Dual Merus MA12070 Amplifier Sabaj A20 fully balanced : pictures inside )

[Toku]

I got another A20a from the same seller (amazon.de). This time it is the updated version.

The differences:
- no relay clicking
- higher power draw. It consumes 2W more power (idle: 3W vs 1W, music: 6W vs 4W).
- a sticker with a serial number underneath
- the previous A20 had an extra hole on the back panel (right side to the 15V DC output) which was masked with some black sticker. This newer version doesn't have such hole.
- and one more thing:

My audio chain is like so: PC -> DAC -> A20A -> Speakers. With the old A20a, speakers are near dead-silent when not playing a sound. But with the new one, there is a constant electronic noise that is audiable and quite annoying at 2 meters from the speakers even at -20db at the amp (it becomes inaudible around -30db).
The noise goes away when I unplug the PC, but that defeats the purpose. The A20a is the only variable here - even the cables are the same.
I tried turning the internal voltage regulator to the max, as suggested earlier in this thread, but that did not change anything.

Any idea what can cause that difference in noise or how to get rid of it?

The noise problem was reported to Sabaj early after the launch of the A20a. But they never admitted that noise was generated.
Many people have reported on this forum how to remove this noise, but it has not been completely removed.
I think the cause is the internal structure of the A20a.

 

[lockheed]

The noise problem was reported to Sabaj early after the launch of the A20a. But they never admitted that noise was generated.
Many people have reported on this forum how to remove this noise, but it has not been completely removed.
I think the cause is the internal structure of the A20a.

I don't think we are talking about the same thing.
The noise I am talking about - when computer is plugged into DAC - was not there in A20a v1 but is present in A20a v2.

 

[Moosi]

I'm quite experienced with modding this amp in terms of silencing it. I've managed to bring it from noise starting from about 60% of volume to being dead silent up to 98% of volume (-2db on the scale). Both with RCA plugged and ears to speakers directly. But this is with the original (clicking relay) version of which lockheed claims to be more silent by default.

Since I've heard this more than once already, I believe by getting rid of the clicking relay, they have somehow "screwed up" the Mornsun PSU leading to more noise leakage. There are multiple approaches for you. From Top (easiest to do) to bottom

a) Connect via balanced XLR should remove most of it
b) Get this thing: https://www.arrow.com/en/products/b64290l40x830/epcos-tdk / wrap it around the AC-IN cable shortly but not immediately before the plug that goes into the amp. Like This
c) Get ferrid bead and plug them to the speakers out cables. Like This
d) Open it up and set the voltage adjustment knob on the PSU to max. (nearly 26V).
e) Open it up and insert a protection/shielding cover over the Speakers-out-panel and ferrite bead to the speakers out cables from inside.
f) If that does not lower the noise enough, get a Meanwell PSU. I'm aware that you are hesitant to spend extra coin, but you can get a EPP-300-24 pretty cheaply around 40-50 Euros. Does not have to be the bigger and more expensive RPS-400-24. The trick with the Meanwells is that they are running in a "no burst mode" when being maxed out to 26V only. If a high power PSU capable of multiple hundred watts is running an amp with 1-3W idle, it must constantly "cut off" supply shortly in order to sustain a stable current at such low levels being energy efficient and comply to european energy saving standarts. That is what is called "burst mode". The PSU is sending small bursts instead of a stable stream of current. This can (but must not) introduce additional noise within the amp. With the Mornsun PSU there is always some sort of burst mode running. The Meanwells have that special function to skip that when being adjusted to max via the knob. The downside is about 6W in idle instead of 2-3W.

If you are hesitant to mod the amp, which I understand and will loose your warranty, I would start with b) and see if that improves it to a degree that satisfies your needs. This kind of Toroid ring is cheap and never a "wasted buy" for it will benefit every electronical device being powered from wall-mains. Any amp powered by internally installed PSUs should have one. Except maybe for those being really carefully designed for repelling any sort of mains leakage (which are few). If you hate all this crap, return it and get the SMSL AO200 or the ToppingPA3. These two should be dead silent by default, but both have a bit less power and "punch" overall.

 

[Toku]

If you hate all this crap, return it and get the SMSL AO200 or the ToppingPA3. These two should be dead silent by default, but both have a bit less power and "punch" overall.

Your opinion is exactly what I felt when comparing the sound quality of the A20a V1/V2 with the PA3s and AO200.

 

[Nozza]

I got another A20a from the same seller (amazon.de). This time it is the updated version.

The differences:
- no relay clicking
- higher power draw. It consumes 2W more power (idle: 3W vs 1W, music: 6W vs 4W).
- a sticker with a serial number underneath
- the previous A20 had an extra hole on the back panel (right side to the 15V DC output) which was masked with some black sticker. This newer version doesn't have such hole.
- and one more thing:

My audio chain is like so: PC -> DAC -> A20A -> Speakers. With the old A20a, speakers are near dead-silent when not playing a sound. But with the new one, there is a constant electronic noise that is audible and quite annoying at 2 meters from the speakers even at -20db at the amp (it becomes inaudible around -30db).
The noise goes away when I unplug the PC, but that defeats the purpose. The A20a is the only variable here - even the cables are the same.
I tried turning the internal voltage regulator to the max, as suggested earlier in this thread, but that did not change anything.

Any idea what can cause that difference in noise or how to get rid of it?

Possibly some sort sort of grounding issue maybe? Can you connect the PC to the DAC via TOSLINK instead, to remove the electrical connection and see if the noise goes away?

Wondering if the new A20a differs in grounding in some way.

I have the latest version (I believe) of the A20a and mine is silent. However my system is not silent if I connect my DAC to my PC via USB (without the USB isolated dongle thing I have). That isn't because of the A20a. It's some issue with noise being transferred from my PC's graphics card. Currently I'm just using TOSLINK.

 

[Roland68]

Possibly some sort sort of grounding issue maybe? Can you connect the PC to the DAC via TOSLINK instead, to remove the electrical connection and see if the noise goes away?

Wondering if the new A20a differs in grounding in some way.

I have the latest version (I believe) of the A20a and mine is silent. However my system is not silent if I connect my DAC to my PC via USB (without the USB isolated dongle thing I have). That isn't because of the A20a. It's some issue with noise being transferred from my PC's graphics card. Currently I'm just using TOSLINK.

You would need a good USB cable (technically good, not expensive) where the + line (voltage, not data) is not connected.

 

[Zek]

You would need a good USB cable (technically good, not expensive) where the + line (voltage, not data) is not connected.

No, + line is not a problem, ground line is main problem.

 

[lockheed]

Possibly some sort sort of grounding issue maybe? Can you connect the PC to the DAC via TOSLINK instead, to remove the electrical connection and see if the noise goes away?

Wondering if the new A20a differs in grounding in some way.

I have the latest version (I believe) of the A20a and mine is silent. However my system is not silent if I connect my DAC to my PC via USB (without the USB isolated dongle thing I have). That isn't because of the A20a. It's some issue with noise being transferred from my PC's graphics card. Currently I'm just using TOSLINK.

Well, obviously the problem is how the A20a v2 handles the noise from the PC carried over the USB cable and TOSLINK would remove it, but it is not an option.

I just tested the silly AudioQuest JitterBug and naturally it did absolutely nothing but maybe slightly changed the pitch of the noise. What a waste of time. People should be prosecuted for selling this voodoo crap.

Ferrite core beads will be tested next and XLR DAC->AMP connection to follow next week.

You would need a good USB cable (technically good, not expensive) where the + line (voltage, not data) is not connected.

No, + line is not a problem, ground line is main problem.

Guys, do tell more about those cables? Is there a way to prepare a USB cable to prevent it from carrying the ground loop noise?

 

[Roland68]

No, + line is not a problem, ground line is main problem.

These are at least 2 different problems that have nothing to do with each other.
What ground wire do you mean? The - wire of the USB power, or the connector ground of the USB connector?

 

[Roland68]

Well, obviously the problem is how the A20a v2 handles the noise from the PC carried over the USB cable and TOSLINK would remove it, but it is not an option.

I just tested the silly AudioQuest JitterBug and naturally it did absolutely nothing but maybe slightly changed the pitch of the noise. What a waste of time. People should be prosecuted for selling this voodoo crap.

Ferrite core beads will be tested next and XLR DAC->AMP connection to follow next week.



Guys, do tell more about those cables? Is there a way to prepare a USB cable to prevent it from carrying the ground loop noise?

The missing + DC line of the USB cable prevents a polluted power supply from the PC from being fed into the DAC.
This only works if the DAC does not need this supply.
The - DC line of the USB cable is required for the DAC to recognize the connection.

 

[lockheed]

The missing + DC line of the USB cable prevents a polluted power supply from the PC from being fed into the DAC.
This only works if the DAC does not need this supply.
The - DC line of the USB cable is required for the DAC to recognize the connection.

Well...
I glued over the + inside the cable plug with an electrical tape and the DAC was not detected by the PC.
Then I tried that with the -/ground and this time DAC was detected AND THE USB NOISE IS GONE! SICK!

 

[Zek]

What ground wire do you mean?

There is no - line, only ground line.

 

[Roland68]

View attachment 187356
There is no - line, only ground line.

I really don't know how to answer something like that.
Haven't you forgotten the shielding of the USB cable and the metal housing of the USB connector that is contacted with it?
They don't create a ground connection between the devices?

 

[Roland68]

Well...
I glued over the + inside the cable plug with an electrical tape and the DAC was not detected by the PC.
Then I tried that with the -/ground and this time DAC was detected AND THE USB NOISE IS GONE! SICK!

In my personal opinion and experience, USB is a sick system anyway. But really cheap and above all much cheaper than Firewire and Thunderbolt.
In addition, most manufacturers do not adhere to these standards (cables, plugs, interfaces, devices) and try to make it even cheaper.
For me it is always amazing how well most device manufacturers have these problems under control and how good the audio quality via USB is now.

Which DAC do you have?
If you're sure you've taped the correct line, your DAC appears to require USB power. However, there seems to be a connection via the connector ground/shield.
An ideal application for a USB isolator (e.g. Intona), but for the money you can also put it in a new DAC that doesn't need something like that.

 

[Moosi]

Interesting, but wrong thread I assume. USB noise is a problem of the DAC and not a problem of the SabajA20a. It cannot do anything about it, since it does not have a USB port. Besides, Amir has tested various USB cables and found small differences (rising with lenght of the cable), but all cables were making noise outside of the audio band.
In general, I would make sure that I do not use USB ports on the back of a computer (where the mainboard is directly attached to) but those on the front of a tower or to use a USB hub. Personally I am using my Monitor conntected to the E30 DAC, since it has a small USB HUB with three ports included. Also make sure you are using a ferrite clip with your USB cable. Most USB cables that come with -or are sold for printers are already shielded with ferrite.

 

[Toku]

Interesting, but wrong thread I assume. USB noise is a problem of the DAC and not a problem of the SabajA20a. It cannot do anything about it, since it does not have a USB port. Besides, Amir has tested various USB cables and found small differences (rising with lenght of the cable), but all cables were making noise outside of the audio band.
In general, I would make sure that I do not use USB ports on the back of a computer (where the mainboard is directly attached to) but those on the front of a tower or to use a USB hub. Personally I am using my Monitor conntected to the E30 DAC, since it has a small USB HUB with three ports included. Also make sure you are using a ferrite clip with your USB cable. Most USB cables that come with -or are sold for printers are already shielded with ferrite.

I think I was the first to find the noise problem on the A20a and report it to Sabaj.
Did you investigate the noise problem on the A20a yourself?

I investigated with two A20a and it has nothing to do with the USB input. Even if all the USB cable and OPT / COAX cable connected to the A20a are removed, the noise will not change. In short, noise is generated even if nothing is connected to the input.
The noise output is quite different between the two A20a units. Even if the input is switched to OPT / COAX, there is a difference in level, but noise continues to appear.

When I was investigating this problem, I found that some products of the same A20a had less noticeable noise. The volume position where noise starts to occur also differs depending on the product.

From these things, I think that the cause of noise is the internal wiring layout.

 

[Moosi]

You are right @Toku I was just referring to Lockheeds post not being related to the A20a for being two different topics:
"Then I tried that with the -/ground and this time DAC was detected AND THE USB NOISE IS GONE! SICK!"

 

[Nozza]

I think I was the first to find the noise problem on the A20a and report it to Sabaj.
Did you investigate the noise problem on the A20a yourself?

I investigated with two A20a and it has nothing to do with the USB input. Even if all the USB cable and OPT / COAX cable connected to the A20a are removed, the noise will not change. In short, noise is generated even if nothing is connected to the input.
The noise output is quite different between the two A20a units. Even if the input is switched to OPT / COAX, there is a difference in level, but noise continues to appear.

When I was investigating this problem, I found that some products of the same A20a had less noticeable noise. The volume position where noise starts to occur also differs depending on the product.

From these things, I think that the cause of noise is the internal wiring layout.

His noise did not appear to be the same problem.
Only occurred when his DAC was connected via USB.

 

[lockheed]

You are right @Toku I was just referring to Lockheeds post not being related to the A20a for being two different topics:
"Then I tried that with the -/ground and this time DAC was detected AND THE USB NOISE IS GONE! SICK!"

If you read my posts carefully, you will note that A20a v1 does not have the USB noise problem, and A20a v2 does. Which means the problem is precisely related to this amp.


Anyway, I tried some tricks you suggested earlier, and the results are indeed surprising. I was actually amazed.
All values are dBs measured with an Android sound meter app on a phone, 20cm from the speaker. Max volume on amps.
Both A20a's had their internal Voltage screw turned to the max.

Background room noise: ~36dB

	Sabaj v1		Sabaj v2
no connection	40.7	high pitch audible whine	40	high pitch audible whine
DAC only	40.7	white noise + quiet low pitch whine	50.8	raspy hum
DAC + USB	40.7	white noise + quiet low pitch whine	50.8	raspy hum
DAC + USB + Toroid ring on 220v cable (4-5 loops)	n/a		38
DAC + USB + Toroid ring + 2-3x ferrite beads per interconnect cable	n/a		34-36	(still audible within 50cm)

 

[Moosi]

Both A20a's had their internal Voltage screw turned to the max.

was it turned to max by default from the manufacturer or did you adjust it to max. by yourself? I believe the default setting is ~23V. Maybe if you have "USB noise" this is related to your DAC being fed from a "dirty" source with a ground loop. If not it should vary from USB plug to USB plug shouldn't it?

Anyways, I do firmly believe the N30 Toroid to the 230V AC-INS (120V AC might be less effective, but still worth a try) is the most effective (and cheapest) mod for this amp. I went even further than that. I insulated the 24V DC-Mains cables going from the PSU to the amp board and I insulated the SPK-cables going from the amp board to the SPK-Panel where the speakers are being plugged to both with a wrappage of copper foil with rubber sleeves on it. Here is a slightshow of it
From all of this I learned that the main source of noise is the proximity between the SPK-panel and the AC-In cables from inside. It is just very close to one another and those AC cables are not properly insulated, they are not even twisted as most amp manufacturers would do with AC lines.
If you can, twist them or at least bend them away from this panel as much as possible. Plus, the Mornsun PSUs may sound decent on paper and claim to be medical, but they are loud as fuck really. If you put your ear to the amp, you can hear it sizzling and buzzing inside the amp. If you can hear that with your ears from outside the case, your speakers will hear it too, for some of it gets picked up by either the SPK cables or this SPK panel and bypasses the MA12070 noise reduction pathways. Swapping it with the Meanwell RPS-400-24 PSU did not significantly reduce the white noise from speakers (not as much as the N30 ring), but did alter the buzzing and sizzling inside the case. There is none anymore. Just a slight hum and no clicks whatsoever, not even one when turned on.
Let's be honest here, this amp is slightly flawed for having a case which is too small to put a 300W+ PSU into it and sandwich it directly underneath the amp board. Other top tier and small size amps such as the Topping PA3&PA5 use an external PSU for a reason. Sabaj (SMSL) probably won't admit, but they know it. They knew it when the made the SMSL DA9 and they knew it when they took the DA9 and updated the PSU to be located at greater distance to the SPK-panel, selling it as the AO200 with better measures (and both have smaller PSUs). As I said, mine is now dead silent with ears hovering 2cm to the tweeter up to 98% of volume and plugged via RCA over a Topping E30 DAC running from galvanically isolated mains powered by a Recom RS3 (9-18V to 5V) fed by a Meanwell Medical 12V adaptor. From -02db I start hearing a hum. Since I'll never play that loudly anyway, I consider this amp now as perfectly silent and with the Meanwell PSU it is as powerful as you can get out of those chips (could do 1,6 [email protected] theoretical max. meaning it could run a passive subwoofer). I'm not a "happy customer" though, I made myself happy with this amp.