Replace Mornsun 350W PSU of Infineon MA12070 AMP

[SoniCCurrent]

Hello everybody.

I own a "SabajA20a" amplifier. This amp has a "Mornsun LOF350-20B24V PSU" capable of 200W/350W peak (14.6A max) running a PCB with two Infineon MA12070 ICs with 24V. Unfortunately they built the device in a sandwich configuration. The case is very narrow. If I put my ear to it, I can hear the PSU "buzzing/summing" inside the case. Im sure there is some EMI. If I turn the volume up louder than ~80% I get "white noise" from my speakers. Im sure this is the same noise I am hearing when I put my ear to the case.

I want to replace this PSU with a better one that is to be installed in a second case which I will put next to the amp.

Plan:

• PSU-Case : HIFI 2000 GALAXY GX147 Aluminium Case 124x170x40mm 3mm Front Panel Black - Audiophonics
• Preferred new PSU : SMPS300(RE,RS) :SMPS300RS Module d'Alimentation a Decoupage 300W / 24V - Audiophonics
• or
• Meanwell RPS-400 : RPS-400-MEAN WELL Switching Power Supply Manufacturer

Questions arose so far and promise to keep on arising. I'm in hope to answer some of them with your help.

1. Is the default Mornsun LOF350-20B24V PSU" a "single rail" or "multi rail" PSU? And do I even need to bother?
2. Which one of the SMPS300 versions (RE, RS, RxE) would be best suitable to power my amp? F.e. the RS is "single rail" and has no GND on the output side, whereas the RE version has a GND at output. Both have GND at the input side (Schuko). Do I need a GND at output? (The Mornsun-PSU I believe has none).
3. Will a more powerful PSU give me more power/headroom with the two MA12070 chips? In theory their max output in Ampere is 8A per chip, so I could go 16A, but the Mornsun does only reach 14.6A.

• SMPS300RS
• SMPS300RE
• SMPS300RhE

Pictures of open amp:

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

 

[SoniCCurrent]

On Wiki it says about "BTL" configs:

Bridged and paralleled amplifiers - Wikipedia

• A bridge-tied load (BTL), also known as bridged transformerless and bridged mono.[...]
  [*]Bridging an amplifier increases the power that can be supplied to one loudspeaker, but it does not increase the amplifier's total available power. Because a bridge amplifier operates in mono mode, a second identical amplifier is required for stereo operation. For bridged amplifiers, damping factor is cut in half. Because the amplifier's bridged output is floating, it should never be grounded or it may damage the amplifier.[5]

What the SabajA20a has, I believe is "PBTL".

• "A bridge-parallel amplifier topology is a hierarchical combination of the bridged and paralleled amplifier topologies, with at least four single-ended channels needed to produce one bridge-parallel channel. The two topologies complement each other in that the bridging allows for higher voltage output and the paralleling provides the current handling capacity needed to drive low impedances, typical in subwoofer applications. It is more commonly used with IC power amplifiers with stereo outputs operating in bridged mode."

I guess this sentence:
Because the amplifier's bridged output is floating, it should never be grounded or it may damage the amplifier. would therefore apply to "PBTL" the very same way as to "BTL".

From that I figure that I need a PSU without a "GND" at the output stage. Therefore I figure this means I would have to pick a "single rail" PSU.

Am I right here?

 

[SoniCCurrent]

I am currently considering a Meanwell "HRP-300N-24" PSU.

Those "n" rated ones have a up to 280% peak current ability, which to me sounds perfectly suited for any amplifier. The 300N can do 336W/14A regularly with peaks up to ~900W for up to 5 seconds. This is mindbending. 😱

Isn't this perfect from amp perspective? Why can I not find anything in the forum about these? Are there severe drawbacks with these?

 

[jean-paul]

Now I am only sideways informed about the Sabaj A20a but I know that the first ones had bad PSU's. These have been changed to another type. Maybe it is a good idea to ask the seller for warranty or to address Sabaj directly. I believe only a small number of bad units were produced. Considering the spectacular reputation of the A20a maybe they are helpful to solve your issue.

The Mornsun PSU is a single voltage PSU. As it can deliver 200W continuously and 350W peak it has more than enough power in all circumstances* (calculate what current is consumed at high volume use and see the reserve of the PSU). Your best bet may not be an external PSU but the same type Sabaj choose to solve the issues.

* please note the MA12070 delivers 35W to 8 Ohm at 1% distortion. In PBTL that is 120W to 2Ohm at 1%. Now the chance that you play louder than 10W per channel continuously is nil. Calculate with 10W output. Now it is also good to design for maximum specified power when powering a load bank but be assured that your ears won't tolerate more than 1W continuously 😀

 

[SoniCCurrent]

Now I am only sideways informed about the Sabaj A20a but I know that the first ones had bad PSU's. These have been changed to another type. Maybe it is a good idea to ask the seller for warranty or to address Sabaj directly.
The Mornsun PSU is a single voltage PSU. As it can deliver 200W continuously and 350W peak it has more than enough power in all circumstances* (calculate what current is consumed at high volume use and see the reserve of the PSU). Your best bet may not be an external PSU but the same type Sabaj choose to solve the issues.

I know this. Thank you, voisin. I don't want to replace it with another one. Since I have to upen it up anyways I am looking for an upgrade PSU. However, I am aware that there is at best minor to gain.
Several reasons:

a) With a new PSU I can run the amp @full 26V (Mornsun can only do 25.2 max) which is recommended by Infineon. Sabaj runs the amp @~23V by default setting. I don't like to know limitations of my gear. Doesn't matter if I can hear a difference. It's a psychological thing.

b) There is white noise coming from my speakers if I turn it louder than about 75%. When I put my ear to the amp, I can hear electrical buzzing (not whining) from within. I suspect since the case if very small and the PSU is sandwiched right underneath the circuits (not much distance to speaker wires too) the noise I can hear from the speakers is the same buzzing noise I can hear from within the case, but picked up and amplified somehow. This is why I want the PSU out of the case. I have no trouble setting up a second case near it. I like cases. They're awesome.

c) The Mornsun ist a good PSU, but there are surely better ones. The Meanwell RPS-500-24 (medical) or HRP-300N-24 seem to be the best alternatives. I am also looking into a SDR-480-24, which has great Ripple&Noise, but I don't like the "rise up time" being 10x that of other PSUs.

Load Regulation:

• RPS-500-24 : 0,08%
• HRP-300N-24 : 0,06%
• SDR-480-24 : 0,55% (Problem?)

Rise Time (max):

• RPS-500-24 : 9.7ms
• HRP-300N-24 : 6.1ms
• SDR-480-24 : 112ms (Problem?)

Dynamic Load - FULL /50% LOAD 50%DUTY / 1KHZ:

• RPS-500-24 : 655mVp-p
• HRP-300N-24 : 233mVp-p (very good)
• SDR-480-24 : 337mVp-p (FULL /Min LOAD
  [*] 90%DUTY/1KHZ)
  

Leakage Current:

• RPS-500-24 : 181.2uA/264VAC (perfect)
• HRP-300N-24 : 1.147mA (Problem?)
• SDR-480-24 : 0.510mA

please note the MA12070 delivers 35W to 8 Ohm at 1% distortion. In PBTL that is 120W to 2Ohm at 1%. Now the chance that you play louder than 10W per channel continuously is nil. Calculate with 10W output.

Please read this post from our voisin here: https://www.diyaudio.com/forums/power-supplies/328306-suitability-power-supply-audio.html#post5564905

• This example is simplified, and looks exaggerated, but reality can be even worse because of reactive currents.
  I gave a class B example, but except for some class A, the problem is the same for all types of amplifiers, including class D: no amplifier has an energy storage capability (not enough for 20Hz anyway)
  [*]The problem arises because the supply's protection kicks in very quickly when dangerous levels are exceeded, which is normal for a regular supply.
  You could think that adding energy storage, in the form of a large parallel cap for example could solve the problem. Not so: the regulation is extremely tight, meaning the internal resistance is small, and the capacitor would need to be really huge to make a difference, but much before that, it will cause problem during the start-up phase.
  Audio supplies are designed taking into account these constraints, but if you use a regular supply in this role, it needs a severe derating: typically a 10W amp would need at least a 50W GP supply.
  

Since the amp can do 120W into 2ohm I would have to calculate 120x5 in order to be totally sure not to have any bottlenecks playing loudly. So I would need a PSU that can do short peaks up to 600W.

 

[russc]

Why don't you try mounting the original psu remotely?
A lash-up would prove or disprove your hypothesis and cost nothing.
Then you could proceed with a permanent fix on the basis of the results.

 

[SoniCCurrent]

Why don't you try mounting the original psu remotely?

I need to change it anyway, since it has problems. Either with a new (same) one, or another (better) one.

 

[jean-paul]

Sabaj choose a better one but then... which one? Maybe you could ask the members that have the Sabaj as seen in other threads?

Shortest path from A to B and optical best solution is the new PSU (internally mounted).... but you could also invest time in improving the Mornsun since you already have it. If you work safely you could try a metal shield screwed with small M3 brackets to the backside and some plastic standoffs for safety. The metal plate should be connected to PE! Is it EMI what is causing the trouble? Then the metal plate will make a difference.

Your power calculations are grossly exaggerated and totally unrealistic, sorry. The original PSU has power in abundance but the issue is the whining. Here you can find the reason WHY the PSU whines. Then maybe you can solve that in the Mornsun and just use the device.

Dual Merus MA12070 Amplifier Sabaj A20 fully balanced : pictures inside ) | Page 7 | Audio Science Review (ASR) Forum

 

[SoniCCurrent]

Shortest path from A to B and optical best solution is the new PSU (internally mounted).... but you could also invest time in improving the Mornsun since you already have it.

I will buy a Meanwell RPS-400-24, replace the PSU in the case and run it @26.13V. I will plug my DAC from the 5V/1A auxiliary port too, since it's only 30mv Ripple/1A.

If you work safely you could try a metal shield screwed with small M3 brackets to the backside and some plastic standoffs for safety. The metal plate should be connected to PE! Is it EMI what is causing the trouble? Then the metal plate will make a difference.

New ball game with the RPS-400 in terms of EMI. It is supposedly much better at it. Don't know if a metal plate will be nessessary with it. But I will buy some copper foil and extra shield those mains-cables (except for ground). Look at the mains-wires. They are not insulated at all and they are located very close to the speaker-outs.


GND is different with the Meanwell. It's a 2-pin with GND via M3 screw, whereas the Mornsun has a GND as a third pin in the terminal. Is there a difference in EMI?

Your power calculations are grossly exaggerated and totally unrealistic, sorry. The original PSU has power in abundance but the issue is the whining.

I have no whining! I have buzzing which is most probably EMI. The SabajA20a can output 416W/26V@16A max. With the RPS-400-24 im safe even with the hardest loads. It can go 470W before hiccup mode. The Mornsun can "only" do 385W before hiccup. Not that it matters much, but I don't to know limitations of my gear and besides the Meanwell is better in about every other aspect too that may or may not matter in the end. (rise time, leakage current, emi, load regulation, Ripple&Noise)

 

[jean-paul]

Hi I think you will have to find out what the PE connection is. And how that is arranged with the Mornsun. If low voltage DC GND is floating I mean. If not you'll probably have a ground loop. Generally you don't want PE at the secondary side. It is used for safety and for the filtering. Please use the terms GND/ground/PE for the right items!

The 30 mV ripple is too much for a DAC. That comes when using switchers. You could build a linear PSU with µV numbers. My own PSU's have below 5 µV ripple/noise easily. If you are after numbers, that is a factor 6000 less ripple /noise...... it makes certainly sense with such low voltages for sensitive equipment. Here clean voltage matters most but using that output is cheapest so probably it will be used by most. It has nothing to do with quality though.

Whining, humming, buzzing ...depends on what one calls it. In my language there is a clear difference when it is about low/lowmid/higher tones and if they are wobbling. Best is a metal plate connected to PE as it is most adequate for shielding when it is radiated garbage which comes with many switchers. It is not the easy/cheap method.

You mention "hardest loads". If you really would use it to the max it won't last an hour before it melts 🙂

 

[SoniCCurrent]

Hi I think you will have to find out what the PE connection is. And how that is arranged with the Mornsun. If low voltage DC GND is floating I mean. If not you'll probably have a ground loop. Generally you don't want PE at the secondary side. It is used for safety and for the filtering. Please use the terms GND/ground/PE for the right items!

Sorry, I am novice to electrical engerineering. I will upgrade the device myself with the knowledge I can get (from this forum and other sources) and have it checked by a real electrician before I use it. This way I can learn a lot without risking my life. So there is a GND (that goes into Schuko) from input. I assume that is for safety. Then there is a PE which is supposed to be grounding small floating currents that may (or may not) appear. This one is supposed to be connected to the case itself or a metal plate installed in the case, right?

Some SMPS have a third GND-pin also for the output-side, but the Mornsun and Meanwell does not. Therefore I would only have to connect the Schuko-GND and the PE connection. If this is wrong, please correct.

The 30 mV ripple is too much for a DAC. That comes when using switchers. You could build a linear PSU with µV numbers.

The DAC im using is a Topping E30. This DAC has internal switchers and filters and is designed for SMPS use with a 5V/1A adaptor. Currently I am using it with a DC-DC 18->5V coverter that has the specs 5V/0.6A@50mv Ripple. I don't hear any noise. I will report if going down from 0.6/50mv to 1/30mv will be benefitial to my ears. I doubt it tho.
The review/measures of the E30 DAC on audiosciencereview.com were taken with the dirtiest power source there is ->A 5V USB plug from a PC. It measured 115db SNR or something with multitone down to 130s. I don't think I would benefit from a linear PSU here, at least not with this DAC.

Whining, humming, buzzing ...depends on what one calls it.

To be more precise: If I turn the amp louder than 70% (no sound playing) I can start hearing some crackling/ant crawling noise from my speakers, but only if I put my ear to it. Turning it louder from there amplifies that "white noise".
If I put my ear to the amp I can also hear a slight crackling/buzzing sound which is some sort of stream of voltage/current noise. It's a power-buzzing sound, like to one you would hear putting your ear to a transformer-case. My suspicion is that this is the same noise I am hearing amplified from the speakers from a certain level (~70%) on. As I said, the case is very small. The PSU is sandwiched underneath the circuit. Not much distance to speaker wires. Mains-cables and speaker-out-cables are not insulated other than plastic covers and are hanging loosely in the air. I am convinced there is some benefits from going to a higher tier medical PSU such as a Meanwell, which has a severely better EMI-rating (CISPR11 vs CISPR32) plus insulating and bonding the wires to the case (more away from speaker-plugs f.e.).

Best is a metal plate connected to PE as it is most adequate for shielding when it is radiated garbage which comes with many switchers. It is not the easy/cheap method.

Since I'm an amateur to electricity I love keeping things simple. Metal plates and PSUs that can go as high a 20A is a combination that scares me. Even a Meanwell-medical can fail. I like fried chicken. I don't want to become one.

You mention "hardest loads". If you really would use it to the max it won't last an hour before it melts 🙂

I know. It's the hold up time right? The time a SMPS can hold up any current.
----
pics:
PCB power input & power out to speakers (-L+L/-R+R):
https://ae01.alicdn.com/kf/H4067360150d746c3a64ce230b256a0d3W.jpg
----

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

IMG-20210511-185742 — ImgBB
https://ibb.co/wr42bzZ
https://ibb.co/9pM85kJ

 

[jean-paul]

If you are a novice this job is already too dangerous. Sorry. You say you like to keep things simple (we all want that but life is different) but on the other hand you want ultimate performance from a 200 $ device. Now, shielding IS simple and mandatory against EMI. It costs a few cents therefor it is often not done. Take seventies analog equipment and inspect how things are done. All is shielded extremely well for best performance. If it would be my device I would see the challenge in making the device as silent as possible with the Mornsun PSU after measuring if the output voltage is clean. I haven't put time in it but the details regarding the relay in the Mornsun would have my attention.

There are some contradictions in all this. It is all fine but it really is better to know what one is doing and how stuff works certainly when working with mains connected stuff. Best performance is only achieved when all areas are executed optimally. Not by focussing on a single item and numbers from paper.

When persevering... please make sure what GND and PE is and if the Mornsun has a connection between GND and PE (this is important). That 20A won't do a thing when one uses load resistors (never connect speakers) and a safe distance when testing. A normal house has an installation that will break the current when a mistake is made. Best tips: work tidy, go for he best, never work on equipment when energized and don't make mistakes 🙂 And if you make them well then you learn from it.

All being said and done, safest and simplest is to address to the seller or Sabaj and get a tested and proven to be OK replacement PSU as this one clearly hampers performance. Chances are you will receive the new PSU for free and then an A20A that is silent.

 

[SoniCCurrent]

If you are a novice this job is already too dangerous. Sorry. You say you like to keep things simple (we all want that but life is different) but on the other hand you want ultimate performance from a 200 $ device.

By "simple" I didn't mean to keep things dangerous or be sloppy. I will test the chassis or any other part with a Multimeter before I touch anything. I will also not alter anything with powers connected and I will wait 5minutes after unpowering before touching with multimeter and/or with hands. I'm very careful. I will buy a metal plate if nessessary and install it if you guide me.

Now, shielding IS simple and mandatory against EMI. It costs a few cents therefor it is often not done. Take seventies analog equipment and inspect how things are done. All is shielded extremely well for best performance. If it would be my device I would see the challenge in making the device as silent as possible with the Mornsun PSU after measuring if the output voltage is clean.

I don't want to wait a couple of weeks for a new PSU with a new relay from China. I got the amp very cheap, even cheaper than it sells normally. I have no problem buying a Meanwell PSU to replace it, since I'll have to replace the PSU anyway (they'll only send a new PSU, not a new amp). It's still a bargain and If I can pull out 10% more out of it with a better PSU I'm happy. This amp is powering a desktop system, not a concert hall. I am very happy with it sound-wise. I just like it to be silent up to 100% vol. if possible and to have about 10-15% more power for my speakers are very demanding (83db sensitivity with long dips to 3.5ohm). Both achieveable I believe.

When persevering... please make sure what GND and PE is and if the Mornsun has a connection between GND and PE (this is important). That 20A won't do a thing when one uses load resistors (never connect speakers) and a safe distance when testing. A normal house has an installation that will break the current when a mistake is made. Best tips: work tidy, go for he best, never work on equipment when energized and don't make mistakes 🙂 And if you make them well then you learn from it.

Thank you for the advice. I learned a lot already. So there is Class I and Class II. I strongly suppose since a Schuko is involved we're talking about a Class I system.

The Meanwell RPS-Series is supposed to be frame grounded the same way as the Mornsun LOF-Series (3 screws FG) with the only difference being the Mornsun having the Schuko-PE from a third pin whereas the Meanwell I suppose has the Schuko-PE connected to one of the M3 screws I suppose. Can you see from the pictures where the PE is on the RPS-400?
Here are some good pictures of it: Mean Well Medical Grade Computer Power Supply Open Frame Rps-400-24 400w 24v Output Pcb Type 80-264v Input With 2 X Mopp - Buy Medical Grade Computer Power Supply,Mean Well Medical Power Supply,Medical Power Supply Open Frame Product on Alibaba.com

This is the Mornsun:

An externally hosted image should be here but it was not working when we last tested it.

Meanwell:
RPS-400-MEAN WELL Switching Power Supply Manufacturer

So that means those 3xM3 screws on the Mornsun same as the Meanwell will have to be frame grounded. What I don't understand is, from the pictures it looks as if there is no metal plate whatsoever underneath the PSU in the case, but only a "blank(?)" platina-board on which the metal spacers rest upon (Picture). Or are there conductive rails inlayed in the board that you cannot see on the pics (I haven't opened it yet by myself, those are not my pics)? Maybe they haven't FG'd this PSU at all, only PE'd. Could this be possible?

It is the leakage current that kills potentially if not properly PE'd, right? And isn't the leakage current so low on those medical Meanwells anyway that you wouldn't harm yourself other than a slight tickling? Touch current <70 uA/264VAC. That's like nothing to humans, right?

 

[jean-paul]

Way too much detail for now. I am just asking to measure if the "-" (often called GND) of the Mornsun is connected to PE at the mains side of matters (despite the documentation). You have "+" and "-" at the output. I understand the mounting holes need to be connected to PE.

There are open frame SMPS that have that and there are that don't. Now we don't "suppose" and "assume" but we measure. So out with that DMM and measure between "-" and PE. From what I gather from the datasheet "-" is isolated.

Cheap is nice but it does not guarantee success of the actions nor is it a parameter of significant influence on performance. You don't want to wait weeks but now I suppose that the manufacturer has educated and skilled engineers and that they made the right choices correcting the error of the previous PSU. So, one can assume good results contrary to own interpretations. Now if they are willing to send one for free the efficiency is highest as well. Classic win-win.

 

[SoniCCurrent]

There are open frame SMPS that have that and there are that don't. Now we don't "suppose" and "assume" but we measure. So out with that DMM and measure between "-" and PE. From what I gather from the datasheet "-" is isolated.

Thank you. Unfortunately the only multimeter I have available right now is this one: Link

It is really cheap. I'm afraid it might blow up or something. I will buy a better one which is 10A proof today from amazon. I will conduct the measurement as soon as I have the better multimeter. For now we will have to assume the datasheet is correct.
What implications would arise from that, assuming the "-" on the Mornsun is "isolated" as you say? Would that make replacement with the Meanwell more complicated due to vastly different grounding techniques or less complicated? In my naive thinking I would just have to unscrew->swap PSUs->rescrew with a metal ring soldered to the yellow/green Schuko-PE-cable, resting between one of the screws (which one we will figure out).

Cheap is nice but it does not guarantee success of the actions nor is it a parameter of significant influence on performance. You don't want to wait weeks but now I suppose that the manufacturer has educated and skilled engineers and that they made the right choices correcting the error of the previous PSU.

The problem with the old PSUs is they shipped the first batch of amps with the relay of the PSU is not being suited for music playback. Playing bass-heavy tracks it starts switching wildly and audibly (ticking noise) because the low hertz spikes trigger the energy saving function. It needs a new relay set to make those switches time dependend not load depended.
I don't think they thought of the PSU having a problem with buzzing/EMI that needs to be fixed. They know this is a budget amp and the case is too small actually to be efficient in those terms. Hence waiting for a new PSU will most probably not do anything to better the noise from the speakers @>70% , but only solve the issue with the relay clicking. Now you might ask "What if the Meanwell also has a relay that is load dependend and starts clicking on bass heavy tracks". The answer is: I don't know. But it has been used by a couple of users from this forum in various DIY amps (TP3255) and they seemed really happy with it. If anything fails I will return it.

 

[jean-paul]

You can't blow up anything when measuring resistance/continuity between "-" and PE (in de-energized state with the plug pulled!!!!). Not with a 10 Euro DMM nor with an antique analog meter nor with a high end bench DMM. In fact this is the least dangerous of this whole operation AND you should master this simple measurement to do anything electrical 🙂 Please forget the "assume" thinking, we don't assume anything in electrical/electronic stuff either for safety or for preventing making stuff kaputt! Electrons are our tiny little friends but since we can't see them they can bite us when assuming things. Assumption is the mother of f*****s and only can be done in the world of alphas where exact answers do not exist.

Why do we want to know this ? Well, having an audio device with "-" connected to PE will likely introduce a ground loop which makes amplifiers/sources humming. When the Mornsun lacks a direct connection the advice is not to replace it for an SMPS with platinum stellar audio properties able to deliver hundreds of clean Amperes that does have a direct connection between "-" and PE. Simple.

BTW I just bought a Sabaj A20a myself so if I am lucky I have the same trouble with the Mornsun as you. Since I bought it locally I think it will be here the day after tomorrow. No promises, no hurry, no expectations.

 

[SoniCCurrent]

You can't blow up anything when measuring resistance/continuity between "-" and PE (in de-energized state with the plug pulled!!!!).

Alright. I thought I should measure it being powered, that's why I wanted a 10A proof DMM. In my electronically naive mind there is no power to measure without power. But there is power everywhere. Even my brain is powered. Just not by the power of knowledge. At least not when it comes to electrons.

Not with a 10 Euro DMM nor with an antique analog meter nor with a high end bench DMM. In fact this is the least dangerous of this whole operation AND you should master this simple measurement to do anything electrical 🙂

I'll plug it off and wait 5mins, screw the back panel open, let the beautiful sandwich come sliding right at me, turn the DMM to the small red banana plug symbol, put the black pin to the "- out" and the red pin to the small hole of the terminal block making physical contact to the crimp holding the green/yellow PE-Schuko cable. If the displays shows "~0" they are connected, if it shows "1" they are not. No worries, no kaputtings.

Please forget the "assume" thinking[..]

That is just in order to gain information about what's possibly coming next. When I was a late teenager I opened up the PSU of my Computer to clean it from dust (plugged off, but without waiting 5mins) and touched the copper wires of the transformer or something and got my fingers warmed up quite a bit quite fast, but it felt actually kind of good. Like a warm, invigorating grip from a very strong hand. I wonder how much current that was. But I will not f*** around with this hand anymore. Since we're still in Covid times, it is preferred not to shake hands anyways.

Why do we want to know this ? Well, having an audio device with "-" connected to PE will likely introduce a ground loop which makes amplifiers/sources humming. When the Mornsun lacks a direct connection the advice is not to replace it for an SMPS with platinum stellar audio properties able to deliver hundreds of clean Amperes that does have a direct connection between "-" and PE. Simple.

How do we know if the Meanwell has a direct connection when I'm only allowed to measure, but need affirmation for it to be suitable in order to buy it in the first place.

BTW I just bought a Sabaj A20a myself so if I am lucky I have the same trouble with the Mornsun as you. Since I bought it locally I think it will be here the day after tomorrow. No promises, no hurry, no expectations.

Thanks. And congratulations on the amp. I doubt you will have the relay problem. Those old amps are not being shipped anymore, at least not if you bought from hifi-express or amazon. Please report if you have white noise from the speakers too from ~70% up.

 

[jean-paul]

You'll be measuring resistance, that is in Ohms/kOhms etc. No power needed for that except the battery in the DMM. No fear for shock when the PSU has had time to discharge but measuring voltage makes sense to be sure so do that. You'll be measuring between the "-" of the 24V side of the PSU and the PE pin of that PSU NOT the "- out" of the speakers. They say here a man has not lived until he has touched 3 phase power (DON'T do it). It was told to me by a man here that is unique as he got a 10 kV shock and still has all limbs (and breathes).

No kaputtings, good philosophy. Boycott assumption with electrical stuff forever.

First the Mornsun then further thinking. Being a man of long posts myself (it takes one to know one) I suggest somewhat shorter posts and one item at the time.

 

[SoniCCurrent]

Update/Measurements

This night I have conducted an operation.

1. Total disassembly of the SabajA20a
2. Inductive measurements conducted as demanded
3. Insulated every power-wire, mains-wires and speaker-wires with a wrappage of copper foil. Speakers-wires got an additional coating with heat shrink tube. 24V-Mains-wires also got additional heat shrink tube coating.
4. Around the Schuko-plug and around the speaker-outs-panel I have build each small "cages" of copper foil. The ends of the speaker cables which are soldered into the panel were not insulated at all. They might have picked up some noise there.
5. Adjusting V-output to max. via screw-nob on the PSU.
6. Replacing the tiny round rubber feet with bigger cubic rubber feet. Elevating the case more, giving it better airflow underneath. Also looks more "high end". 😛

1.
Measurements:

2.
Before operation "copper"

3.
Starting with the speaker outs, L&R each seperated.

4.
Adding the heat shrink tube and building the cage for the speaker-out-panel.

5.
Copperfoil+heat shrink tube coating for the PSU-to-PCB Mains.

6.
Now the AC/230V Mains getting a copper cage.

7.
Important to note, both single screw holders for both PCBs are connected to "-" GND and show Inductive ~0004. Same goes for both screws of course.

8.
Last picture before reassambly

... Now I am sitting here, afraid to power it. Please give permission if this is okay to do in terms of safety. (I have made sure for no part of the copper foil to touch any open lines.)

 

[jean-paul]

Resistance.... it is resistance/continuity that you measured. It looks like you did it carefully but I wouldn't have taped mains wiring with copper tape even if it would make the amplifier perform best. The metal plate bent in the right form is safer and more easy to remove when servicing. What happens if the copper tape comes loose?

If you have measured continuity/resistance and nothing is shorted things can be powered on. Safe distance, power switch to on, 2 small resistors 10 Ohm as loads, plug in your hand. Then plug in the plug and see what happens. There is no other way I am afraid. Wait a few minutes, measure output voltage.

BTW the faston connectors for the mains voltage of my Sabaj A20-a did not have enough pressure on the contacts. I had to correct that with a pliers.