Are you addressing my post? Inductance is in ANY wire. Small caps and small inductance give high freq ringing. It is all well documented, google "tank circuit", I think Thorsten also mentioned it before. Again, one deciding factor (apart from the C and L values of course) is the rate of rise (or decay) of the current, the delta-I/delta-T. This can be seen in amplifier tests with fast square waves. I have a measurement somewhere looking at the amp output with square waves at the amp output point, and also after 6 feet of twin lead. The latter showed a lot of ringing. Of course, for AUDIO this test is irrelevant.
Jan Didden
Jan Didden
Yes, all " things " have inductance, but questions is, how is value of this. If we shall do dissection of all problems, we shall soon on " superstring " level 😉 .
Upupa Epops said:
Sure, but what I mean is that in the sizes we often use in our projects, like several inches of cable, and 10's of pF of capacity, the L effects can readily be seen, IF you exite the circuitry with fast pulses. Fast pulses like a couple of uSec rise/fall times. Don't ask me exactly how many nH that cable is or how many pF that stray capacity, because I really don't know and I am not that good at guessing, but I SEE those effects on the scope. I see it, recognise it for what it is, and do something about it or conclude that it has no relevance to audio reproduction.
Upupa, I am not at all talking about audible differences, I am talking about things anybody with a reasonable scope with say 20 or 30MHz bandwidth can measure.
Jan Didden
Konnichiwa,
Maybe, but a good way to build an oscillator is to attached a pair of Batteries of suitable voltage to an LM3875 Power Op-Amp without applying any additional bypassing.
Sayonara
Upupa Epops said:
Maybe, but a good way to build an oscillator is to attached a pair of Batteries of suitable voltage to an LM3875 Power Op-Amp without applying any additional bypassing.
Sayonara
Konnichiwa,
Up to here we are completely factual.
Hmmm. Now we are in hypothesis, assumption and presumption.
It has been observed that resonances in output transformers in the range above 100KHz and coming only to a few db lift at the resonance frequency seem to induce audible artifacts in the audio range. A friend of mine (A transformer maker) noticed when some people complained about the sound and the cause was eventually traced to a well ultrasonic resonance.
Yup, I know anecdotal evidence is no evidence, but it would be sufficient to at least eleict a "reasonable doubt" about the relevance.
Too many people are by far too quick in shooting from the hip as to what is (potentially) audible and what is not. I notice that an awfull lot is audible and life would be much easier if things behaved like some people wish to assume, sadly they don't.
Sayonara
janneman said:
Up to here we are completely factual.
janneman said:
Hmmm. Now we are in hypothesis, assumption and presumption.
It has been observed that resonances in output transformers in the range above 100KHz and coming only to a few db lift at the resonance frequency seem to induce audible artifacts in the audio range. A friend of mine (A transformer maker) noticed when some people complained about the sound and the cause was eventually traced to a well ultrasonic resonance.
Yup, I know anecdotal evidence is no evidence, but it would be sufficient to at least eleict a "reasonable doubt" about the relevance.
Too many people are by far too quick in shooting from the hip as to what is (potentially) audible and what is not. I notice that an awfull lot is audible and life would be much easier if things behaved like some people wish to assume, sadly they don't.
Sayonara
Hi,
it looks good in practice too. When one is switching 1200A at 800V in less than 2us, every nH of inductance matters. I know what ESL is, believe me.
Since I deal mostly with SMPS, all the caps lying around are typically low ESR ones, and smaller cans measure from 6nH to 20nH. The worst piece I could find was some 10 years old Iskra (local brand, now belly up) general purpose axial can 4700uF/40V. It measures 75nH at 1 and 10MHz.
I have also checked Epcos range of snap in lower voltage caps. All the datasheets where ESL is stated say 20 or 10nH. Now I would really like to know your measuring method and on what samples it was taken. I suppose that high ESL of electrolytics is another audio myth that needs debunking.
Best regards,
Jaka Racman
have you ever watched one oscillate?Kuei Yang Wang said:
i was quite amuzed by the fellow who built a '3875 amplifier without a heat sink -- this will definitely cycle off and on and create all sorts of nasties -- and there is even more interesting phenomena as the SPIKE protection kicks in -- sometimes it hesitates before the thermal protection kicks in and the distortion will go from 0.03% to 10% in a flash. I found that when the additional 10uF supply bypass caps were installed the hesitation went away and the amp can be pumped up to max -- my heat sink is now a wind tunnel, however.
Kuei Yang Wang said:
Nah, I think it's still factual. The fact that square waves into a few feet of zipcord elicit ringing from said cord has no relevance to audio because the conditions of a square wave in the zip cord donot arise with audio signals.
Of course, you can have an amp that is marginally stable or has protection system problems.
Jan Didden
Kuei Yang Wang said:
Although there are a only handful of participants to this discussion, I note over 3800 views of the thread. So, I'm guessing that there are quite a few of us observers that appreciate that you did make the effort to present an analysis.
Sheldon
I've already said that this PSU is too simple to be understood or taken seriously.
In fact, the more you speculate, the more you complicate.
I arrived to this schematic in a couple of hours testing and listening.
In my mind, I had previously the feeling that this was the way to go. It was.
You guys can discuss this for years...
Why not test it and listen?
Why not measure it after building it?
I advised the values that work best in most situations, you have a good starting point.
90% of the job is done, isn't it?
Is it hard to try a couple of resistors and caps?
In fact, the more you speculate, the more you complicate.
I arrived to this schematic in a couple of hours testing and listening.
In my mind, I had previously the feeling that this was the way to go. It was.
You guys can discuss this for years...
Why not test it and listen?
Why not measure it after building it?
I advised the values that work best in most situations, you have a good starting point.
90% of the job is done, isn't it?
Is it hard to try a couple of resistors and caps?
I so totally agree!!! Just try it, it is fairly simple to do😉
If you do not like it, (unlikely) go back to your original setup😱
How hard can that be? I really do not get this discussion.
Just do it! You will not be disappointed. I actually made CarlosFM's
PSU for several amps, and it works really sweet😎
Thanks a lot Carlos!
Steen.
carlosfm said:
because hes allready made up his mind that this snubber thingy is not going to help. easy as that.
I was under the impression that this forum was for exchanging good ideas on DIY hifi ? Correct me, if I am wrong!
OK, guys here is one reason to forget about all that "witchhunting"!!! I am just listening to SADE "Diamond life" on vinyl. Great, great stuff😉 That is so good , I almost forgot about all that sh*t, you guys can come up with in your "witchhunting"😉 😉 😉
Sorry about this rather harsh comment!
"In case you don't get it"! The reffered "witch" is , Ironically:
CarlosFM. I really do not get that fact. That nice fellow really did give something great and usefull stuff, to you fellow DIY'ers to try!!
You ought to be thankfull
😉
I feel sorry for you guys, not wanting to try that snupper thing!
Steen.😉
OK, guys here is one reason to forget about all that "witchhunting"!!! I am just listening to SADE "Diamond life" on vinyl. Great, great stuff😉 That is so good , I almost forgot about all that sh*t, you guys can come up with in your "witchhunting"😉 😉 😉
Sorry about this rather harsh comment!
"In case you don't get it"! The reffered "witch" is , Ironically:
CarlosFM. I really do not get that fact. That nice fellow really did give something great and usefull stuff, to you fellow DIY'ers to try!!
You ought to be thankfull
😉
I feel sorry for you guys, not wanting to try that snupper thing!
Steen.😉
Steen,
Nobody is which-hunting. What we were discussing (among other things) is that listening reports from whoever, who has a vested interest, are just too unreliable to conclude: "yes, if I do that, my system will sound better". It may (after all, if you expect it, it most probably will), but there are so many other variables it's just a nice story, not more.
That is exactly the way those multi-kilo dollar cable peddlers, brilliant pebbles, magic pens, whatever, etc make their killing. They insinuate that if you can't hear the enormous improvement with their stuff, you're one of these suckers with wooden ears. And who wants to be that, right?
Now, carlosfm is clearly a victim of this, when he says (re-read that last scentence):
What else can I say? Maybe I should follow the advice of an earlier poster: Forget this thread, and move on. On the other hand, I feel strongly about this. I hate to be manipulated, even by myself.
Jan Didden
Nobody is which-hunting. What we were discussing (among other things) is that listening reports from whoever, who has a vested interest, are just too unreliable to conclude: "yes, if I do that, my system will sound better". It may (after all, if you expect it, it most probably will), but there are so many other variables it's just a nice story, not more.
That is exactly the way those multi-kilo dollar cable peddlers, brilliant pebbles, magic pens, whatever, etc make their killing. They insinuate that if you can't hear the enormous improvement with their stuff, you're one of these suckers with wooden ears. And who wants to be that, right?
Now, carlosfm is clearly a victim of this, when he says (re-read that last scentence):
carlosfm said:
What else can I say? Maybe I should follow the advice of an earlier poster: Forget this thread, and move on. On the other hand, I feel strongly about this. I hate to be manipulated, even by myself.
Jan Didden
I don't see why it needs to be listening/empirical needs to conflict with the vs. theoretical+measurement science.
Trying it ones own system of course has value because as pointed out you can get sound you like better.
Understanding the theory/science can uncover improvements and ways to make better guesses/approximations for better places to start with particular components and layouts. It provides the springboard.
So seeking to know why the CarlosFM schematic worked for him and others is a really good thing. It does not imply a disbelief in the result by the participants. (Though some may and some may not.)
Trying it ones own system of course has value because as pointed out you can get sound you like better.
Understanding the theory/science can uncover improvements and ways to make better guesses/approximations for better places to start with particular components and layouts. It provides the springboard.
So seeking to know why the CarlosFM schematic worked for him and others is a really good thing. It does not imply a disbelief in the result by the participants. (Though some may and some may not.)
steenoe said:
Hi Steen,
It is people like you that encourage me to post and share my findings with the other diyers.😎
Unfortunately, my life seems much more peaceful if I keep my things to myself...
janneman said:
😕
Jan, I've been very quiet lately.
But it seems that it is not enough.
Did I open this thread?
Seriously, what's all this fuss about?
Don't like it, don't eat it.
EDIT: Jan, I see... you didn't get it.
I'm going to quote myself and insert two new words, to make it clear:
carlosfm said:
Clear?
carlosfm said:
Hi Steen,
It is people like you that encourage me to post and share my findings with the other diyers.😎
Unfortunately, my life seems much more peaceful if I keep my things to myself...
😕
Jan, I've been very quiet lately.
But it seems that it is not enough.
Did I open this thread?
Seriously, what's all this fuss about?
Don't like it, don't eat it.
OK Carlos, let's just agree to disagree, then. Probably for ever. You ARE a very prolific writer, and you provide a lot of interesting things for all of us.
It's just this fundamental thing, that, ohh, forget it. I'm gonne try this wine my girl friend brought me. Then again, maybe I should forget the wine, and try the girlfriend? Nah, I'm civilised, remember!
Jan Didden
Early in this thead there was a question on whether the regulated supply could benefit also from this approach. I don't think it got addressed.
What impedence characteristics are likely/typically added by the regulator (LM338 for example)?
What does this tend to look like to the chipamp?
In CarlosFM's regulated supply values he has a 47nF cap at the output of the LM338 (near the reg.) and a 33uF-100uF cap at the LM3886. I don't think there is a bypass cap on the 33uF-100uF cap at the Chipamp IIR. The power supply before the regulators was transformer>4700uF+4700uF>47uf.
What impedence characteristics are likely/typically added by the regulator (LM338 for example)?
What does this tend to look like to the chipamp?
In CarlosFM's regulated supply values he has a 47nF cap at the output of the LM338 (near the reg.) and a 33uF-100uF cap at the LM3886. I don't think there is a bypass cap on the 33uF-100uF cap at the Chipamp IIR. The power supply before the regulators was transformer>4700uF+4700uF>47uf.
- Status
- Not open for further replies.