Ampapa D1 Stereo Amplifier Review

[john61ct]

TBH I would rather just drop it, but I love to be corrected if it helps my learning.

I have no idea what exactly you are saying with that big post full of tables and charts nor that long youtube.

Please quote my text that you think is wrong, and **concisely** state, just what you think is right

one at a time...

TIA

 

[ICIETDIYEUR]

It is however simple to understand, it is necessary to properly adapt the PSU with the AMPAPA D1 amplifier according to at least 3 criteria:

- the release of heat in IDLE.

- the power (W) as well as the voltage (U) and the intensity (I) depending on the load used (4 or 8 Ohm impedance speakers).

- The possibility for the amplifier to operate without overheating (or 'clipping' or stopping/protecting itself or its power supply) and above all with suitable distortion that is expected of a so-called 'Hi-Fi' installation (less than or equal to 0.1%, i.e. -60dB).

We are on a site that talks about science and you tried to approach it: I thought you would understand easily.

 

[john61ct]

You are fighting a straw man. I asked you to quote the words I wrote that you feel were wrong.

None of the general statements you just wrote address anything I wrote afaict, nor do I see what you are recommending from a practical POV

And again "current" or Amps is the correct term, using "intensity" for that is just confusing, an obsolete holdover from the original French.

 

[NTK]

The heating up of the output LC filter inductor is due to core loss, or magnetic loss, as explained in the TI application report. In calculations, a parallel resistance Rp is used to account for the loss. Different inductors have different values of Rp, and therefore different amount of heat generation under idle conditions*.

Since core loss is modeled as a parallel resistance with the ideal inductor, it is obvious that the loss will be roughly proportional to the square of the PWM rail voltage. The idle power consumption numbers given in post #178 were 48 V → 9 W and 36 V → 6.5 W. If we exclude the datasheet BTL mode switch loss of 2.5 W, the ratio of power loss is (9-2.5)/(6.5-2.5) = 1.6. The ratio of core loss due to voltage increase is (48/36)² = 1.8. These numbers are close enough and explain the source of the idle power consumption increase, as there are other system losses that have not been not considered.

* The output of the TPA32xx chip, which uses AD modulation, is not zero at zero output. The output waveform is a square wave at the PWM frequency with 50% duty cycle. Therefore, it is fully expected that there will be power loss/dissipation at the output LC filter.

 

[ICIETDIYEUR]

Hello NTK.

I never said that a loss/dissipation of power at the output LC filter was abnormal: I simply said that it is proportional to the PVDD voltage (that of the PSU), the greater PVVD, the greater the heat dissipation.

Furthermore, to follow up on what is observed in the videos, I simply indicated that it was not wise to increase the voltage of the PSU (PVDD in fact) if the amplifier cannot follow 'thermally' and if this results in a restoration of the signals with distortion values that cannot be accepted in 'Hi-Fi'.

What is your view on this?

PS: please excuse me john61ct if I don't answer you, it's just that I perhaps have the bad impression of maintaining a 'dialogue of the deaf' with you, without wanting to disrespect you.
Maybe I'm having trouble making myself understood, maybe it's the translations.

 

[john61ct]

no worries at all

 

[ICIETDIYEUR]

People hi

Here is (another) one of my thoughts...

A speaker with 89dB/W/m reproduces 89dB with only 1W: this is what we must understand (if the values given by the manufacturer are real).

To obtain 92dB, you must increase by 3dB and for this you must increase the power by 2, or 2W (only).

To obtain 99dB, you must increase by 10dB and for this you must increase the power by 10, or 10W (only too).

Now, I invite you to equip yourself with a sound level meter and increase the sound until you obtain these 99dB

Knowing that the musical reproduction is in impulse mode, this level will increase enormously during passages with low frequencies but provided that the amplifier and its power supply have been designed suitably.

You will see that at this level it is already very loud, provided that your speakers can reproduce this level without too much distortion, which is not always obvious if we simply refer to reviews of certain models carried out by Amirm.

It will therefore be necessary with a given amplifier (here the AMPAPA D1) to have very precisely calibrated the power (W) of the PSU as well as its voltage (V) and its current (I) and this is not done simply by purchasing (like everyone else) a 48V/10A Gan power supply...

 

[NTK]

...
Now, I invite you to equip yourself with a sound level meter and increase the sound until you obtain these 95dB
...

This is taken from the manual of the NIOSH Sound Level Meter app. The numbers are just for reference. The LAeq is the average SPL. The LCpeak is the maximum pressure during the measurement period, which the amplifier will have to supply the required power to the speakers to generate this sound pressure.

Notice the required headroom over average? Measure it yourself with the sound level meter app, and see for yourself what is the required headroom in your music listening session.

NIOSH Sound Level Meter App

The NIOSH Sound Level Meter app measures workplace noise to determine if the levels are damaging.

www.cdc.gov

BTW, a 3 dB level difference is double the power, not 10x.

 

[NTK]

Hello NTK.

I never said that a loss/dissipation of power at the output LC filter was abnormal: I simply said that it is proportional to the PVDD voltage (that of the PSU), the greater PVVD, the greater the heat dissipation.

Furthermore, to follow up on what is observed in the videos, I simply indicated that it was not wise to increase the voltage of the PSU (PVDD in fact) if the amplifier cannot follow 'thermally' and if this results in a restoration of the signals with distortion values that cannot be accepted in 'Hi-Fi'.

What is your view on this?

PS: please excuse me john61ct if I don't answer you, it's just that I perhaps have the bad impression of maintaining a 'dialogue of the deaf' with you, without wanting to disrespect you.
Maybe I'm having trouble making myself understood, maybe it's the translations.

This is from the datasheet of Coilcraft inductors often used in class-D amplifier output filters. These inductors are rated to work at 125 °C. Therefore 50 or 60 °C is hardly a problem.

Amir measured this 3e A7 Mono with a 52 VDC power supply. I see no problem with distortions.

[Edit] This is an official product review thread, and this is my last off topic post. My apology to Amir and Rick.

 

[ICIETDIYEUR]

Hello NTK.

Correct, I was in too much of a hurry: I will correct differences in level

The 3eAudio has 2 TPA 3255 chips in PBTL mode, it is not comparable, The 52V PSU allows the current/intensity (A) to be reduced, we can also look at the curves in the TI datasheets.

Amirm does not specify for how long this power was maintained, nor in what conditions (at least the type of load) and even less the temperature reached by the amplifier...

It would have been interesting to know the intensity measured with this power as well.

In any case, as you pointed out just before: in real listening conditions we are far from the reality of what was measured.

NB: the inductors that you show are not those of the AMPAPA D1 amplifier, but what is problematic is not their temperature but that of the PCB in general and all of these components.

Generally, excessive heat degrades the initial/optimal characteristics of components, no ?

 

[Head_Unit]

adjustable 30–200Hz high-pass filter

thank you Jesus Buddha Allah Tom Cruise! At long last we start to see some motion in this direction. AND a REMOTE CONTROL oh lord the Rapture must be nigh! All these cute little amps never have a damn remote, grrrrrrr.

AND old time VU meters...actually it was VU meters that got me into my career! Went into a hi-fi shop, listening room is blasting like Maynard Ferguson or something, really powerful, quite loud, and the amplifier meters were sitting around 1/100 of a watt. !!!!!! Sales guy comes in "can I help you" "ah, how the heck can these speakers play so loud on so little power?!?" "oh they're very efficient due to this horn (Altecs)"
HOOKED.
Right then and there decided to become a loudspeaker designer, which eventually after "all that fooling around in the basement" as my mother puts it, I did.

 

[john61ct]

The 3eAudio has 2 TPA 3255 chips in PBTL mode, it is not comparable

I was definitely comparing them when you responded to me (as if you disagreed, but without clarity.)

The ABILITY to withstand / make use of over 48V is what delivers high SPL from TPA3255, using two of them in PBTL is yet another further design advantage.

Others may not agree on the relative importance of the power Watts and current Amps, for me they are a dealbreaker, at least for my main front pair. If my surround pair turn out to be very efficient not require much of either then I might get this Ampapa for the bouncy lights and (maybe) HPF

It may not be a "fair" comparison but obviously they CAN and should be compared.

> It would have been interesting to know the intensity measured with this power as well.

Again, the English term is just "current". Amps is the only other commonly understood synonym for clarity, that I-word is simply confusing. Just trying to help you with your English skills...

> I think The 52V PSU allows the current/intensity (A) to be reduced

Yes, further reason high current is rarely a problem, if overheating does in fact result, the user has several easy solution choices. There is no indication of such overheating actually occuring, with either 3e A7 , much less this Ampapa model.

 

[Hipster Doofus]

Which EMIs are you using? I've passed on a couple pair of the DLS 529s (mismatched or bad cosmetically, etc) but I'm intrigued by them.

DLS 1 Similar to the one behind Ringo

 

[john61ct]

nice flex!

 

[Scrivs]

I was definitely comparing them when you responded to me (as if you disagreed, but without clarity.)

The ABILITY to withstand / make use of over 48V is what delivers high SPL from TPA3255, using two of them in PBTL is yet another further design advantage.

Others may not agree on the relative importance of the power Watts and current Amps, for me they are a dealbreaker, at least for my main front pair. If my surround pair turn out to be very efficient not require much of either then I might get this Ampapa for the bouncy lights and (maybe) HPF

It may not be a "fair" comparison but obviously they CAN and should be compared.

> It would have been interesting to know the intensity measured with this power as well.

Again, the English term is just "current". Amps is the only other commonly understood synonym for clarity, that I-word is simply confusing. Just trying to help you with your English skills...

> I think The 52V PSU allows the current/intensity (A) to be reduced

Yes, further reason high current is rarely a problem, if overheating does in fact result, the user has several easy solution choices. There is no indication of such overheating actually occuring, with either 3e A7 , much less this Ampapa model.

I think what he meant to say all along is: “Unless you need the ultimate in clean power/SPL ability for (insert TPA325X appliance here), go with the lowest Voltage PSU that will satisfy your routine “SPL” needs, as higher Voltage supplies will incur more heat loss (amplifier board) at idle, and in general, than lower ones, all else being equal.”
Not sure why core loss for the output inductors was brought up, aside from their contribution to idle losses/heating, which varies with PVDD.

 

[Scrivs]

It’s been my experience that buck converter I.Cs, LDOs, and things like that are what fails at higher PVDD on these small boards in their little enclosures.

 

[john61ct]

I think what he meant to say all along is: “Unless you need the ultimate in clean power/SPL ability for (insert TPA325X appliance here), go with the lowest Voltage PSU that will satisfy your routine “SPL” needs, as higher Voltage supplies will incur more heat loss (amplifier board) at idle, and in general, than lower ones, all else being equal.”
Not sure why core loss for the output inductors was brought up, aside from their contribution to idle losses/heating, which varies with PVDD.

Unless you get a voltage-adjustable PSU like Mean Well, IMO that is too conservative, you're short-changing yourself unless you KNOW FOR SURE you will NEVER want to surpass SPLx.

At least with a vendor like 3e, when the data sheet says 52V is fine, IMO you can safely run 48V all day long, even 50V should not be risky.

To be cautious, testing / monitoring temps is a good idea, just turn down the volume is the better response than permanently kneecapping the amp.

Of course if you have spare PSUs around at different voltage / current levels, swapping them per context is as good as using a voltage-adjustable unit

But these days good PSUs are costing more than the amps.

 

[ScottySmallsMom]

would the load dependency make this amp not a great choice for C6B (mostly nearfield) ?

 

[ICIETDIYEUR]

People hi

Scrivs perfectly understood what I was trying to explain in an 'unskillful' way, perhaps coming from the 'language barrier' or am I simply not 'pedagogical' enough in a certain way and so I thank him

An excellent suitable power supply does not necessarily cost more than the amplifier (here the AMPAPA D1 model) associated with it, here is an example accompanied by some tables which I hope will allow me to better understand my reasoning ->

NB: the 48V model would achieve 130W RMS per channel with a load of 8 Ohms (speakers)

IMPORTANT: you will notice that the 36V and 48V models with a 4 Ohm load would only allow the advertised powers to be achieved 'at peak' and not at continuous operation as would be the case with 8 Ohm loads.

This MEANWELL LOP-500-48 power supply (the LOP-600-48 is almost the same price) is excellent (at a price of around €70, shipping included), although it requires to be integrated into a protective box, and will certainly be more efficient in operation at full power over a long period of continuous use than 'the well-known one' sold under the name 48V/10A Gan but I could be wrong (?)

So i think it could be interesting in the future to measure and compare these two power supplies in 'intensive use'.

 

[low_art]

DLS 1 Similar to the one behind RingoView attachment 515456View attachment 515455

Very nice- definitely not the "bookshelf" DLS. I'd love to play some Beatles through that system.