Hello guys,Well I was talking with guy and He said to me that no matter what you do with this chip LM3875 ,the sound will always be "crab".this is what He said in one of his post:
The integrated chip you talking about cost around $3.30 (hXXp://www.national.com/mpf/LM/LM3875.html). This chip provides 56W continues power and 100W peak power. The site you talking about tries to get huge profit taking more than $1000 for box which has couple of this chips and couple more details. The copper sink is so small it's unbelievable. My very old center in a basement has couple amplifiers like that (two integrated chips very similar or probably same) and has much bigger sinks. The chip you talking about has adequate distortions when power lower than 56W. Then it goes up very steep and at 100W it has distortion about 10%. I'm not talking you should listen it at 100W.
" lanchile, do this experiment:
Load this amplifier with 8 ohm 100W resistor (well, i know it's not so easy, resistor should be 100W, or you can create one using a many resistors in parallel creating one 8 ohm.). Then connect its input to sound generator (use sinusoidal output of 1000HZ let say, but you can use other frequency if you want) and output to oscilloscope. Set volume to max and adjust input signal value to have output signal value amplitude +-30V (controlling by oscilloscope difference from max to min will be 60V) for 56W. In that case your power supply MUST have voltage not less that +- 30V for bipolar scheme or more than +60V for not bipolar scheme. If you got that output signal on a 8Ohm resistor with no visible distortion - fine. If you see visible distortion - it mean that distortion is around 10% already. Don't forget to load both chips in same time, so you will have a real test. If you don't have oscilloscope you can connect the voltmeter and it should show you around 21V (you will not see distortions though).
Good luck".
I think he never buid one ,for what I can see lol.
The integrated chip you talking about cost around $3.30 (hXXp://www.national.com/mpf/LM/LM3875.html). This chip provides 56W continues power and 100W peak power. The site you talking about tries to get huge profit taking more than $1000 for box which has couple of this chips and couple more details. The copper sink is so small it's unbelievable. My very old center in a basement has couple amplifiers like that (two integrated chips very similar or probably same) and has much bigger sinks. The chip you talking about has adequate distortions when power lower than 56W. Then it goes up very steep and at 100W it has distortion about 10%. I'm not talking you should listen it at 100W.
" lanchile, do this experiment:
Load this amplifier with 8 ohm 100W resistor (well, i know it's not so easy, resistor should be 100W, or you can create one using a many resistors in parallel creating one 8 ohm.). Then connect its input to sound generator (use sinusoidal output of 1000HZ let say, but you can use other frequency if you want) and output to oscilloscope. Set volume to max and adjust input signal value to have output signal value amplitude +-30V (controlling by oscilloscope difference from max to min will be 60V) for 56W. In that case your power supply MUST have voltage not less that +- 30V for bipolar scheme or more than +60V for not bipolar scheme. If you got that output signal on a 8Ohm resistor with no visible distortion - fine. If you see visible distortion - it mean that distortion is around 10% already. Don't forget to load both chips in same time, so you will have a real test. If you don't have oscilloscope you can connect the voltmeter and it should show you around 21V (you will not see distortions though).
Good luck".
I think he never buid one ,for what I can see lol.
It doesn't sound like he knows what he is talking about. The distortion goes up because the output clips. All amps have high distortion when they clip. This is not really what is talked about when discussing distortion. This is clipping. Increase the supply voltage and amazingly the 'distortion' goes away. You have to use the amplifier in the range it is designed for to avoid such 'distortion'. The LM3876 chip is designed for around 56W applications. If you want higher you have to use something like the LM3886 or bridge or parallel configurations. If he understood what he was talking about then he'd know this.
-SL
-SL
What do you think?
I have build three LM3875 integrated amp.the sound is fantastic,I do not think He ever build or hear a Gainclone amp.I had a "Portal Panache" integrated amp and when I build my third gainclone using the best parts as I could and fallowing all the recommendations and tips from all the pro here.I got an integrated amp that kicked Portal Panache *** big time.I am not talking about power but pure sound.I put it side by side test.and the $1800 portal panache passed to history!.I trust my ears I do not need an oscilloscope to tell me this is a good amp or not!.many people when they see that this chips are $6 dollars,they say "NO WAY IT CAN BE GOOD".well,but if this chip would cost $60 dollars... well everyone would say "WOW THIS CHIP IS THE BEST".
the best thing to say (I think) is "BUILD ONE" and then we talk.
I am not saying a good build Gainclone is the best,but it is for sure better than many commercial amps out there no doubt.
I have build three LM3875 integrated amp.the sound is fantastic,I do not think He ever build or hear a Gainclone amp.I had a "Portal Panache" integrated amp and when I build my third gainclone using the best parts as I could and fallowing all the recommendations and tips from all the pro here.I got an integrated amp that kicked Portal Panache *** big time.I am not talking about power but pure sound.I put it side by side test.and the $1800 portal panache passed to history!.I trust my ears I do not need an oscilloscope to tell me this is a good amp or not!.many people when they see that this chips are $6 dollars,they say "NO WAY IT CAN BE GOOD".well,but if this chip would cost $60 dollars... well everyone would say "WOW THIS CHIP IS THE BEST".
the best thing to say (I think) is "BUILD ONE" and then we talk.
I am not saying a good build Gainclone is the best,but it is for sure better than many commercial amps out there no doubt.
I don't know if the gainclone sound better than the high end system (because i haven't listen to a very expensive one yet) but trust me the chip has very good sounding. It will beat the 200 dollars yamaha amplifier sale in Future shop here in Canada. However, it also depends on what kind of speakers you have.
lanchile07 said:80%? I do not think so! I would say more like 50% maybe
Everything is important, and a system has to be synergistic. But i do feel that once you have achieved a certain level -- and that doesn't necessarily cost a lot -- the speaker is the least important part.
dave
I agree with gainphile here.
Most guys will use an amp with 1% THD or even 0.1 % THD at high volumes.
Problem is, at high volumes, the speaker THD is at 10% or more.
Amplifier contribution to the sound is important, but minimal. Way better to improve speakers than improve the amplifier.
Problem is, top of the line speakers aren't that expensive, for example Seas TBFC/G tweeter at 30$ is considered top of the line by Zaph.
So, we can only improve the amplifier !
Let wish for new speakers technology breakthrough !
Most guys will use an amp with 1% THD or even 0.1 % THD at high volumes.
Problem is, at high volumes, the speaker THD is at 10% or more.
Amplifier contribution to the sound is important, but minimal. Way better to improve speakers than improve the amplifier.
Problem is, top of the line speakers aren't that expensive, for example Seas TBFC/G tweeter at 30$ is considered top of the line by Zaph.
So, we can only improve the amplifier !
Let wish for new speakers technology breakthrough !
Can someone please explain to me how an amplifier can be rated for a certain wattage RMS (56W) and then a peak power rating (100W).
This doesn't make sense? and never has to me??
With my lm3886 i use 35V rails.
I did exactly just what was posted in the firt post.
I had an oscilloscope, and connected it to my input and output signal. I made up a HUGE 8 ohm resistor with a few 25W aluminium resistors in parralel.
I then put in a 1kHz sinewave and turned up the voltage just before the output sine started to clip , the peak voltage was just over 28V.
Which calculates to just about 50W RMS.
The heatsink got HOT but it stayed there.
So my amp cannot deliver anything above 28V into 8ohms. IS the higher wattage ratings for lower impedances?
Is it Peak because it can deliver huge amounts of current into much lower ohm loads but only for the 1/10th of a millisecond the filter caps can deliver it? Because the power supply can't deliver that much power?
Are these peak powers just what the chip theoretically can handle with a perfect power supply and still only for a few nanoseconds like the surge current rectifier diodes can handle?
anyway
The power i am getting out of the current setup is More than i would EVER need. But all depends from person to person i guess.
This doesn't make sense? and never has to me??
With my lm3886 i use 35V rails.
I did exactly just what was posted in the firt post.
I had an oscilloscope, and connected it to my input and output signal. I made up a HUGE 8 ohm resistor with a few 25W aluminium resistors in parralel.
I then put in a 1kHz sinewave and turned up the voltage just before the output sine started to clip , the peak voltage was just over 28V.
Which calculates to just about 50W RMS.
The heatsink got HOT but it stayed there.
So my amp cannot deliver anything above 28V into 8ohms. IS the higher wattage ratings for lower impedances?
Is it Peak because it can deliver huge amounts of current into much lower ohm loads but only for the 1/10th of a millisecond the filter caps can deliver it? Because the power supply can't deliver that much power?
Are these peak powers just what the chip theoretically can handle with a perfect power supply and still only for a few nanoseconds like the surge current rectifier diodes can handle?
anyway
The power i am getting out of the current setup is More than i would EVER need. But all depends from person to person i guess.
It's a numbers game. That 28V peak you measured was 20V RMS, which gives the same heat in an 8 ohms load that a 20V DC would. So, its RMS power is 20^2 / 8 ohms = 50 Watts.
But its peak value is 28V. So, at the instant that the signal is at its peak, it delivers 28^2 / 8 ohms = 100Watts peak... (And of course when the signal crosses zero, that instanteneous power is zero as well). It doesn't have any practical value, but looks better in the brochure.
Jan Didden
But its peak value is 28V. So, at the instant that the signal is at its peak, it delivers 28^2 / 8 ohms = 100Watts peak... (And of course when the signal crosses zero, that instanteneous power is zero as well). It doesn't have any practical value, but looks better in the brochure.
Jan Didden
simon5 said:
Do keep in mind that without the spectrum those numbers are meaningless. A loudspeaker will produce 2nd & 3rd order distortion which are naturally occuring.
An amplifier can be producing higher order harmonics which are unnatural.
For example (i'm just pulling numbers form the air for illustration) you can find (say) 0.001% 5th order distortion from the amplifier annoying thru the 2% 2nd order distortion from the speaker which you don't even notice.
dave
Naturally occuring distortion I agree, but they aren't in the source, so it's still distortion. According to your numbers, the speaker would be distorting like 2000 times more than the amplifier.
Sure it's better to have a nice sounding speaker than a bad sounding one, I agree with you, but then where do you draw the line between what the guy wanted you to hear and what you are actually hearing.
Maybe the guy is untalented but your speakers produce so much 2nd order distortion that it sounds wonderful... 😉
OTOH, just asking, I wonder but maybe a speaker can produce 5th order distortion also ? Like cone breakup modes or else ?
Thanks for your reply, you made a good point there.
Sure it's better to have a nice sounding speaker than a bad sounding one, I agree with you, but then where do you draw the line between what the guy wanted you to hear and what you are actually hearing.
Maybe the guy is untalented but your speakers produce so much 2nd order distortion that it sounds wonderful... 😉
OTOH, just asking, I wonder but maybe a speaker can produce 5th order distortion also ? Like cone breakup modes or else ?
Thanks for your reply, you made a good point there.
simon5 said:
I don't no how accurate the number is, but it is in the ballpark of how well our ear/brain can detect something that it hasn't encountered in the million years or so we have been evolving.
The high order feedback is usually casued by feedback, so a servo woofer could have higher order products. The same mechanism is why a simple no feedback tube amp can sound so good despite having high (single number) THD.
dave
simon5 said:
You can beat that one to death and never come up with an answer. No point in getting nuerotic about it (ie the "audiophile" that is never happy with his system)
I think it is more important to be able to enjoy the music and get an emotional connection with it.
dave
Emotions ? Aren't we all macho men AFAIK ? 😀
By the way, at the recording stage, could the recording already contains that 5th order distortion ? (Microphone pre-amp, and else) It would mean that we are all doomed from the start, no ?
That whole debate is very interesting, I guess I will have to do my own research about it someday because everyone has their own opinion about it.
Pro-Against Feedback, Pro-Against Tubes, Pro-Against Discrete, Pro-Against ICs, Pro-Against tuning with measurements, Pro-Against tuning with ears, Pro-Against exotic wires, Pro-Against exotic XOs components, Pro-Against exotic vocabulary to explain sound... etc. All those topics with little physics/science behind them but so much talk.
By the way, at the recording stage, could the recording already contains that 5th order distortion ? (Microphone pre-amp, and else) It would mean that we are all doomed from the start, no ?
That whole debate is very interesting, I guess I will have to do my own research about it someday because everyone has their own opinion about it.
Pro-Against Feedback, Pro-Against Tubes, Pro-Against Discrete, Pro-Against ICs, Pro-Against tuning with measurements, Pro-Against tuning with ears, Pro-Against exotic wires, Pro-Against exotic XOs components, Pro-Against exotic vocabulary to explain sound... etc. All those topics with little physics/science behind them but so much talk.
Some time ago I tried the "full power test" with my 3886 based Gainclone ala P-A Sjöström.
Look here: http://www.keld.eu/gainclone/test.htm
Look here: http://www.keld.eu/gainclone/test.htm
The question is: Did you guys build yourself a pair of 25.000usd pair of monoblocks to compare sound?
🙂
🙂
simon5 said:
Indeed. What happens -- at least here -- is that the recordings that are poorly done, with poor equipment get listened to less.
dave
Hi
This is the one I build: http://sjostromaudio.com/joomla/index.php?option=com_content&task=view&id=65&Itemid=27
Pics here: http://www.keld.eu/gainclone/
Its non inverting with buffers and DC servo.
This is the one I build: http://sjostromaudio.com/joomla/index.php?option=com_content&task=view&id=65&Itemid=27
Pics here: http://www.keld.eu/gainclone/
Its non inverting with buffers and DC servo.
- Status
- Not open for further replies.