Connex seem to be difficult to find in Europe and can be quite expensive. The Mean Well Power Supplies on the other hand are very easy to find and quite often available on eBay either used or as good as new.
The Mean Well Power Supplies are very common in multiple manufacturers equipment and can often be found as excess stock or as spare parts.
I bought a load of Digipos 24 volt 250 Watt Power Supplies for less than £10 each. They are Mean Well SP-240-24 in a nice plastic case. Adjustable voltage up to about 30 volts. You just have to disable the remote control feature unless of course you want to use it to turn on the Power Supply remotely.
Does anyone have suggestions or insight on PSU that could be sourced in the EU for two 3E amp's (3255 2ch + 3251 4ch)?
The time to ship something right now is ridiculous, a month within EU seems to be the norm right now.
They are all under the power that the Amp can deliver, but I'm guessing it will be enough for me.
Also, I don't have any ground connection in my house, if some are more sensitive towards not being grounded.
If you have any other suggestions just fire away.
I have looked at Connex, MeanWell and possibly Hypex.
For the 3255:
MeanWell MEAN WELL SE-600-48 Switching Power Supply SMPS 600W 48V 12,5A - Audiophonics
Connex SMPS500QRV2 Switching Power Supply Module 500W / +/- 40V - Audiophonics
Hypex (though I'm not sure if it would work) HYPEX SMPS400 / 180 Switching Power Supply Module 400W / 2x46V - Audiophonics
For the 3251 (24-30V) two different Connex:
CONNEX SMPS300REh Power supply Module 300W +/-30V - Audiophonics
SMPS300RS Module d'Alimentation a Decoupage 300W / 32V - Audiophonics
Some of the Power Supplies you have listed are dual rail and not suitable for TPA325x amplifiers that use a single rail.
Unless you are going to drive all your amplifiers flat out then you probably don’t need such high capacity supplies. Also the heat sinks on most of the TPA325x boards are too small to run at the full specification for any period of time !!
I was aware that it could be a problem with dual rail, must have missed it with the ones I posted. 
The 'problem' with the Meanwell that's lower output seem to be that they either have 24V or 36V, but they say that it's 10% adjustable so 24V should work for the 3251. The ones with less Watt also have a higher R&N value, but maybe it's of low consequence going from 150mV to 200-250mV...
Or is it possible to use something in-between the PSU and Amp to clean up the signal, capacitor perhaps..?
The MW RSP 500W in 27V and 48V is what I'm looking at right now.
(Yes, I'm new at this. lol )
The 'problem' with the Meanwell that's lower output seem to be that they either have 24V or 36V, but they say that it's 10% adjustable so 24V should work for the 3251. The ones with less Watt also have a higher R&N value, but maybe it's of low consequence going from 150mV to 200-250mV...
Or is it possible to use something in-between the PSU and Amp to clean up the signal, capacitor perhaps..?
The MW RSP 500W in 27V and 48V is what I'm looking at right now.
(Yes, I'm new at this. lol )
Get the SMPS300RS. I just ordered things form Audiophonics two weeks ago and it took 1 week form France. I use a 36V version.
//
Aha - sorry!
But if you think you can live with 35V you can get the 32V which seem to be in stock..
From Connexion page:
"The stock available output voltages are adjustable within approx. 10% and the ordered voltage value is at the middle of the interval. For example the 24V version can be adjusted from 22.5V to 26V."
For a 32V it would mean 35V.
Higher voltage keeps the amp delivering high output in higher impedance speakers. If you are on say 4 ohm, the lower voltage ones are fine. Also less heat.
//
But if you think you can live with 35V you can get the 32V which seem to be in stock..
From Connexion page:
"The stock available output voltages are adjustable within approx. 10% and the ordered voltage value is at the middle of the interval. For example the 24V version can be adjusted from 22.5V to 26V."
For a 32V it would mean 35V.
Higher voltage keeps the amp delivering high output in higher impedance speakers. If you are on say 4 ohm, the lower voltage ones are fine. Also less heat.
//
Last edited:
<
Please stop to promote!
There is an forum for that.
And do you realy think we regular people hear difference at yours
and Mean Well which are very good and little noise even with 200uV ripple?
I dont.
Maybey there is some small 3rd harmomic that differs or other stuff.
Do we hear that?
But there is an big market I understand to sell stuff that is best but in DIY it must be cheap I think.
But people take it easy on all these mods and those who running a game as company.
I am rather amazed of the backlash that any kind of measurements brings.
The discussion was about psu and I will make ours available for free to a few members to test on different Tpa3255 boards.
Tpa3255 has almost no Psrr, so you are listening direct to psu noise. Some of the post say that you won't hear 200uv vs 1uv. In fact the meanwell supply has 150mv = 150000 uv vs 1uv. Do you think you can hear that Solve?
If you want to ignore science and measurements, that is your right. We have the right to be here and share our mods and test data. For example, drmordor now is using bigger capacitance on avdd since we last shared our test results. Everyone wins.
Please stop imposing " cheap" and no test data, because you think you don't hear it. We all like the tpa3255 because TI engineers improved the measurements compared with tpa31xx.
Can you hear that?
The discussion was about psu and I will make ours available for free to a few members to test on different Tpa3255 boards.
Tpa3255 has almost no Psrr, so you are listening direct to psu noise. Some of the post say that you won't hear 200uv vs 1uv. In fact the meanwell supply has 150mv = 150000 uv vs 1uv. Do you think you can hear that Solve?
If you want to ignore science and measurements, that is your right. We have the right to be here and share our mods and test data. For example, drmordor now is using bigger capacitance on avdd since we last shared our test results. Everyone wins.
Please stop imposing " cheap" and no test data, because you think you don't hear it. We all like the tpa3255 because TI engineers improved the measurements compared with tpa31xx.
Can you hear that?
No I don't hear any difference in the lowest ripple PSU I had 13uV and that Mean Well with 200uV.
But sure you can measure it.
You have built up an succesfull TPA3118 with help of people's mods and sell it comercial in an pretty big company.
But we as DIY can't buy PCB or have any schematic.
Please start an thread in Vendor forum for the company you work at so all now which you present.
I will start the thread as soon as I get the new pcb...sorry for the delay but nowadays things are slower
Meanwhile, like drmordor, jplester and many others, I will keep on posting my test data and how I am able to reduce thd....independant of the fact that you cannot hear it.
I also hope to post a novel way of reducing distortions in output filters that's not in any white papers as a thx to this community. I will wait for a double blind test to make sure measurements and sound quality agree.
Meanwhile, like drmordor, jplester and many others, I will keep on posting my test data and how I am able to reduce thd....independant of the fact that you cannot hear it.
I also hope to post a novel way of reducing distortions in output filters that's not in any white papers as a thx to this community. I will wait for a double blind test to make sure measurements and sound quality agree.
Linear or smps, the main thing is low noise in the 20khx bandwidth. Tpa3255, like every other class d open loop has non existant power supply noise rejection and even when you close the loop, you only gain a few dbs.
Datasheet says _65db of psrr but that's with no signal. Just a marketing trick.
D
Deleted member 148505
Any comparable power supply in the market?
1.5uVrms is lower than the output noise of TPS7A, ultra low noise post switching regulator
Bucks bunny, Tpa3255 datasheet specifies 65db psrr in btl with no signal.. It's clear that once you apply signal you will have less psrr.
Also there is an article by ti engineers that explains that psrr of btl in class d is incorrect compared with same numbers on class ab. It simply does not capture the modulation of analog signal by psu noise.. but you can see it on IMD.
Jplester I know that it sounds crazy... and yet that's what I see on fft and AP. Note also that I have designed multiple psu, 15w smps resonant with 1.5uv, dual linear with less than 1uv 40w. My first smps project was a llc, so I took the old project and applied the new lessons.
In fact I would be more than happy to make the new llc available, pm me.
Also there is an article by ti engineers that explains that psrr of btl in class d is incorrect compared with same numbers on class ab. It simply does not capture the modulation of analog signal by psu noise.. but you can see it on IMD.
Jplester I know that it sounds crazy... and yet that's what I see on fft and AP. Note also that I have designed multiple psu, 15w smps resonant with 1.5uv, dual linear with less than 1uv 40w. My first smps project was a llc, so I took the old project and applied the new lessons.
In fact I would be more than happy to make the new llc available, pm me.
Last edited: