I have found the datasheet. I like those blocks, by the looks at least but also by the performance.hjelm said:
http://www.epcos.com/inf/20/20/db/fc_05/MKT_B32560_64.pdf
The resonance is at 1.2 MHz which is enourmously higher that a normal cap... approx times 2.
moving_electron said:
Not to throw the fat into the fire, but -- specifically, what not to do is regulate the power supply of an amp chip like the overture series ! Especially not with an off-the-shelf regulator like the LM338. The LM338 which Carlos uses has, as expected, a rising impedance curve with frequency. The LM338 is also noisy compared to a discrete regulator, and the cost savings (vs discretes) is de minimis. Douglas Self provides a litany of reasons for not regulating a power amplifier -- and his books are readily available from Amazon.
A regulated power supply for a GC may provide some psychological compensation for the builder, however, so would a glass of porto...
Konnichiwa,
!!!???
Based on the datasheet from Epcos that where it should fall (approximatly). If you want to complain, complain to them.
Them's the ones.
Actually, I prefer the right part for the right job. So, wherever they are apropriate I use them. There is no "magic" part that is equally suited everywhere. These Cap's are good for Audio Application that need good HF behaviour and better linearity than ceramics. I's all a matter of application.
No idea. Batteries are pretty decent out to around 5 - 10KHz, then the impedance goes up, that is a lot of inductance (relatively speaking). We need to bypass the battery to make the amplifier stable, it's as simple as that. A bigger capacitor shifts the frequency where the impedance peak occours downwards AND reduces the Q of the tank circuit, it an attendant much lower peak impedance.
Modelling the battery as 0.001R/1uH and including 1m heavyish duty 50R CI coax cable as lumped 0.15R/250nH/100pF (including contatct resistance) my bypass suggests a resonance at around 65KHz with a peak impedance of a little over 1R, snubbers show little/no effect on Z, Phase otation of current vs. voltage looks to be no more than +80/-60 degrees.
BTW PSU impedance should remain low well past 10MHz.
Sayonara
peranders said:
!!!???
peranders said:
Based on the datasheet from Epcos that where it should fall (approximatly). If you want to complain, complain to them.
yldouright said:
Them's the ones.
yldouright said:
Actually, I prefer the right part for the right job. So, wherever they are apropriate I use them. There is no "magic" part that is equally suited everywhere. These Cap's are good for Audio Application that need good HF behaviour and better linearity than ceramics. I's all a matter of application.
yldouright said:
No idea. Batteries are pretty decent out to around 5 - 10KHz, then the impedance goes up, that is a lot of inductance (relatively speaking). We need to bypass the battery to make the amplifier stable, it's as simple as that. A bigger capacitor shifts the frequency where the impedance peak occours downwards AND reduces the Q of the tank circuit, it an attendant much lower peak impedance.
Modelling the battery as 0.001R/1uH and including 1m heavyish duty 50R CI coax cable as lumped 0.15R/250nH/100pF (including contatct resistance) my bypass suggests a resonance at around 65KHz with a peak impedance of a little over 1R, snubbers show little/no effect on Z, Phase otation of current vs. voltage looks to be no more than +80/-60 degrees.
BTW PSU impedance should remain low well past 10MHz.
Sayonara
Konnichiwa,
Well, I though you said all stacked film MKT's have a 90KHz resonance?
Also, to double the frequency you need to quarter the ESL, plus at the time (early 90's) I could not find much else that offered similar performance.
Sayonara
peranders said:
Well, I though you said all stacked film MKT's have a 90KHz resonance?
Also, to double the frequency you need to quarter the ESL, plus at the time (early 90's) I could not find much else that offered similar performance.
Sayonara
Hi guys!
I was just over at my brother's to compare gainclones. He has a normal nigc with 1000uf caps. His speakers (sensitive monitors) and source are much better than mine.
He has been very pleased with the amp ever since I built it for him, and he did not think that there was anything wrong with the bass either.
We droped in my amp to see if we could note any degradation of the treble and upper midrange.
There was a degradation, however very slight and almost not hearable. Then we tested his amp again with some strings and drums, and what we found were that the 1000uf gc dropped much treble detail when the drums were beating.
This did not occur with my snubber 40000uf in total gc.
We decided to get hi-cap psu parts for him.
I was just over at my brother's to compare gainclones. He has a normal nigc with 1000uf caps. His speakers (sensitive monitors) and source are much better than mine.
He has been very pleased with the amp ever since I built it for him, and he did not think that there was anything wrong with the bass either.
We droped in my amp to see if we could note any degradation of the treble and upper midrange.
There was a degradation, however very slight and almost not hearable. Then we tested his amp again with some strings and drums, and what we found were that the 1000uf gc dropped much treble detail when the drums were beating.
This did not occur with my snubber 40000uf in total gc.
We decided to get hi-cap psu parts for him.
postpunk said:
That's what happens with the low capacitance PSU, even with "claimed" easy to drive speakers.
I've been telling this for long...
postpunk said:
But that is too much, anyway.
I don't see the point of going higher than around 20,000 uf.
Between 10,000uf and 20,000uf is fine, in almost every possible implementation.
postpunk said:
Remove half of the capacitance from your PSU and give the caps to your brother.😎
wima has something to say about snubbers and their caps: too http://www.wima.com/igbt.htm
another wima link: http://www.wima.com/applicguide.htm just another application guide. a good one though.
note that most of us diyers don;t have to cope with high voltage spikes (if, that is, your fridge isn;t on the same group as your audio stuff 😉)
another wima link: http://www.wima.com/applicguide.htm just another application guide. a good one though.
note that most of us diyers don;t have to cope with high voltage spikes (if, that is, your fridge isn;t on the same group as your audio stuff 😉)
I definately plan on implementing the snubbers when I build my BrianGT gainclone, but I'm a little unsure of exactly which design is "correct". As far as I can tell the schematic attached at the bottom is the one most recently updated by CarlosFM. I've heard some people talking about "100uF" caps and "100nF" caps, are the capacitors listed in this schematic as "100nF" correct or should they be "100uF"? Also, Kuei has been recommending a particular Epcos capacitor, should this capacitor be used in the "100nf"/"100uF" position, or is he recommending it for a different position? I guess it would help if someone could just very clearly spell out what the best implementation of this snubber is (as in a known good schematic with all values correct) and also any recommended types or brands for the particular components.
http://www.diyaudio.com/forums/attachment.php?s=&postid=533226&stamp=1103055025
http://www.diyaudio.com/forums/attachment.php?s=&postid=533226&stamp=1103055025
XELB said:Can you control deep bass with only 20,000uF ?
That's new for me.....
This is all new for you.😀
Hmmmm, CarlosFM once again started a long discussion!
Not that he started this thread,,,,, but still!!!!
I think of that guy as being really dedicated, seeking for perfection
all the time! He does actually come up with some really usefull things, now and then. Not many of us DIY'ers put that much effort into our work😉 I like that persistance, and try the ideas for myself.
I am not a scientist, like most of DIY'ers are not, and do not need scientific
evidence to try an idea😉 If it sounds good, it has got to be good!
If it doesnt, unplug the blessed thin😉
I am asking just one thing: Why does some of you guys want to
"ground" the snubbered high cap PSU😉 😉
Steen.
Not that he started this thread,,,,, but still!!!!
I think of that guy as being really dedicated, seeking for perfection
all the time! He does actually come up with some really usefull things, now and then. Not many of us DIY'ers put that much effort into our work😉 I like that persistance, and try the ideas for myself.
I am not a scientist, like most of DIY'ers are not, and do not need scientific
evidence to try an idea😉 If it sounds good, it has got to be good!
If it doesnt, unplug the blessed thin😉
I am asking just one thing: Why does some of you guys want to
"ground" the snubbered high cap PSU😉 😉
Steen.
I'm having a heck of a time identifying a decent part number for a 1R2W or 2R2W resistor....
I can come up with some Ohmite stuff at 5 or 7 bucks but surely this is not what we are talking about here...
Got a Mouser, Newark or RS number?
I can come up with some Ohmite stuff at 5 or 7 bucks but surely this is not what we are talking about here...
Got a Mouser, Newark or RS number?
jackinnj said:
Course people seem to really like the regulated JLH amp.
Regarding the regulated supply impedence rises with frequency but in absolute terms it is lower than the snubberized spice simulation up to 1MhZ.
Early in this thread I wrote:
"I did not find much on the National site but the STMicroelectronics datasheet of their TO-3 packaged LM338 has an interesting impedence vs. frequency graph. The output impedance falls from 0 to 100Hz and then rises substantially (at least in relative terms) by several orders of magnitude until it reaches 1 Ohm at 1Mhz.
Here is the link to the datasheet. The graph is on page 5:
http://www.st.com/stonline/books/pdf/docs/2576.pdf "
Does not say what happens above 1Mhz though.
In another post I refered to the
It indicates that a typical voltage regulator could be modeled as having 1uH to 2 uH of inductance is series with the output.
In Kuei Yang Wang's spice analysis of carlosfm's circuit he wrote "Finally, the bypassing/snubber I personally would suggest is shown in royal blue. We start again from 64mOhm at 10KHz, our impedance peaks with 1 Ohm at 140KHz and stays low to above and beyond 10MHz, assuming the 1uF bypass is placed close to the Amplifier Chip."
So is the regulator's impedence of 1 Ohm at 1Mhz really that bad? It is better in the 140Hz to 1Mhz region and then presumably becomes worse than the snubberized.
Having impedence lower in the 0-1Mhz region could be beneficial could it not?
chipco3434 said:
What about the Mills MRA-5 non-inductive wirewound resistors at
www.percyaudio.com ?
They are relatively small even though they are 5w. $2.15 ea but there is a minimum order of $25.00.
"Non inductive, wire wound..."
I guess that's really the information I need and will ease my search process. Thanks.
"Tyranny of 52%..." You got that right!
I guess that's really the information I need and will ease my search process. Thanks.
"Tyranny of 52%..." You got that right!
curva said:
Perhaps, but I think it is best not to read too much into skepticism. Scepticism, and not taking things at face value is one way we get to the underlying reasons of why something works and identifying when it works and when it or something slightly different does not work.
I have faith that the technique works and when I try it I will like it. But knowing why it works is important to understanding how to make it work in instances where the components are not exactly the same. Or to identify an even better combination. It is possible that carlosfm found a very beneficial combination but that it is a "local" maximum and that there is a better optimum out there that can be found if the theory is understood well by more people.
An intellectual discussion of "why should this be so" or "I don't see how that can be the case since..." is an excellent way to move the knowledge forward - as long as folks resist the very human tendency to feel it is a personal attack.
Many, many scientific discussions follow this same discussion path as this thread. Some leads to a new techniques or theory. Sometimes it is found that an imperial discovery does not generalize particularly well.
So whether some folks discussion the “why” of carlosfm’s “snubberizing” effect have not tried it yet is not particularly relevant. If that were able to try it and found positive results, they would still asking/discussing the same questions and exploring the same areas.
Very likely😉
On the other hand, they would never have discovered this, anyway. Being too shortsighted.
I presume this is ultimately about music, isn't it??
Please correct me if being wrong here!
So, what I mean is: try the snubbered thing for yourself
If your music sounds better, well great.
If not, unplug...... Hate to repeat myself, but this thread invites me to😱
Still I have a feeling this thread was started with a nother purpose!!
Namely to gather a lynchmob, to hang CarlosFM!!!!
That is just too bad. CarlosFM just wants the best for he's
fellow DIY'ers, so he feel an urge to share his findings.
And the findings of he's, are just great in my EARS😀
To the other guys, please stay on thread, please
Steen.
- Status
- Not open for further replies.