Other examples are one pass amp and aural exciter.
The rising thd with frequency and volume is sought after by a lot of ears.
And by warming the chip the hf thd gets over the masking threshold of normal ears.
I have read somewhere that some prefer 2. Harmonics,(tubes). Other 3. Harmonics. (Transistors)
(Or it is the thd under the masking threshold and the whole thing only placebo?)
The rising thd with frequency and volume is sought after by a lot of ears.
And by warming the chip the hf thd gets over the masking threshold of normal ears.
I have read somewhere that some prefer 2. Harmonics,(tubes). Other 3. Harmonics. (Transistors)
(Or it is the thd under the masking threshold and the whole thing only placebo?)
This has to be the case seeing as it's the only measurable difference. There are a lot of people who like the sound of harmonic distortion.
A long term solution might be to introduce THD in a preamp section so you don't have to stress your amplifier which will shorten it's life/put your speakers in danger of DC in a worst case scenario.
Last edited:
FRESHNESS. MARCE FRESHNESS
marce
Thank you asking me the right question .
What better I get by heating It?
This IC amp delivers the freshest sound ever ,only if you make it run at HELL° temperature.
OPTIONAL READING
By walking down a street,you encounter a loudspeaker hanging on a wall, diffusing a program.You can without doubt descern whether the sound is live (fresh) or playback (decayed),without having any idea about its source. The simpler is the circuit ,fresher remains the sound.It has been shown that, even silver solder decays less than ordinary one.There is no any scientific method to evaluate or qualify it,but personelly I associate the sound freshness with the french bread the " la baguette" better known as sandwich bread. From all hot ,to green rotten, it can have its equivalent sound, at all stages.This bread ,when ultra fresh, is cracky or say crispy, but for very short time, after, it becomes soft.It is as this cracky stage ,what an amplifier for me should sound like.Whereas most amps are designed to sound like soft industrial type bread .As is the case with food ,once the sound loses it's freshness nothing can bring it back,which makes it the most precious aspect of sound. You should not be aware of this ,so that you can appreciate " rotten caviar" rather then "fresh radish".
Japanese gave the most interest for freshness and presence, thanks to "tea houses".The highlight of these establishments was not the tea but the hifi sound. Being destinat for profesionel use, hence money, encouraged great number of designers to bring horns and single ended amps to the highest level.
Nowadays ,if you need the most freshest amp, it should be the simplest one possible,and is still the same 1927 direct heated two stage triode SE amp.
But there comes the lm3886. Two years ago, I discovered this IC, singing far fresher sound than any of my SE designs, but only at tweeter sounds.Why? It is much more difficult to achieve freshness at highs than at lows.The defect is explained in the picture of the guts of this beetle .The upper lungs are the power transistors and the intestines are the low levels.When the power transistors suddenly go hot ,the heat flows downwards to the colder intestines such as a tsunami ,heating the transistors sequentially .Each transistor heated, gets it's beta enhenced, hence conducts more current ,transforming heat variations into electricals ones, creating by this, internal feedbacks .This can be seen on the bias vs temperature curve which is invers function of the input transistor's beta ,as the curve shows passing from 25°C to 125°C the beta gets enhanced by 50% .To anheal thermal/electrical transformation ,the solution is to heat the die at a temperature about which the betas of all transistors(as all at low level have similar characteristics) doesn't vary any more .The bias curve shows to be 125°C.
But the output transistors now have little Tj margine left.The designer resolved this by creating for this IC only ,a new type of transistor ,unique on planet Earth available to common mortels, that can operate at junction temperature of 250°C and this ,offered with Spike protection,if you please ,so that you can use it safely up to such high temperature.
Now,you tell me ,Why is that? the National Semiconductor Corporation would have (as would define an ex-employee) "shot himself in the foot" investing huge amount of R&D resources for extraordinary high temperature transistor,when they could simply use ordinary ones and achieve identical performance, as OPA's and TDA's do.
Further more, Why no one looks at the table shown ,which calculates the heatsink size.The minimum heatsink temperature preconized by the manufacturer at which it can still deploy it's maximum output power,does not start at 50°C nor 60°C( std. max.) but 90°C MIN.That is the coldest object you must mount this IC on, is a boiler, and can make it singing 80W.
END OPTIONAL READING
Since 15 years ,"experts" were advising diy'ers to run behind a taxi named Gaincard , to arrive the same destination, but, keeping thousands of dollars of taxi fare in their pockets.Now they realize that since 15 years they were running behind a streetcar named desire.
There is still another defect to repair before perfecting THE successor of 300B . Please, be patient.
KOKORIANTZ
marce
Thank you asking me the right question .
What better I get by heating It?
This IC amp delivers the freshest sound ever ,only if you make it run at HELL° temperature.
OPTIONAL READING
By walking down a street,you encounter a loudspeaker hanging on a wall, diffusing a program.You can without doubt descern whether the sound is live (fresh) or playback (decayed),without having any idea about its source. The simpler is the circuit ,fresher remains the sound.It has been shown that, even silver solder decays less than ordinary one.There is no any scientific method to evaluate or qualify it,but personelly I associate the sound freshness with the french bread the " la baguette" better known as sandwich bread. From all hot ,to green rotten, it can have its equivalent sound, at all stages.This bread ,when ultra fresh, is cracky or say crispy, but for very short time, after, it becomes soft.It is as this cracky stage ,what an amplifier for me should sound like.Whereas most amps are designed to sound like soft industrial type bread .As is the case with food ,once the sound loses it's freshness nothing can bring it back,which makes it the most precious aspect of sound. You should not be aware of this ,so that you can appreciate " rotten caviar" rather then "fresh radish".
Japanese gave the most interest for freshness and presence, thanks to "tea houses".The highlight of these establishments was not the tea but the hifi sound. Being destinat for profesionel use, hence money, encouraged great number of designers to bring horns and single ended amps to the highest level.
Nowadays ,if you need the most freshest amp, it should be the simplest one possible,and is still the same 1927 direct heated two stage triode SE amp.
But there comes the lm3886. Two years ago, I discovered this IC, singing far fresher sound than any of my SE designs, but only at tweeter sounds.Why? It is much more difficult to achieve freshness at highs than at lows.The defect is explained in the picture of the guts of this beetle .The upper lungs are the power transistors and the intestines are the low levels.When the power transistors suddenly go hot ,the heat flows downwards to the colder intestines such as a tsunami ,heating the transistors sequentially .Each transistor heated, gets it's beta enhenced, hence conducts more current ,transforming heat variations into electricals ones, creating by this, internal feedbacks .This can be seen on the bias vs temperature curve which is invers function of the input transistor's beta ,as the curve shows passing from 25°C to 125°C the beta gets enhanced by 50% .To anheal thermal/electrical transformation ,the solution is to heat the die at a temperature about which the betas of all transistors(as all at low level have similar characteristics) doesn't vary any more .The bias curve shows to be 125°C.
But the output transistors now have little Tj margine left.The designer resolved this by creating for this IC only ,a new type of transistor ,unique on planet Earth available to common mortels, that can operate at junction temperature of 250°C and this ,offered with Spike protection,if you please ,so that you can use it safely up to such high temperature.
Now,you tell me ,Why is that? the National Semiconductor Corporation would have (as would define an ex-employee) "shot himself in the foot" investing huge amount of R&D resources for extraordinary high temperature transistor,when they could simply use ordinary ones and achieve identical performance, as OPA's and TDA's do.
Further more, Why no one looks at the table shown ,which calculates the heatsink size.The minimum heatsink temperature preconized by the manufacturer at which it can still deploy it's maximum output power,does not start at 50°C nor 60°C( std. max.) but 90°C MIN.That is the coldest object you must mount this IC on, is a boiler, and can make it singing 80W.
END OPTIONAL READING
Since 15 years ,"experts" were advising diy'ers to run behind a taxi named Gaincard , to arrive the same destination, but, keeping thousands of dollars of taxi fare in their pockets.Now they realize that since 15 years they were running behind a streetcar named desire.
There is still another defect to repair before perfecting THE successor of 300B . Please, be patient.
KOKORIANTZ
Attachments
Are you sure ur not reading the meaning of the temperature graphs and descriptions backwards?
So the chip will produce those temperatures when pushed to a certain watt output?
I simply just do not believe in electronics running too hot. I'll stick with my big heatsinks and optional cooling fans.
So the chip will produce those temperatures when pushed to a certain watt output?
I simply just do not believe in electronics running too hot. I'll stick with my big heatsinks and optional cooling fans.
I have Sf300 amp that is based on lm3886. It is heavy modified... it runs quite on high working temperature...around 60-70oC. It is pushed a little into classA and on same heatsink are bipolars as a part of lm317-337 regulator and constant current resistor to push regulator always in on state.
So...if amp is cold it is ok but nothing special. Almost as all 3886 kits that i was listening to. But after 30-40 min when becomes hot it plays damm good. Wider, bigger 3d stage. Very fast. Fine microdynamics. Much more neutral. It plays equaly dynamic from bottom til top. Not pronounced more in midle of sounstage like when it is cold.
I am not dreaming a lot of my friends can say the same. They are suprised how fine it sounds. Because of small child and listening at low level , for me first 10w are important and here i enjoy very much with this amp. For louder listening choose another amp.
It touch my emotions and that is what i have missed with 90% amp that i had. Very simillar was Hiaraga la monstre but ...i missed some wats.
So...if amp is cold it is ok but nothing special. Almost as all 3886 kits that i was listening to. But after 30-40 min when becomes hot it plays damm good. Wider, bigger 3d stage. Very fast. Fine microdynamics. Much more neutral. It plays equaly dynamic from bottom til top. Not pronounced more in midle of sounstage like when it is cold.
I am not dreaming a lot of my friends can say the same. They are suprised how fine it sounds. Because of small child and listening at low level , for me first 10w are important and here i enjoy very much with this amp. For louder listening choose another amp.
It touch my emotions and that is what i have missed with 90% amp that i had. Very simillar was Hiaraga la monstre but ...i missed some wats.
KOKORIANTZ,
I think you misunderstand MTBF and FIT data and how it is derived.
The accelerated life test at 125C is performed to allow a reasonable sample size (15838)to produce usable data in an acceptable amount of time (1000 Hrs).
The failure rate at elevated temperature is used to extrapolate the MTBF and FIT at 55C.
This is the data presented for the component, not the MTBF at 125C.
Furthermore, since there were no failures during the 1000 hour life test, they only have a 60% confidence level.
Testing for longer than 1000 hours might produce different MTBF data.
I think you misunderstand MTBF and FIT data and how it is derived.
The accelerated life test at 125C is performed to allow a reasonable sample size (15838)to produce usable data in an acceptable amount of time (1000 Hrs).
The failure rate at elevated temperature is used to extrapolate the MTBF and FIT at 55C.
This is the data presented for the component, not the MTBF at 125C.
Furthermore, since there were no failures during the 1000 hour life test, they only have a 60% confidence level.
Testing for longer than 1000 hours might produce different MTBF data.
Funny enough, Rth usually decreases with temperature. Not that it is of much help, though
.
Funny enough, Rth usually decreases with temperature. Not that it is of much help, though
I'm guessing you're talking about the channel resistance (RDS(ON)) of a MOSFET. Yeah. That has little relevance for the LM3886 which is all BJT....
Tom
The Gimp
Permit me to inform you the little knowledge I have about MTBF, as once I use it for railway applications.
That's true the tests are executed as I mentioned at 125°C . the working temperature 65° is mentioned to be qualified.About 1400 units were tested at 125°C for 1000 hours with 0 failure. The statistics gives this unrealistic value of 138 million hours.This value will simply participate with MTBF of other components ,To end up reduced to only few years.Exemple ,A power transistor in a crt Tv can have itself million of hours MTBF but the whole TV can be less than five years.
How to determine the MTBF at higher temperature than at which it is qualified.
Imagine the maximum temperature at which the component (can be a bearing or spring in mechenics or capacitor , transistor etc.) is destroyed to be a cliff at which you are standing.Passed the limit your MTBF equals zero.If you stand few centimeters away from the edge your MTBF gets higher but is still very low, as a windgust or inattention you are falling.If you stand a meter away ,your MTBF now sharply enhenced,but no one knows, you wouldn't let your children play so near .As you see the MTBF varies nonlinear from value of qualified to zero.
Each type of component( bearings or steal for springs) have its own statistic curve shows how the MTBF falls to zero.
The lm3886 has thermal shutdown at 165°C This means it' MTBF goes zero (the cliff) about 170°C. Running it at 125 °C instead of 65°C ,as if it is running at 105°C if it had 150°C max.This gives small percentage of MTBF decrease relative to 65°C.
But,admitting that it gets DRAMATIC reduction of 99% .Instead of 138million hours reduced to 1.38 million ,rounding it further to 1million only.If this component is used at average of 1 hour per day ,that is 1000 hours every 3 years,this means ,after 3000 years ,that is in the year 5016, your descendent switches on ,50% chance, it will fail.Of course this is statistics for the IC alone.
Permit me to inform you the little knowledge I have about MTBF, as once I use it for railway applications.
That's true the tests are executed as I mentioned at 125°C . the working temperature 65° is mentioned to be qualified.About 1400 units were tested at 125°C for 1000 hours with 0 failure. The statistics gives this unrealistic value of 138 million hours.This value will simply participate with MTBF of other components ,To end up reduced to only few years.Exemple ,A power transistor in a crt Tv can have itself million of hours MTBF but the whole TV can be less than five years.
How to determine the MTBF at higher temperature than at which it is qualified.
Imagine the maximum temperature at which the component (can be a bearing or spring in mechenics or capacitor , transistor etc.) is destroyed to be a cliff at which you are standing.Passed the limit your MTBF equals zero.If you stand few centimeters away from the edge your MTBF gets higher but is still very low, as a windgust or inattention you are falling.If you stand a meter away ,your MTBF now sharply enhenced,but no one knows, you wouldn't let your children play so near .As you see the MTBF varies nonlinear from value of qualified to zero.
Each type of component( bearings or steal for springs) have its own statistic curve shows how the MTBF falls to zero.
The lm3886 has thermal shutdown at 165°C This means it' MTBF goes zero (the cliff) about 170°C. Running it at 125 °C instead of 65°C ,as if it is running at 105°C if it had 150°C max.This gives small percentage of MTBF decrease relative to 65°C.
But,admitting that it gets DRAMATIC reduction of 99% .Instead of 138million hours reduced to 1.38 million ,rounding it further to 1million only.If this component is used at average of 1 hour per day ,that is 1000 hours every 3 years,this means ,after 3000 years ,that is in the year 5016, your descendent switches on ,50% chance, it will fail.Of course this is statistics for the IC alone.
I understand you and i agree with you. For me it is unimportant if lm3886 failure after 1000 hours if plays way better hot than cold. You buy new one for 10€ and 1 hour of soldering and show goes on. Tubes are also not longlasting and nobody care. A lot of youu give 100€ and more for good signal tubes and much more for power tubes.
- Status
- Not open for further replies.