Yesterday with couple frinds of mine we tested POWERSOFT K10 and POWERSOFT K3 on dummy load 2 channels loaded with 2 ohms each, one channel with 2 ohms and one channel wit 4 ohms (K3), scope probe and dummy load put directly on 1 + and 1 - on speakon, (only one channel with scope, because non ellectrically insulated or ballanced probes), knowing each ch is floating + and - outputs, BTL out.
They was supplied correct, K10 directly from main distro from 40A fuse, and K3 on main soket from 20A fuse.
Those amps having PFC (power factor correction), must be imune at main tension from 100 to 265Vac.
All I can say is "the fun" is lasting just on small burst regim couple tens of miliseconds, thereafter it limits output power and current (you'll see how and how much) until thermal limit occurs inevitably.
On squarewave signal they are ok until 4-6Khz, at any amount of power and after...................
But better, guys watch youself.
You can make coments pro or against, I don't have any comercial interest on any amp producer, I just tried to show you all, exactly how those amps perform on 2 ohms load per channel on sinewave signal.
Just from mine point of view they must sustain min 3-5 seconds Pmax on 2 ohms load until current limit occurs as official specs to be really 100% for any musical program compressed or not, as known, DUBSTEP is most energy amp demanding program material, there crest factor is less than 3db, most normal live programs or pop rock gave at lest 9-12 db crest factor.
Powersoft K10 2 Ch/2 ohm:
Short peak of 103-105Vrms/2 ohm in limiting.
https://www.youtube.com/watch?v=QjGHKoTKRjk
Powersoft K10 1 Ch/2 ohm:
Same perform, from where we can understund limitings occurrs from final stage not power supply:
https://www.youtube.com/watch?v=wD4MPEjk23c
Powersoft K10 in sinusoidal, triangular and squarewave part 1:
https://www.youtube.com/watch?v=mQpRZmjvri8
From 4-6Khz and above, personal I don't like what I see, response on squarewave signal is worse than PKN XD2500, tested here at min 9:45 https://www.youtube.com/watch?v=gYSUsRi880A ...
The fact as at 10Khz squarewave becomes sinusoidal, shows poor rise time and fall times (more than 16uSecs) , incomparably worse than a big power TD class amplifier or a classH or old Macrotech 5000VZ.
Powersoft K10 in sinusoidal, triangular and squarewave part 2:
https://www.youtube.com/watch?v=FpeAOjQMy3I
And now, the small brother, POWERSOFT K3, in same testing conditions:
https://www.youtube.com/watch?v=flkGIeehOyY
Personally I use these amps exclusively on subbass-bass-midbass duty (20-200Hz), at most on mid until 3-4 Khz with adecvated external limiter preseted to cut any higher level wich can put amp to deliver more that 85-90% from its's clean sinusoidal output on 2 ohms load, because at short higher peaks, his internal limiter acts not so good, lefting 1-2 secs on outputs, hard clipped signal, until it cuts outs for clean sinusoidal signal on output, compairing with the amp tested there in same condition, where limiter acts almost perfect:
https://www.youtube.com/watch?v=7tWFikrMxHo
Waiting for your comments!
They was supplied correct, K10 directly from main distro from 40A fuse, and K3 on main soket from 20A fuse.
Those amps having PFC (power factor correction), must be imune at main tension from 100 to 265Vac.
All I can say is "the fun" is lasting just on small burst regim couple tens of miliseconds, thereafter it limits output power and current (you'll see how and how much) until thermal limit occurs inevitably.
On squarewave signal they are ok until 4-6Khz, at any amount of power and after...................
But better, guys watch youself.
You can make coments pro or against, I don't have any comercial interest on any amp producer, I just tried to show you all, exactly how those amps perform on 2 ohms load per channel on sinewave signal.
Just from mine point of view they must sustain min 3-5 seconds Pmax on 2 ohms load until current limit occurs as official specs to be really 100% for any musical program compressed or not, as known, DUBSTEP is most energy amp demanding program material, there crest factor is less than 3db, most normal live programs or pop rock gave at lest 9-12 db crest factor.
Powersoft K10 2 Ch/2 ohm:
Short peak of 103-105Vrms/2 ohm in limiting.
https://www.youtube.com/watch?v=QjGHKoTKRjk
Powersoft K10 1 Ch/2 ohm:
Same perform, from where we can understund limitings occurrs from final stage not power supply:
https://www.youtube.com/watch?v=wD4MPEjk23c
Powersoft K10 in sinusoidal, triangular and squarewave part 1:
https://www.youtube.com/watch?v=mQpRZmjvri8
From 4-6Khz and above, personal I don't like what I see, response on squarewave signal is worse than PKN XD2500, tested here at min 9:45 https://www.youtube.com/watch?v=gYSUsRi880A ...
The fact as at 10Khz squarewave becomes sinusoidal, shows poor rise time and fall times (more than 16uSecs) , incomparably worse than a big power TD class amplifier or a classH or old Macrotech 5000VZ.
Powersoft K10 in sinusoidal, triangular and squarewave part 2:
https://www.youtube.com/watch?v=FpeAOjQMy3I
And now, the small brother, POWERSOFT K3, in same testing conditions:
https://www.youtube.com/watch?v=flkGIeehOyY
Personally I use these amps exclusively on subbass-bass-midbass duty (20-200Hz), at most on mid until 3-4 Khz with adecvated external limiter preseted to cut any higher level wich can put amp to deliver more that 85-90% from its's clean sinusoidal output on 2 ohms load, because at short higher peaks, his internal limiter acts not so good, lefting 1-2 secs on outputs, hard clipped signal, until it cuts outs for clean sinusoidal signal on output, compairing with the amp tested there in same condition, where limiter acts almost perfect:
https://www.youtube.com/watch?v=7tWFikrMxHo
Waiting for your comments!
Last edited:
I know that pdf.
Upper are real tests on real dummy loads, not what they "claim" on that pdf or prospects.
Upper are real tests on real dummy loads, not what they "claim" on that pdf or prospects.
"The fact as at 10Khz squarewave becomes sinusoidal, shows poor rise time and fall times (more than 16uSecs)"
A typical slew rate issue is occurring when big 10kHz signal gets more distorted into an almost triangle waveform than small signal because of higher dv/dt.
In this case, it might not even be slew rate issue, since the small signal bandwidth of Powersoft K series is "only" ~23kHz. So the lowest harmonic of the 10kHz square is 30kHz and that is filtered away by the bw limit so no wonder almost pure sinusoidal wave remains. Acoustically that makes NO difference at all as there is NOBODY on earth who can tell the difference between 10kHz square and 10kHz sine wave by ear as long as there is no oscilloscope.
A typical slew rate issue is occurring when big 10kHz signal gets more distorted into an almost triangle waveform than small signal because of higher dv/dt.
In this case, it might not even be slew rate issue, since the small signal bandwidth of Powersoft K series is "only" ~23kHz. So the lowest harmonic of the 10kHz square is 30kHz and that is filtered away by the bw limit so no wonder almost pure sinusoidal wave remains. Acoustically that makes NO difference at all as there is NOBODY on earth who can tell the difference between 10kHz square and 10kHz sine wave by ear as long as there is no oscilloscope.
Diferences between squarewave and sinewave on audio amplifier can be easy heared by the ear, try-it especially at low power, couple watts.
Compairing with that, classD amps are FAR away from high quality sound, especiallt above 12Khz.
Watch from minute 6:10.
https://www.youtube.com/watch?v=4abEScdfLxc
Maybe ear can be fooled at untrained people, but, always will be a "but".
Compairing with that, classD amps are FAR away from high quality sound, especiallt above 12Khz.
Watch from minute 6:10.
https://www.youtube.com/watch?v=4abEScdfLxc
Maybe ear can be fooled at untrained people, but, always will be a "but".
Last edited:
Square consists of fundamental plus 3rd, 5th, 7th, 9th
Sine is only fundamental.
Take a square signal and listen to it, maybe 1000Hz, then change to sinewave by filtering out the harmonics. NOT by just flipping the switch on your signal generator as this would most likely change the amplitude of the fundamental component too.
At 1000Hz you would easily notice the difference, also at 2000, 3000, maybe 3500. At 4000 it gets very difficult, since the first harmonic is 12000 will be completely masked by the 4000Hz. Some good ears might hear the difference at 5000, maybe even 6000 but above 7000Hz that there is NO CHANCE AT ALL to hear the difference between sine and square, since the first harmonic is already above 20kHz and cannot be heard even by children. only bats or dogs or the like..
Note that the amplitude of the fundamental component must keep equal - To avoid dealing with this trouble, just turn up each tone from zero slowly until the test person himself says stop loud enough so the test person does not get any clue what signal sounds louder... Test person must not have view of the signal generator of course.
Sine is only fundamental.
Take a square signal and listen to it, maybe 1000Hz, then change to sinewave by filtering out the harmonics. NOT by just flipping the switch on your signal generator as this would most likely change the amplitude of the fundamental component too.
At 1000Hz you would easily notice the difference, also at 2000, 3000, maybe 3500. At 4000 it gets very difficult, since the first harmonic is 12000 will be completely masked by the 4000Hz. Some good ears might hear the difference at 5000, maybe even 6000 but above 7000Hz that there is NO CHANCE AT ALL to hear the difference between sine and square, since the first harmonic is already above 20kHz and cannot be heard even by children. only bats or dogs or the like..
Note that the amplitude of the fundamental component must keep equal - To avoid dealing with this trouble, just turn up each tone from zero slowly until the test person himself says stop loud enough so the test person does not get any clue what signal sounds louder... Test person must not have view of the signal generator of course.
DjLeco Goodwork..!!!
Good to see the REAL power of K10, so much fake power they claim.
You are right about dubstep power demands.
Good to see the REAL power of K10, so much fake power they claim.
You are right about dubstep power demands.
Sounds like you are having some heavy intermodulation, causing components in the audible band.
All class D amps? Which ones have you tested?
I have tested alot of classD (PKN XD, QSC PLX , Crest LT, Ecler DT series) and class I amplifiers crown exclusive Itech series.
https://www.youtube.com/watch?v=N06eMWeaqig
https://www.youtube.com/watch?v=gYSUsRi880A from min 11
https://www.youtube.com/watch?v=hnBBa8FhzUc
https://www.youtube.com/watch?v=kX27yqarkCM
All of them, on squarewave above 5-8Khz, gets triangular or sinewave on output, because of 20-40uH output coil filter used.
Squarewave response at 10Khz is long far away fron AB class amplifier.
Compare it with that amp from min 6:10 sec.
https://www.youtube.com/watch?v=4abEScdfLxc
When classD will be able to offer response like the one above, for 10Khz squarewave, I will lay down my hat.
To be like that, the switching freq must be around 1,5-2Mhz, as can be used 2,2uH as output filter, being almost like bucherot filter from AB class amplifier.
If they can do that, efficiency will be more less then actually (250-500Khz switching freq), but I don't think they will do very big power amplifiers at more than 500Khz switching freq.
Even Labbruppen (TD class, wich is B or ab class with switching downconverter rails supply amplifier), dropped their switching freq from 614Khz , first fp series, to less (autooscilant downconverter).
I don't say classD is no good, but I'm very pretentious on sound quality, as if I will have classD or classI amplifiers, I will use exclusively on bass duty.
https://www.youtube.com/watch?v=N06eMWeaqig
https://www.youtube.com/watch?v=gYSUsRi880A from min 11
https://www.youtube.com/watch?v=hnBBa8FhzUc
https://www.youtube.com/watch?v=kX27yqarkCM
All of them, on squarewave above 5-8Khz, gets triangular or sinewave on output, because of 20-40uH output coil filter used.
Squarewave response at 10Khz is long far away fron AB class amplifier.
Compare it with that amp from min 6:10 sec.
https://www.youtube.com/watch?v=4abEScdfLxc
When classD will be able to offer response like the one above, for 10Khz squarewave, I will lay down my hat.
To be like that, the switching freq must be around 1,5-2Mhz, as can be used 2,2uH as output filter, being almost like bucherot filter from AB class amplifier.
If they can do that, efficiency will be more less then actually (250-500Khz switching freq), but I don't think they will do very big power amplifiers at more than 500Khz switching freq.
Even Labbruppen (TD class, wich is B or ab class with switching downconverter rails supply amplifier), dropped their switching freq from 614Khz , first fp series, to less (autooscilant downconverter).
I don't say classD is no good, but I'm very pretentious on sound quality, as if I will have classD or classI amplifiers, I will use exclusively on bass duty.
Last edited:
I am not sure I would call 3 "a lot".
And what does square wave response at 10 kHz have to do with "high quality sound"?
What's your view on, let's say, the hypex nc400?
What do you think that a 10kHz or 20kHz rectangle looks like when stored in CD format and reproduced with the very best CD players?
If you can live with the sound quality of a CD, then a rise time of 10us in the following equipment is not a bottle neck.
This does not mean that the amps you tested must sound great.
I am just not convinced that a rise time of 10us must cause a poor sound.
This being said by a fan of rectangle esthetics during amp design 😉
Typically I am adjusting loop gain and frequency compensation in my class D amps in way which results in rise times of approx 5us and no overshoot.
So far I think that's faster than what audio needs and have shifted my efforts more on things like IMD and reproduction of complex signals. I.e. 1kHz triangle with 70% of max voltage and superimposed 10kHz rectangles with 5-10% and see if the rectangle is reproduced reasonably well along the entire triangle.
I tested more classD amps but I just enumerate couple of them, the ones with some power on output.
Amps under 500Watts on classD are not important to me, as one good amplifier (AB or AB+B) can deliver 500 or more watts on 8 ohms load, at maximum sound quality Crown Macrotech especially).
To understand how amps perform better than one, let's take by example 2 amplifiers, for let's say, 400-500Watts power on 2 ohms load no matter how powerfull are one or other but both to be able to deliver minimum 600Watts on 2 ohms load, so clipping will not occur when demandind those 500 watts on output.
Powersoft K3, wich have rise and fall times of output voltage (peak to peak) of around 22-24 microseconds, as seen on video from first post , need 22-24 microseconds to swing from 10 to 90% of Vpp excursion.
The second amp, as you see, wich have rise and fall times of output voltage (peak to peak) of around 2,5 microseconds, need 2,5 microseconds to swing from 10 to 90% of Vpp excursion.
That means, second amplifier is almost 9,8 times FASTER than first amplifier.
As high freqs need faster response times to be perfectly accurate, I let you to think wich performs better for mid hi or fullrange program, regarding just sound quality, not randament or burst power.
Amps under 500Watts on classD are not important to me, as one good amplifier (AB or AB+B) can deliver 500 or more watts on 8 ohms load, at maximum sound quality Crown Macrotech especially).
To understand how amps perform better than one, let's take by example 2 amplifiers, for let's say, 400-500Watts power on 2 ohms load no matter how powerfull are one or other but both to be able to deliver minimum 600Watts on 2 ohms load, so clipping will not occur when demandind those 500 watts on output.
Powersoft K3, wich have rise and fall times of output voltage (peak to peak) of around 22-24 microseconds, as seen on video from first post , need 22-24 microseconds to swing from 10 to 90% of Vpp excursion.
The second amp, as you see, wich have rise and fall times of output voltage (peak to peak) of around 2,5 microseconds, need 2,5 microseconds to swing from 10 to 90% of Vpp excursion.
That means, second amplifier is almost 9,8 times FASTER than first amplifier.
As high freqs need faster response times to be perfectly accurate, I let you to think wich performs better for mid hi or fullrange program, regarding just sound quality, not randament or burst power.
Last edited:
So I assume that means you haven't tested the nc400 for example?
But considering the highest frequency signal the amps need to be able to reproduce, what does it matter?
When you say "faster response times" I assume you mean slew rate?
What difference does "response time" make as long as you can reproduce a 20 kHz signal at full amplitude?
Simply analyse music program and you will see that the content of high frequencies is low - there is absolutely no indication that large signal rectangle response would be a relevant criteria for audio.
But if you demand it - you should also demand it from the source (again pointing to CD players) and also from the speakers.
IMHO you are forcing to much one single criteria, which only of partial relevance.
Imagine a tractor for the field. Will it be better, after changing the transmission to achieve a top speed of 150km/h?
Of course 150km/h is a better measurement value than just 50km/h, but is this the right direction for improving the tractor?
Or imagine a Porsche and now change the chassis that it can carry 5t instead of just some hundreds kg. Of course the ability of carrying 5t is better than the ability of transporting just two grown ups and a dog - but would this improved construction be a better sports car?
The measurements you show are telling of course some technical properties, but I would not dare to judge on the sound from the shown values.
Even the slowest one is in the same category like a CD record playback.
Julf, I'm not presenting here hypex, nc400,or other low power class D amps, I'm speaking about 2x6000Watts (2x5900 watts by specs) claimed by amplifier wich is K10 for 2 ohms load.
As you see, he is not able to sustain that power.
Other things is not so important.
Would you pay 5000 euros for that kind of amplifier wich on 2 ohms can sustain less than half power declared on oficial specs?
And those 2500 watts per channel delivered for couple seconds and after thermal limiting?
I didn't posted here too argue with anybody, personally I don't like how this amp is perform for max power on sinewave, and at any power for squarewave, tested at 3,5 Vrms on out, until 68Vrms on out.
Also tested on squarewave, his response is POOR, 23usec, rise and fall times, is not good for me.
If for you or other is ok, go buy it!
Show me one classD wich has 5usec rise and fall times at 90% from P-max and I will love that.
Till then, those monster burst amplifiers are just with lies specs, I wanna see thoose 2x5900 watts sustained for min 10 seconds until limiting final stage or thermal.
Burst amplifiers are like a short loud fart, and after small farts until end...
As you see, he is not able to sustain that power.
Other things is not so important.
Would you pay 5000 euros for that kind of amplifier wich on 2 ohms can sustain less than half power declared on oficial specs?
And those 2500 watts per channel delivered for couple seconds and after thermal limiting?
I didn't posted here too argue with anybody, personally I don't like how this amp is perform for max power on sinewave, and at any power for squarewave, tested at 3,5 Vrms on out, until 68Vrms on out.
Also tested on squarewave, his response is POOR, 23usec, rise and fall times, is not good for me.
If for you or other is ok, go buy it!
Show me one classD wich has 5usec rise and fall times at 90% from P-max and I will love that.
Till then, those monster burst amplifiers are just with lies specs, I wanna see thoose 2x5900 watts sustained for min 10 seconds until limiting final stage or thermal.
Burst amplifiers are like a short loud fart, and after small farts until end...
Last edited:
You definitely don't need it for full slope.
If you have fast and proper response for small and medium signals and independend of the load and also remaining blameless in combination with LF large signals - that's what serves music program and various speaker impedances.
Have a look ...
Shown measurements are unbridged with various loads down to 2R.
Bridged the amp serves 2kW into 4R.
Posting #316 (the PA-adjustment V1.3 with 7-8us rise at lower levels and 10us at high levels)
http://www.diyaudio.com/forums/class-d/224052-systemd_2kw-any-interest-open-design-32.html
And the Hifi-adjustment V1.5 with 3-4us rise time at lower and mid power
posting #424
http://www.diyaudio.com/forums/class-d/224052-systemd_2kw-any-interest-open-design-43.html
And in posting #431 some full power sines in bridged mode
http://www.diyaudio.com/forums/class-d/224052-systemd_2kw-any-interest-open-design-44.html
..agree... often power ratings are just meaningless numbers under irrelevant conditions. Especially single burst ...old game of catching customers... ..method prooved to work since the first power amp has been advertised.
Repetitive bursts with realistic duty cycles might make sense in some sort.
I want to see that measurement made with 2 ohms load at minimum 160Vpp on output, to be some power there, not at +/- 12 volts.
I want to see it on movie, not static image, as I have presented
I want to see it on movie, not static image, as I have presented
Fair enough, the nc400 is only 650 W max. But please don't state stuff like "classD amps are FAR away from high quality sound". Some might be, some not. What you should have said is "of the 6000 W amps I have tried, none have looked good in terms of being able to sustain the specified power".
Actually, seems square waves are an important thing for you.
I have no opinion on those, and they haven't been discussed until now.
How often do you listen to squarewaves?
Show me one that sounds good, and I will love that, whatever rise time it has.
Rise time, as long as it is high enough to accommodate 20 kHz at full amplitude, is not a measure of quality.
You want a lot:
...firing a 2 Ohms tweeter with 10kHz-20kHz rectangles of at least 3.2kW.
...watching movies.
Both nothing that I intend to serve.
Not going for a design mistuning according to your request,
and also not going into the business of Vin Diesel.
- Status
- Not open for further replies.