Topping A90 Headphone Amplifier Review

[Atanasi]

Is 2.9 V confirmed for single ended? Because if so, then I would assume XLR should take 5.8ish V. There is 2.9 V on product page as input sensitivity on high gain but the numbers are all same for single ended and balanced so it doesn't make sense.

The common implementations of balanced outputs double the voltage (+ 6 dB boost), but balanced inputs don't halve the voltage. That is, if you set the amp to unity gain and send 1 V to a balanced input, the same 1 V comes out of the unbalanced headphone output. The input sensitivity is the same with both unbalanced and balanced inputs. I don't know why they specifically told to use 5-8 V range. It is a valid input range when using A90 at low or medium gain.
It doesn't matter very much as long as you get the desired output volume, and no device is set to so low volume that it has to be amplified significantly at a later stage, because that would lose SNR.

 

[Jarlaxle]

The common implementations of balanced outputs double the voltage (+ 6 dB boost), but balanced inputs don't halve the voltage. That is, if you set the amp to unity gain and send 1 V to a balanced input, the same 1 V comes out of the unbalanced headphone output. The input sensitivity is the same with both unbalanced and balanced inputs. I don't know why they specifically told to use 5-8 V range. It is a valid input range when using A90 at low or medium gain.
It doesn't matter very much as long as you get the desired output volume, and no device is set to so low volume that it has to be amplified significantly at a later stage, because that would lose SNR.

Now it makes a little more sense. Now that I see there was a similar discussion about this topic at page 79 where @JohnYang1997 says lower the volume to avoid clipping, and user @Ddd confirmed their amp, fed by D90 at 4Vrms, clipped at high gain at past volume 1-2 o'clock which is kinda sad. I would have quoted the posts but it is a back and forth convo so I tried to explain in short. Anyone interested can check it out.

 

[Selarom]

Anyone here that enjoys the A90 paired with a SMSL SU-9n / SU-9? Can you share a picture of the stack? Is it a good pairing?

 

[Ddd]

Now it makes a little more sense. Now that I see there was a similar discussion about this topic at page 79 where @JohnYang1997 says lower the volume to avoid clipping, and user @Ddd confirmed their amp, fed by D90 at 4Vrms, clipped at high gain at past volume 1-2 o'clock which is kinda sad. I would have quoted the posts but it is a back and forth convo so I tried to explain in short. Anyone interested can check it out.

Well. From what I look back, A90 output capabilty at M or H gain are almost the same if I'm not wrong
(both Gain settings have Same output swing level 49Vp-p of dynamic range)

M Gain 2.0x , Max input 8.8V
Multiply get A90 max voltage output 17.6V

H Gain 6.0x , Max input 2.9V
Multiply get 17.4V ..

0.2V difference negligible? M gain better options?

Plus 17V output is deafening loud. Don't think anyone can get there. So although A90 limit is there, my ears volume limit clips way earlier before A90 clips at high gain XD joking....
So I'll say don't worry about the clip at H gain.. It will only H gain clip past 1pm only if your music is constantly 0dB full range 4Vrms which in real world rarely goes above 3Vrms from my experience..
And music is not constant.. Is a varying voltage as the music goes..
So when use with source that are below 2.9V... Don't worry, the A90 doesn't clip at H gain past 1pm.


Only H gain lower the input sensitivity to help boost the headphone output to its maximum when use with source that are not capable of providing standard balanced 4V output to A90 (below 2.9V balanced source)

If your source have standard balanced output 4V or more , high variable output voltage range up to 8.8V at M gain you can ALREADY fully extract the max output potential capabilty of A90 voltage output up to 17.6V...

Basically the M gain are everything of A90 and H gain is an extra overdrive mode for sources that can't provide A90 with enough juices to get its max output....

Lastly from what I found out previously. The input sensitivity in A90 is not fixed.. Is changing with the volume you used. At H gain when I provide it with 4Vrms.. before 1pm it doesn't clips.. Suggest the input sensitivity before 1pm is probably higher than 2.9V.. And at 1pm is where the input sensitivity reach lower than full range 4Vrms

Reason I say this is, A90 in this way is better at least you still can use 4Vrms source at H gain up to 1pm.
While my past experience with THX 789 amp , with a 4Vrms capable source.. No matter the output volume even at the lowest. Is a distorted clip sound.. As if the full gain value (6.6x) is applied at the input signal and clips the amp even before reaching the headphone output..
So maybe can say A90 gain value amount increase with volume output...

Above are just my impression and opinion. I may be wrong. Correct me if I am. I'm also still learning thanks

 

[Jarlaxle]

Well. From what I look back, A90 output capabilty at M or H gain are almost the same if I'm not wrong
(both Gain settings have Same output swing level 49Vp-p of dynamic range)

M Gain 2.0x , Max input 8.8V
Multiply get A90 max voltage output 17.6V

H Gain 6.0x , Max input 2.9V
Multiply get 17.4V ..

0.2V difference negligible? M gain better options?

Plus 17V output is deafening loud. Don't think anyone can get there. So although A90 limit is there, my ears volume limit clips way earlier before A90 clips at high gain XD joking....
So I'll say don't worry about the clip at H gain.. It will only H gain clip past 1pm only if your music is constantly 0dB full range 4Vrms which in real world rarely goes above 3Vrms from my experience..
And music is not constant.. Is a varying voltage as the music goes..
So when use with source that are below 2.9V... Don't worry, the A90 doesn't clip at H gain past 1pm.


Only H gain lower the input sensitivity to help boost the headphone output to its maximum when use with your source that are not capable of providing high balanced output to A90 (below 2.9V balanced source)

If your source have standard balanced output 4V or more , high variable output voltage range up to 8.8V at M gain you can ALREADY fully extract the max output potential capabilty of A90 voltage output up to 17.6V...

Basically the M gain are everything of A90 and H gain is an extra overdrive mode for sources that can't provide A90 enough juices to get its max output....

Lastly from what I found out previously. The input sensitivity in A90 is not fixed.. Is changing with the volume you used. At H gain when I provide it with 4Vrms.. before 1pm it doesn't clips.. Suggest the input sensitivity before 1pm is probably higher than 2.9V.. And at 1pm is where the input sensitivity reach lower than full range 4Vrms

Reason I say this is, A90 in this way is better at least you still can use 4Vrms source at H gain up to 1pm.
While my past experience with THX 789 amp , with a 4Vrms capable source.. No matter the output volume even at the lowest. Is a distorted clip sound.. As if the full gain value (6.6x) is applied at the input signal and clips the amp even before reaching the headphone output..
So maybe can say A90 gain value amount increase with volume output...

Above are just my impression and opinion. I may be wrong. Correct if I am. I'm also still learning thanks

If what you say is true about medium and high gain, interesting design choice. Definitely a different approach from SP400 which is gain limited and almost never clips. Also I decided to keep my SP400, not because of the power capacity. Both devices are more powerful than I need as you said. (And I have an Emotiva BasX A100 just in case ) But a remote control and r2r volume relay for better preamp capabilities are nice things to have. Thank a bunch though

 

[Ddd]

If what you say is true about medium and high gain, interesting design choice. Definitely a different approach from SP400 which is gain limited and almost never clips. Also I decided to keep my SP400, not because of the power capacity. Both devices are more powerful than I need as you said. (And I have an Emotiva BasX A100 just in case ) But a remote control and r2r volume relay for better preamp capabilities are nice things to have. Thank a bunch though

Yeah exactly. Having more than enough power than we need it...

From my check, SP400 didn't mention it's max input sensitivity.

But from my rough calculation from its max balanced voltage output 16.25V from its 440mW at 600 ohm...

With lower Gain value than A90,
11.6dB is 3.8x...

So the max input voltage maybe 4.26Vrms

Just my guess. If providing higher input voltage greater than 4.26V at that gain settings probably will clip it as well...

So my guess is maybe SP400 assume user use 4Vrms input and they lower the maximum gain value to make sure the user won't use it over its clipping max voltage output... Guess...

While A90 may also done the same thing just not stated maybe.. Which may be the same thing as SP400 before 1pm ...

Probably both having the same thing in mind but state their specs differently..
A90 lower H gain input sensitivity value to know clipping level at 1pm position
So anything before 1pm the gain value amount is lower than its maximum 6.0x probably around same as SP400 3.8x.. Which also means before 1pm input sensitivity 4V is usable.

while SP400 lower H gain value at specs to know the maximum gain value limit with assumed fixed 4.26V input as max input sensitivity

Hope it helps

 

[Jarlaxle]

Yeah exactly. Having more than enough power than we need it...

From my check, SP400 didn't mention it's max input sensitivity.

But from my rough calculation from its max balanced voltage output 16.25V from its 440mW at 600 ohm...

With lower Gain value than A90,
11.6dB is 3.8x...

So the max input voltage maybe 4.26Vrms

Just my guess. If providing higher input voltage greater than 4.26V at that gain settings probably will clip it as well...

So my guess is maybe SP400 assume user use 4Vrms input and they lower the maximum gain value to make sure the user won't use it over its clipping max voltage output... Guess...

While A90 may also done the same thing just not stated maybe.. Which may be the same thing as SP400 before 1pm ...

Probably both having the same thing in mind but state their specs differently..
A90 lower H gain input sensitivity value to know clipping level at 1pm position
So anything before 1pm the gain value amount is lower than its maximum 6.0x probably around same as SP400 3.8x.. Which also means before 1pm input sensitivity 4V is usable.

while SP400 lower H gain value at specs to know the maximum gain value limit with assumed fixed 4.26V input as max input sensitivity

Hope it helps

Actually somewhere along the line, SMSL or at least Mandy confirmed that SP400's max input voltage is 10 Vrms and their calculations made with 5.5 Vrms which is stacks M400's max output value. Although Amir's measurements are taken with 4V input, potential seems to be higher. There are also people using their SP400 with RME ADI-2 DAC FS with 7 V output with great results. That's why I said different approach. Maybe should have gave some context earlier. Thanks again.

for context this is where I read: https://www.audiosciencereview.com/...smsl-sp400-review-headphone-amp.19346/page-16

 

[Ddd]

Actually somewhere along the line, SMSL or at least Mandy confirmed that SP400's max input voltage is 10 Vrms and their calculations made with 5.5 Vrms which is stacks M400's max output value. Although Amir's measurements are taken with 4V input, potential seems to be higher. They are also people using their SP400 with RME ADI-2 DAC FS with 7 V output with great results. That's why I said different approach. Maybe should have gave some context earlier. Thanks again.

I see..

Great then. But for the higher voltage input acceptance. The gain should be lower and not max H gain I believe?

Like the A90 8.8V at M gain maybe...


Anyhow. Good to know. Great info

 

[Jarlaxle]

I see..

Great then. But for the higher voltage input acceptance. The gain should be lower and not max H gain I believe?

Like the A90 8.8V at M gain maybe...


Anyhow. Good to know. Great info

To be honest no idea, but I doubt it because there really isn't much of a difference in SP400 different gain settings.

 

[akatenev]

Do I need any special tool for disassembly A90?

 

[Jarlaxle]

Do I need any special tool for disassembly A90?

I remember Goldensound trying to tear A90 apart for his review but couldn't.

 

[the_brunx]

Has anyone else experienced ground noise problems with their A90?

 

[Talisman]

Excuse the ignorant question, with this A90, if I sent Rca input, I could have a balanced output at full power and then use a dac SE?

 

[Veri]

Excuse the ignorant question, with this A90, if I sent Rca input, I could have a balanced output at full power and then use a dac SE?

SE input will not give full power as opposed to balanced input. But it will still give plenty.

 

[staticV3]

Excuse the ignorant question, with this A90, if I sent Rca input, I could have a balanced output at full power and then use a dac SE?

You need a 1.85Vrms signal amplitude at the RCA input to reach full output @16Ω, High gain, Bal out.
sqrt(7.6*16)*10^(-15.5/20)≈1.851

At 32Ω you need 2.4Vrms RCA in to max out the A90 in high gain, Bal out.
At 300Ω and above you need 2.9Vrms.

 

[Talisman]

You need a 1.85Vrms signal amplitude at the RCA input to reach full output @16Ω, High gain, Bal out.
sqrt(7.6*16)*10^(-15.5/20)≈1.851

At 32Ω you need 2.4Vrms RCA in to max out the A90 in high gain, Bal out.
At 300Ω and above you need 2.9Vrms.

Sorry, I didn't specify well, I would like to use it as a preamp and my question is, if by sending a single ended input signal, I can get a balanced output from the xlr outputs.
In practice, does it convert single ended to balanced?
Thanks

 

[Veri]

Sorry, I didn't specify well, I would like to use it as a preamp and my question is, if by sending a single ended input signal, I can get a balanced output from the xlr outputs.
In practice, does it convert single ended to balanced?
Thanks

Yes, all outputs are active. It's a "real" pre-amp function not some passive pre-out

 

[Talisman]

Sì, tutte le uscite sono attive. È una funzione di preamplificazione "reale" non una pre-uscita passiva

Thanks!

 

[Tokyo_John]

Finally got an A90 to pair with my Asahi Kasei DAC MQA version D90. Listening to my favorite test tunes with Ultrasone Ed 8 Ruth over simple SE for starters.

Wow.

 

[kyleTheDude]

What's the better combo? A90 + D10b or L50 + E50