I just would like to add that not everybody shares this opinion o experiences this amp this way....in my rig this amps actually is more pleasing than a lot of other amps of different topology (and price range)...which is of course also purely subjective....
.....and no, I do not get paid to post this and I am not affilated to the manufacturer either....
Last edited:
To the noise:
I have started modding this amp by wrapping copper foil and a shrink sleeve to the speaker-cables going from the PCB to the SP-out-panel (from where to connect the Speakers). If I would have known about ferrit beads at that point, I would have used those instead. The panel itself got a card board cover too (lighting match box inlay). In addition I have done the same to the 24V-Mains cables.
The white noise from the speakers changed quite a bit after that. It changed from "ant crawling noise/white noise" to a "high pitched hum". The new "hum" is coming in from -15db on the volume and is only hearable if you put your ear to the speaker. The old "ant crawling noise" was coming a little earlier (-22db on) but was also only hearable by putting the ear to the speakers.
After that I swapped the Mornsun PSU with a Meanwell RPS-400-24. While it did improve the sound overall, especially on higher volume (better&cleaner dynamics, more punch), there was no improvement in the white noise. It basically stayed the same as described above. So the problem must not be the Mornsun PSU itself radiating too much, but most likely the speaker cables to the sp-out-panel.
To the big 2200 E-FK caps feeding the Merus chips only being 25V each, I will ask the manufacturer (again) if it's safe to operate this amp @26V.
I have started modding this amp by wrapping copper foil and a shrink sleeve to the speaker-cables going from the PCB to the SP-out-panel (from where to connect the Speakers). If I would have known about ferrit beads at that point, I would have used those instead. The panel itself got a card board cover too (lighting match box inlay). In addition I have done the same to the 24V-Mains cables.
The white noise from the speakers changed quite a bit after that. It changed from "ant crawling noise/white noise" to a "high pitched hum". The new "hum" is coming in from -15db on the volume and is only hearable if you put your ear to the speaker. The old "ant crawling noise" was coming a little earlier (-22db on) but was also only hearable by putting the ear to the speakers.
After that I swapped the Mornsun PSU with a Meanwell RPS-400-24. While it did improve the sound overall, especially on higher volume (better&cleaner dynamics, more punch), there was no improvement in the white noise. It basically stayed the same as described above. So the problem must not be the Mornsun PSU itself radiating too much, but most likely the speaker cables to the sp-out-panel.
To the big 2200 E-FK caps feeding the Merus chips only being 25V each, I will ask the manufacturer (again) if it's safe to operate this amp @26V.
Last edited:
Anybody already figured out which caps are used for the input of the amp and to the MA1207?
The series on these caps say "U2P". The "U" would suggest a Nichicon cap, but I can't find a "U2P" in the portfolio of Nichicon...
Closest is "UUP", but that is a bipolar cap and these are clearly not bipolar.
The series on these caps say "U2P". The "U" would suggest a Nichicon cap, but I can't find a "U2P" in the portfolio of Nichicon...
Closest is "UUP", but that is a bipolar cap and these are clearly not bipolar.
I am familiar with this also from my FX 502s pro (TPA 3250). Same thing there (although never had noise). Suggestions were made in that thread that this comes from the SMPS.
Looked for information about desoldering these SMD Caps. This turn out to be a PITA with the leadfree solder. Best way seems to just push them down and turn, so they just snap off. Of course destroying the cap, but best for the PCB... Through hole is a lot easier...
Last edited:
A first mod to see if 'it's worth the effort', would be to put a larger cap on pin 17 AVVD, as this gave the biggest improvement in the TPA3250 amp (Bass, dynamics etc.).
In that amp I could just solder a 470uF over the 1uF smd cap, but here this cap is ridiculously small (the left one in the circle). I can't imagine this is a 1uF /35V cap... as suggested by Infinion...
In that amp I could just solder a 470uF over the 1uF smd cap, but here this cap is ridiculously small (the left one in the circle). I can't imagine this is a 1uF /35V cap... as suggested by Infinion...
Attachments
Hummm...submiting (bias) or operating a lytic cap at its rated voltage is no good for its lifespan nor regarding distortion/linearity (see Cyril Bateman investigations).
Last edited:
Today Dubai2000 came over for visit to proudly present the matching dac, the Sabaj a20d. Annoyingly the 15 volt output of the amp was not functioning, but with a little bit of improvisation we managed get an external 15 volt power supply working.
After quite some time of not so happy listening we realised, that the Sabaj A20a amp was set on the E2 setting, which resulted in the sound which was described above: thin, lack of bass an a little bit harsh. That explained why I was unhappy with the sound the last two three days.
My amp seems to be a lemon, because it switches automatically to this setting, whenever I enter the menu.
I guess everyone should check whether it is running in E1, the direct mode. It sounds so much better...
After quite some time of not so happy listening we realised, that the Sabaj A20a amp was set on the E2 setting, which resulted in the sound which was described above: thin, lack of bass an a little bit harsh. That explained why I was unhappy with the sound the last two three days.
My amp seems to be a lemon, because it switches automatically to this setting, whenever I enter the menu.
I guess everyone should check whether it is running in E1, the direct mode. It sounds so much better...
Where do I find those Bateman investigations, do you have a link?
I believe as long as you stay below V+7% you should be fine. 26V should be ~4% therefore okay, as long as there is not much heat in the case and as long as the ripple does not get too high. Since I swapped the PSU with a RPS-400-24, there should be notably less ripple coming to those caps, for it's great in those terms. Therefore I should be safe (I figured) using 26V.
https://www.mouser.com/pdfDocs/UCC_ElectrolyticCapacitorTechnicalNotes.pdf
5-6-1 Overvoltage shows 107% WV has small effect and 114% WV much more on lifetime.
Well, had hard time googleing "Cyril Bateman capacitors" led me to the first link : Cyril Bateman's Capacitor Sound articles | Linear Audio NL
This is what I don't understand. The new MA12070 amp from Topping does not seem to have 2200 FK caps at all, but only 470 35V caps, Link, just as the DIY boards you can buy from audiophonics, just as the reference-board from infineon.
Why do they use those instead?
Why do they use those instead?
Bigger caps does not mean better sound... contrary, the more uF's the slower and less involving it seems to get (my personal experience).
With conventional power supplies the challenge is to get the 100/120 Hz ripple as low as possible. To achieve that you needed a lot of uF's.
Now with the switching power supplies delivering very low ripples, and also far outside the audio range, you won't need big caps for that.
You need enough to deliver the peaks drawn by the amp(chip) to keep the voltage stabile. Depends on resistance/inductance also from pcb tracks and wires to the power supply. Then a smaller low ESR cap with high ripple current can be better than a big one.
Thanks, yes X5R probably.
Read just the other week that for audio you should use NPO or COG and avoid other ceramics.
This is well known. Problem is the form factor: a 220nf C0G mesures 3.2*1.6 mm.
Last edited:
Yes, but saw it again last week in a document about components for audio, found in one of the threads here.
Interesting documents by the way, these of Baterman. Instead of using Elna and such you can better use a (cheap) bipolor as blocking cap... (when you don't have place for a film/foil).
Found a 1uF/35V 0603 size COG: GRT188C8YA105KE13J
Still 1.6mm x 0.8mm instead of 1 mm x 0.5 mm...
Must admit, 0402 and then trying to soldering something on it is beyond my soldering capabilities....
Interesting documents by the way, these of Baterman. Instead of using Elna and such you can better use a (cheap) bipolor as blocking cap... (when you don't have place for a film/foil).
Found a 1uF/35V 0603 size COG: GRT188C8YA105KE13J
Still 1.6mm x 0.8mm instead of 1 mm x 0.5 mm...
Must admit, 0402 and then trying to soldering something on it is beyond my soldering capabilities....
Last edited:
Thank you, interesting. But why did they feel the need to go for those big 2200 caps (in all dual merus amps, DA-9, AO200, A20a)? There are 35V 470's down the board feeding the 15V transformer, so I figure they were not running out of them.
Usually one Merus chips has two of them for each polarity. Now with the two chips we should have 4x470uf, but there was probably not enough space on the board, so they went for 2x2200uf instead? But why not choose 2x1000uf, would'nt that have been sufficient already? Maybe the Ripple&Noise coming from the Mornsun LOF350 PSU was relatively high so they had to filter it a little more?
When I was looking for a replacement PSU I was scanning all the report sheets from Meanwell for all SDR,HRP,EPP/RPS-300, -400, -500 Series PSUs for measured Ripple (dynamics 1KHZ/120HZ) and that for each 18V,24V,27V models. What I found was that they differed pretty vastly, even within the same model line, even with different V-output.
1khz / 120Hz (50% duty min-max)
LOF350-20B24 - ? / ?
EPP/RPS-300-24 - 590mv / 542mv
EPP/RPS-400-24 - 258mv / 516mv [x]replaced with
EPP/RPS-500-24 - 655mv / 780mv
EPP/RPS-300-27 - 198mv / 231mv
EPP/RPS-400-27 - 442mv / 556mv
EPP/RPS-500-27 - 595mv / 1220mv
HRP-300N-24 - 233mv / 283mv [x]with ~250% peak ability(650W+) for 3 seconds, great new PSU for audio, but enclosed type with FAN only.
Last edited: