom than the A90.
 
The beauty of stepped relay volume control is perfect channel matching which we see here:
Topping A90 Discrete Measurements channel match  Headphone Amp Preamplifier Balanced.png
Would be great to see evolution of SINAD or SNR vs attenuation on such a graph (and compare to digital attenuation)
Not all attenuators work the same on that point.
We should probably also see some benefit of the switched/relays attenuator.
 
Last edited:
Excellent review, ordered one! @amirm did you measure any difference between the Ground and Lift settings?
 
Please note that the noise performance of the XLR output is always gonna be slightly worse than the 6.35mm output same for the A90.
The 6.35mm output performance is essentially the same as A90.
 
It's disappointing that it's $100 more expensive than its predecessor, but we're excited about the new performance. Order complete!
Both models will be on at the same time for time to come. This provides a different flavor of functionality. It's not gonna be right if this is cheaper than the Pre90 right? Cheers.
 
For those of us who are not audio engineers, can you explain in layman's terms why one would choose this over other similar products from Topping or others? I would like to be able to be able to appreciate it better.
 
Would be great to see evolution of SINAD or SNR vs attenuation on such a graph (and compare to digital attenuation)
I have not been able to make this work in the past. As you change the levels the analyzer THD meter goes through the roof and then back down. So you get a lot of glitches.
 
Excellent review, ordered one! @amirm did you measure any difference between the Ground and Lift settings?
I did not. Truth to be told, I only saw that when I was taking its picture. :) I had no mains noise anyway since I was using differential connections. Will test with RCA at some point to see if it helps.
 
Back
Top Bottom