• These commercial threads are for private transactions. diyAudio.com provides these forums for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members, use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

DUAL Merus MA12070 fully balanced Amplifier pictures inside

the true Ripple should be 10-20mV which is superb and only beatable by linear powering, which doesn't make sense here, because the DAC is suited for SMPS use.
That is an assumption. Practically every device I tried till now had benefit from a better PSU. Especially with a DAC even when it has switchers internally (which should be replaced ASAP).

What's the assumption? The 10-20mV or that cleaner power doesn't lead to a better SMPS-DAC?

I'm not sure about a SMPS-DAC benefitting from it. But I have tried three different sources.
a)Huawei Tablet adaptor 5V/2A
b)some higher grade 5V/2A adaptor from a medical vaporizier I own
c)a selfmade 15V-5V/0.6A converter from Recom (measured 5.000001V)

I'm very picky. I use to listen to about 5 tracks with very subtle cues and symbols I know very well in order to compare. With the Huawei adaptor I get the feeling that something is a little off. It sounds a subtle bit more dull. The better 5V/2A adaptor from the vape is doing a better job. Yet I feel I had the best clarity and detail from the Recom module . The output is extremely stable and datasheet says 50mV max. but for all models. My assumption is 20-25mV for this thing with my DAC (5V/0,5A). The 5V out of the RPS-400 should be on the same level of low noise, if not better, but it does not have the superb stability of the output (Recom:5.000001V /RPS: 4.9425V). Do you think the stability of the output is a factor here, besides ripple?
 
Last edited:
I'm very picky. I use to listen to about 5 tracks with very subtle cues and symbols I know very well in order to compare. With the Huawei adaptor I get the feeling that something is a little off. It sounds a subtle bit more dull. The better 5V/2A adaptor from the vape is doing a better job. Yet I feel I had the best clarity and detail from the Recom module . The output is extremely stable and datasheet says 50mV max. but for all models. My assumption is 20-25mV for this thing with my DAC (5V/0,5A). The 5V out of the RPS-400 should be on the same level of low noise, if not better, but it does not have the superb stability of the output (Recom:5.000001V /RPS: 4.9425V). Do you think the stability of the output is a factor here, besides ripple?

And why not give GaN (Gallium Nitride) technology a try? There are several GaN chargers out there, I'm curious how it would perform here.

Anker Nano II 65 W Chargeur USB C, GaN II Technologie Charge Rapide pour iPhone 13/12/11/Pro/Mini, Samsung Galaxy S20/S10, iPad Pro, MacBook Pro/Air, Dell XPS 13, Note 20/10+, etc. : Amazon.fr: High-Tech
 
Member
Joined 2008
Paid Member
Both versions of the chip have I2C inputs, for controlling various operational parameters.

Only the -P version accepts I2S input (i.e. digital audio). The non-P version is line-level analog input only.

I used a Raspberry Pi, PulseAudio, and a simple ma12070p board to make a "sound server".

It's a great chip, but working with it physically is kind of challenging - the form factor isn't particularly well-suited to DIY soldering. At least not to me, I need to spend some time improving my SMD soldering skills! I actually made another PCB that used massively overkill components, and it worked for a while, then stopped. I haven't had time to fully debug it, but I think the soldering was marginal from the start, and simply handling it pushed it over the edge. That said, I realized the fab house I've been using (JLCPCB) will do SMD assembly, so I'm thinking about giving that service a test.
 
I was curious about the ma12070 and getting enthusiastic reviews of the Sabaj A20a, I bought it. Also because it has balanced inputs it seems a nice fit for my LX-mini speakers.

I have listened to the A20a now for about 2 weeks. Yes, it does some things right. sturdy and wide stage etc. You hear it has a good power supply.
But it sounds very thin and bright, lack of bass (as others also mentioned) and lack of dynamic, there is no punch. (I like to feel the music, voices. I am not really into "audiophile" sound, as being always everything sounding polite and clean)

Put my modified FX Pro 502 back in. And although it sounds more intimate, having not that large soundstage. I like it much more to listen to. Dark, bass, punch, and very nice highs, good mids.

So I guess I have to take the A20a apart and see what can be optimized.
 
I own the FX Pro 502 based in the TPA3250 chip, it is a pretty old amplifier but works fine. For your info : this is the only TPA3250 finished amplifier with some "air core inductors". )

Again, appreciation remains a matter of taste and probably tied to the paired setup. For me the Sabaj A20a is better in every way )
 
I have listened to the A20a now for about 2 weeks. Yes, it does some things right. sturdy and wide stage etc. You hear it has a good power supply.
But it sounds very thin and bright, lack of bass (as others also mentioned) and lack of dynamic, there is no punch. (I like to feel the music, voices. I am not really into "audiophile" sound, as being always everything sounding polite and clean)
The A20a is transparent, detailed and true to the source. If the track has low bass, you will get low bass. If it has room-shaking bass you will get that. Seriously on bass heavy tracks I feel it's even close to being too much (woofers visibly bumping out of my speakers). The power is definately there, it's just not being shown with every track. I know that there are amps that give you heavy bass on any track regardless, but most of them especially cheaper ones are distorting the sound/EQing a "U" form, raising both ends of the frequency spectrum, that is probably why you sense a lack of "dynamics" with high frequencies too. You can put the A20a in "E3" mode, which is "dynamic bass" that will basically do the same. Or -best- use a subwoofer.
-------------
I finally conducted an operation on the A20a.

[x]Replaced Mornsun PSU with Meanwell RPS-400-24 (@~26V)
[x]Drilled a hole in the back panel, put a cable from the 5Vout of the Meanwell PSU through to feed my DAC from it.
[x]speaker-out-cables, speaker-out-panel and 24v-mains cables have been shielded additionally. (copper foil wrappage with rubber coating)
[x]Rubber feet replaced with much bigger rubber feet. (better ventilation underneath+better looks)












It's too early to draw a full conclusion, but it's very detailed now, probably more detailed than before. I am partially hearing new cues in known tracks and there is more detail/nouance in the bass, but do not overestimate this, because the DAC is running on the 5Vout of the Meanwell now (which is lower ripple than any other 5V adaptor except for linear ones). That probably contributes most to it. Unfortunetely I am also having a ground loop from this 5Vout making the speakers whine/sing <60% volume. Trying to find a solution to it. Other than that, it sounds great, very "clean", "deep" and "black".

[x]Next step is plugging the FAN-out to my DAC to see if there is improvement with the RPS-400 in theory being able to go up to 470W (instead of 290W w/o FAN). Both before hiccup.
 
Last edited:
Nice mod ) Very cool thanks ! :)
Did you insulate the speakers cables from inside ? I just put a Ferrite on it and it is really helpful to avoid any RFI.

Yes, I insulated from inside. I'm not familiar with ferrite. Do you put it to the speaker cables or inside the case?

Today I listened to tracks I know really well and use for comparison because they are overblown with details/cues/echoes. I was really impressed. The A20a sounds a fair bit less compressed, especially when heavy multitone kicks in there seems to be better seperation. It's also more airy/big and more "in your face" with the Meanwell PSU. The cues/symbols seem to "pop" more. Even tho the amp does not play louder and the bass does not sound as if it goes deeper or harder, I get the impression/sensation of being "punched in the chest" on louder volume which I can't really remember having had that notable before. That goes for other tones too to a lesser extent.

-"Tool - chocolate chip trip" was mindblowingly detailed, even the triple/quadruple super fast bass cascades were clean. The cymbal hit at the end stayed/echoed for tremendously long. Never had it that good. *24/96 FLAC
-"Sephira - The 3rd Drink" was really punching my guts out on vol -26. Everything stayed clean and distortion free with great detail and ability to follow every tone. *16/44 FLAC via Roon+HQplayerDSD512

The dynamic load measures we have from the report say:
Full/Min load 50%duty/1KHZ = 258mv
50%duty=200W which is in reality never pulled @1KHZ on loudest volume, probably 1/20 maximum(10W). 258mv/20 = 12,9mv real ripple (assumably even less)
Full/Min load 50%duty/120HZ = 516mv
50%duty=200W which is in reality never pulled @120HZ on loudest volume, probably 1/5 maximum(40W). 516/5 = 103,2mv real ripple (playing light bass, assumably even less)

The MA12070 chips can suppress 70mv ripple. I'm not sure if that doubles using two chips, but I strongly assume not.
If my figures are somewhat halfway correct, there should hardly be any ripple anymore with the Meanwell down to light bass levels with the exception of deep bass (where ripple is not that notable anyway).
For the Mornsun we do not have any measures/reports to compare, but my ears have chosen carefully.

[x]I can recommend swapping the Mornsun with the Meanwell. If you are an adventurer, go for it! If someone is satisfied with the A20a and/or a lazy lad, don't.
 
Last edited:
edit:
This album on vol. -30. Really amazing! The dynamics are incredible, the amp sounds a lot more like a A/B class amplifier. I need to up my recommendation. The more I hear the better it gets. :)
Tropico Paradise Island - The Music | The Daniel Indart Project | LMS Records
*standby power will double with the Meanwell (~4W->8W), for it has no such low-power-relay-switching as the Mornsun. This may contribute also to the better dynamics, since those low-energy-states of mass market PSUs are known to be bad for audio performance.
 
Last edited:
Are you aware that the power caps are only rated for 25V?
The E on these Caps indicate the 25V rating...
Are you sure? Which ones exactly do you mean, do you have a picture?

I have been asking the manufacturer Sabaj and got this answer:
"Yes, we take a lot care about this! Our caps are 50V or above.[..]there is a potentiometer on it to adjust the voltage to close to 26V."
 
Are you sure? Which ones exactly do you mean, do you have a picture?
I have been asking the manufacturer Sabaj and got this answer:
"Yes, we take a lot care about this! Our caps are 50V or above.[..]there is a potentiometer on it to adjust the voltage to close to 26V."
Several 35v here : Sabaj A20a modification... worth the effort?
(never found a lytic rated under 50v sounding as well as 50v and over, even for 3.3v rail...)
 
Last edited: