Sabaj A10h Review (Headphone Amplifier)

[Roland68]

Two big problems:
a) Crosstalk
b) Ground loops and/or returns from power supply (with some PC you can hear background noise).

I couldn't detect any excessive crosstalk on my A10h and A20h, even compared to other HPAs. Could it be that it's caused by the cabling (RCA or headphone cable) or a device connected upstream?

Ground loops always come from device configurations. In connection with PCs, the USB ground or the ground of the sound connection is usually the original problem. The A10h is probably not to blame for that.
It's best to list the entire chain.

 

[Saponetto]

Neither my A10h has crosstalk problems.

 

[gcompari]

I couldn't detect any excessive crosstalk on my A10h and A20h, even compared to other HPAs. Could it be that it's caused by the cabling (RCA or headphone cable) or a device connected upstream?

Ground loops always come from device configurations. In connection with PCs, the USB ground or the ground of the sound connection is usually the original problem. The A10h is probably not to blame for that.
It's best to list the entire chain.

I attach the A10h crosstalk: as you can see there are only 60dB of isolation between channels.
If you use IEMs, you can remove one of the IEMs from your ear and sweep the removed channel (only) with a 600Hz to 5kHz sinewave: you can clearly hear the signals generated in the other channel.
I attach the crosstalk of the JSD Labs Atom Amp (first series, not "+" or "2"), Topping L30ii and Topping A50s: I think you can see the diffference...

In exactly the same configuration (same PC, same DAC, same cables, same connections, same headphones) I just had to replace the Sabaj A10h with the JDS Labs Atom Amp or the Topping L30ii or the Topping L30 to have no more noise o problems. Note that the Atom Amp, the L30ii and the L30 have a transformer and not a switching power supply as the A10h does...
You can see also: https://www.l7audiolab.com/f/sabaj-a10h/

 

[vins555]

I have Sabaj A20h and when I plug the headphones in 6.3mm it has no crosstalk, but when I plug in 4.4mm balanced it has a bit of audible crosstalk, Anyway the crosstalk is way less then in Asgard 2.

 

[gcompari]

I have Sabaj A20h and when I plug the headphones in 6.3mm it has no crosstalk, but when I plug in 4.4mm balanced it has a bit of audible crosstalk, Anyway the crosstalk is way less then in Asgard 2.

Very strange: I no longer have the A20h to measure the crosstalk of the 6.3mm output: the 4.4mm balanced output was very similar to the A10h 6.3mm output.
(I sent back to the dealer my A20h because the crosstalk levels)
Normally, the balanced design (except for the unbalanced "dual mono" design) is the only way to reach very low levels of crosstalk.
I attach the crosstalk graphs of E1DA #9038S (balanced, max crosstalk -106dB @20kHz) and #9038D (unbalanced, crosstalk -75dB).
Please note that to avoid any switch off or automatic noise reduction of the channel w/o input, I drove the not sweeped channel by a fixed inaudible frequency.
The E1DA #9038S is one of (DAC+)HPA I own whose crosstalk levels are inaudible (at least for me: I can hear crosstalk to -85dB; probably younger people may hear lower levels).

 

[vins555]

I didn't measure, I only used my ears(unpluged one side of headphone). I can say that Little Dot Mark 2 has no audible crosstalk and it's not balanced, Creative AE-5 headphone output has no audible crosstalk also. But Asgard 2 has it more then Sabaj A20h, The champ in crosstalk is IFi ZEN CAN, this amp has a lot, in both balanced or unbalanced.

 

[gcompari]

I didn't measure, I only used my ears(unpluged one side of headphone). I can say that Little Dot Mark 2 has no audible crosstalk and it's not balanced, Creative AE-5 headphone output has no audible crosstalk also. But Asgard 2 has it more then Sabaj A20h, The champ in crosstalk is IFi ZEN CAN, this amp has a lot, in both balanced or unbalanced.

I found some measurements of the Creative AE-5: you can see a -58dB of crosstalk at 16 ohms (from the attached graph). I think it is very audible.

You can find the review of the Little Dot MK III on ASR:

Review and Measurements of Little Dot MK III Tube Headphone Amplifier

This is a review and detailed measurements of the Little Dot MK III Tube Headphone amplifier. It is on kind loan from a member. It costs US $289 including Prime shipping. The unit has been sitting for a while (my apologies to its owner) so please excuse a bit of dust in the pictures below. I...

www.audiosciencereview.com

there isn't the crosstalk graph (unfortunately), but from the other measurements I can see, I wouldn't be surprised to see high crosstalk values (may be covered by the noise).

To understand if there is an audible crosstalk you have to make some steps:
- Use IEMs: they help to isolate any external noise.
- Check if your amp can work w/o the load on one channel only, otherwise you need to realize a dummy load with a power resistor of the same value of the impedance at 1 kHz of your IEMs; some HPA switch from "Phones" to "Preamp" mode if you disconnect the output load.
- You must set the HPA volume level with a track with a significative dynamic (for example. the 6^ Symphony of Beethoven, 6th mov., half time) at a realistic audio level (thinking to be in front of the stage of a theater).
- You must install a signal generator in your audio source (phone, tablet, PC): for example (PC only) RTA DSS3 (free for the signal gen.).
- You have to generate at 0dBFS a sweep of frequencies form 600 to 5kHz in the "A" channel.
- Disconnect the IEM of the "A" channel (leaving from the ear the IEM it isn't enough: you can hear some sound from it if it is connected) and connect the dummy load (if necessary to avoid abnormal behaviors of the HPA: switching to "preamp" mode, switching off, etc.).
- Concentrate on the channel "B" IEM: you mustn't hear absolutely nothing.

If you don't hear absolutely nothing, the crosstalk of your amp is inaudible (at least for you).

To check your sensitivity to crosstalk:
- Use a DAC or HPA that has a dB level display and/or a digital level volume control graduated in dB (1dB step, or better 0,5dB step).
- Play at a reasonable high level (a realistic level of a orchestral "fortissimo") a 0dBFS 3kHz sinewave (or better a continuous sweep 600Hz<->5kHz) and go down the volume control until you can hear absolutely nothing.
- The difference from the original listening volume and the volume that you don't hear anything is you (subjective) crosstalk threshold sensitivity.

The full balanced design you permit to realize very low crosstalk amps: it seems to be a "necessary" condition, but not "sufficient".

Topping L30ii is one of the best unbalanced HPA ever measured by ASR (if not the best) and has an audible crosstalk...

 

[vins555]

Anyway it's not a problem, even on Ifi ZEN CAN while this amp has a lot of crosstalk. But others not even close to ZEN CAN, so not a problem it's still stereo. If you really need to ged rid of it, then double mono is the way?

 

[gcompari]

A good "Dual Mono" (normally very expensive and very bulky: double PSU, special case, etc.) or a good "Full Balanced" HPA/DAC+HPA: so far I have found low (inaudible) levels of crosstalk (<85dB at any frequency) only in: E1DA #9038S, Tanchjim Space Dual and Tanchjim Luna (all balanced).

I ordered the Full Balanced Topping L70 (I found it @259.23$, but it will be shipped on January) that in some test measures very well about crosstalk:

https://cdn.l7audiolab.com/wp-content/uploads/2022/10/1666693096-Crosstalk.jpg

I tried Topping A50s (balanced), L30, L30ii, L50, Sabaj A10h, A20h (balanced), SMSL C200, JDS Labs Atom Amp, but the crosstalk is always audible.

 

[vins555]

But in fact for the headphones some audiophiles do like to add some crossfeed

 

[gcompari]

But in fact for the headphones some audiophiles do like to add some crossfeed

I leave the "magic button" to "some audiophiles": from the HiFi DIN 45500 (year 1966) the crosstalk has always been defined as a defect, as you can see:

Crosstalk - Wikipedia

en.wikipedia.org

Many years ago I bought a Meier's "Porta Corda" HPA with the "Crossfeed" option, like this:

Shop Meier Audio Porta Corda III USB & Discover Community Reviews at Drop

Portable amplifier with USB DACDiscontinued.Now being sold as a DIY kit.

drop.com

After trying it, I asked Jan Meier how to totally remove the crossfeed network (the front "crossfeed" switch didn't completely remove the "crossfeed" effect) who told me which capacitor and resistor to unsolder: after that, the Porta Corda became a very notable HPA.
If you're interested, I can show you why "crossfeed" destroys one of major advantages of headphones listening over listening with speakers.
But I think we have to start a different thread to avoid annoying people who are interested only in the Sabaj A10h.

 

[asoka]

Hi!
About using Sabaj a 10h as an preamp...
I do not clearly understand if the preamp signal will pass also thru the amplifier chip on the sabaj...also changing sound signature and scene width
or the signal is passed only thru stepped relays and resistors as only an passive volume without further interference with original signal?
I also do not understand how the gain stages work, because they work also in the pre output...this suggesting that the signal is routed also thru amplifier chip?
The stages are low -6 db normal at 0 and high at +9 db.
I use this sabaj as an additional step to lower the signal power of my dac that output to much, between the dac and the actual hifi tier preamp to be able to use the later preamp volume pot in an usable way...without sabaj prior dampening the signal the volume pot of the second high quality preamp could be operated only on the start of the scale because to loud sound after that.

It is ok and better this solution of using sabaj as preamp in cascade with the second preamp? or better to use just an 10-12 db signal dampener to the input ports of the amplifier?
What method do you think that will better preserve the sound and scene?
I could say that using sabaj in this chain I could not notice any sound degradation, but that rca dampeners I do not tried to see if they do bad to sound.

 

[Roland68]

Hi!
About using Sabaj a 10h as an preamp...
I do not clearly understand if the preamp signal will pass also thru the amplifier chip on the sabaj...also changing sound signature and scene width
or the signal is passed only thru stepped relays and resistors as only an passive volume without further interference with original signal?
I also do not understand how the gain stages work, because they work also in the pre output...this suggesting that the signal is routed also thru amplifier chip?
The stages are low -6 db normal at 0 and high at +9 db.
I use this sabaj as an additional step to lower the signal power of my dac that output to much, between the dac and the actual hifi tier preamp to be able to use the later preamp volume pot in an usable way...without sabaj prior dampening the signal the volume pot of the second high quality preamp could be operated only on the start of the scale because to loud sound after that.

It is ok and better this solution of using sabaj as preamp in cascade with the second preamp? or better to use just an 10-12 db signal dampener to the input ports of the amplifier?
What method do you think that will better preserve the sound and scene?
I could say that using sabaj in this chain I could not notice any sound degradation, but that rca dampeners I do not tried to see if they do bad to sound.

Yes, the signal is amplified in the Sabaj A10h, it is not a passive preamplifier.
It has been well tested here and is very transparent and very low-noise, perhaps even better than your second preamplifier. The price is no longer a big factor in something like this.

But you have both preamplifiers, test them directly against each other, perhaps the A10h will surprise you more than you think. And if not, that's no problem, then at least you have certainty.

 

[asoka]

Yes, the signal is amplified in the Sabaj A10h, it is not a passive preamplifier.
It has been well tested here and is very transparent and very low-noise, perhaps even better than your second preamplifier. The price is no longer a big factor in something like this.

But you have both preamplifiers, test them directly against each other, perhaps the A10h will surprise you more than you think. And if not, that's no problem, then at least you have certainty.

Yep! The hifi preamp it is a very high level and really well reviewed... it is an Acoustic Invader and it function as class A and it is j-fet based. Until it arrived I used the sabaj few days and when switching to this higher tier one the change in sound quality was not striking as expected.

But may the rca fixed attenuators and removing sabaj could be a better solution?
I need to test the sound thru sabaj only and thru expensive preamp only...but I am very bad at spotting differences in short term when changing devices it is really an hard struggle to perceive them and I tend to base on the opinion of people more able to tell the differences