This is one reason why all the major PCB design software companies are adding Power Delivery Integrity software to their list of add ons. Went to a seminar on the subject of PDS recently,was very informative. The main conclusion is that decouplers are guesed at and generaly overcompensated for (ie to many of them).
Hi, I'll give you results of what I hear (it's the only important thing!) and don't care about the measure!
Parralleling a PSU cap is only good in a CLC or CRC PSU and at its output only: parralleling the input cap will help noise disturbance wich is going out of the rectifiers to cross the PSU and you will hear it!
But in the end after the resistor or better the inductor it will help to open the bandwith in the highs but for the best result you have to use big MKP capacitor about 20 or 22uf is an optimum and there you will hear better highs with more air around. Don't use lilttle values most of the time you'll hear nothing
Have a good day: David
Parralleling a PSU cap is only good in a CLC or CRC PSU and at its output only: parralleling the input cap will help noise disturbance wich is going out of the rectifiers to cross the PSU and you will hear it!
But in the end after the resistor or better the inductor it will help to open the bandwith in the highs but for the best result you have to use big MKP capacitor about 20 or 22uf is an optimum and there you will hear better highs with more air around. Don't use lilttle values most of the time you'll hear nothing
Have a good day: David
Capacitors are useful at the point of load, where they provide a low impedance return path; placing additional ones at intermediate points is therefore superfluous, and can aggravate resonance problems.
A lot of linear power supply PCBs have small film onboard film "decoupling" caps close to the output pins of the board, which is said to improve transient response of the regulators.
Wouldn't these be entirely useless and aggravate resonance problems once run to another amplifier board with (relatively) long wires?
I am thinking that these output bypass should most likely be removed since they wouldn't be doing anything for the load.
Wouldn't these be entirely useless and aggravate resonance problems once run to another amplifier board with (relatively) long wires?
I am thinking that these output bypass should most likely be removed since they wouldn't be doing anything for the load.
thanks Elvee, probably my specific question would be better illustrated with a picture of the board. i'll try and rustle one up, youve pretty much answered my question though so thanks. i didnt place a heap of different places on the board, just the reference bypass and regulator output caps called for in the datasheet (a polymer and np0, although the DS calls for a slightly higher ESR of ~12mOhm) but i also put optional 1206 size pads at the output pins for each regulator in case i wanted to power something that didnt have its own decoupling and thus would have a cap closer to the point of load. nothing is very far away in this build though, as its contained in a 120mm x 30mm x 90mm area
all devices that are actually on the pcb that need power just have local decoupling and are only 20-30mm from the regulator
all devices that are actually on the pcb that need power just have local decoupling and are only 20-30mm from the regulator
Last edited:
If there is a substantial wiring downstream they are indeed useless.
If the wiring is clean, with the supply and return wires tied or twisted together, the resistance will dominate and they will be harmless.
If the wiring is cobweb style, as it is advised by some, the effect will be disastrous.
Indeed, and in addtion they will stress the stability of the regulator.
Thats a scientific aproach to design and improving audio, hope you dont design real world products.
Both are improtant, measurements being the more important as they give us data and results that are non ambigous, repeatable for diferrent circuitry etc.
I'll go even firther, a power supply is there to create a stable DC voltage, you cant hear DC, its DC... If you can hear your power supply then its a BAD design.
Last edited:
Ah, that's easy:
For simple conductor geometries (e.g. rectangular, circular, or other shapes of cylinders):
R = rho x L / A
where
L is length in meters (m),
A is cross-sectional area in m^2 (square meters, i.e. meters squared),
rho is the Resistivity of the conductor material in Ohm-meters (There are tables, on the web, with the resistivities of lots of different materials). Note that rho will vary with temperature.
At room temperature, rho of copper is 1.7 x 10^-8 Ohm meters.
At room temperature, rho of aluminum is 2.6 x 10^-8 Ohm meters.
From the equation for R, it is obvious that to get a higher resistance you could use a smaller cross-sectional area or a larger length, or change the conductor material. Knowing rho and the length, for example, you would calculate the cross-sectional area needed to give a desired resistance. Then, knowing the copper thickness of your PCB layer, you can calculate the trace-width that would give the desired cross-sectional area and thus the desired resistance for the given length. Alternatively, you could solve for the length and adjust that instead.
You will probably also want to google "trace inductance calculat*".
Cheers,
Tom
Last edited:
Well, even though the power supply VOLTAGE is "DC" (actually, we often hope for a "FIXED" DC voltage, i.e. unchanging), the power supply CURRENT should be almost exactly the music signal(!), if the load impedance is a fixed value (no, it's usually not, but my point remains valid).
With that in mind, everyone can see that the primary "signal path" for what you hear from your speakers is the current through the power supply, then through the active power devices in the amplifier, then through the speakers and back through the ground returns to the power supply. It's all about the current. The power supply voltages are only there to enable the current to flow whenever and however commanded/allowed.
At the same time, we can see that the input "signal path" ENDS at the active power amplification devices. It just stops, there! It controls the active power devices (e.g. power transistors, chipamps, etc), which can be thought of as current-modulating "valves", which allow precise currents to be pushed or pulled through them, because of the existence of the power supply voltages.
The signal that actually makes the sound is the large currents that go through the power supply, the active power amplification devices, and the speakers.
So yes, every component in the power supply is in the main sound-signal path, and could possibly affect the sound quality even more-directly than the components in the input "signal path".
And I'll go further: (This is just one of my "pet peeves" but I'm "on a roll", now: ) Forget the myth that shunt components are "not in the signal path" like series components are. That's simplistic nonsense.
Sorry to have blathered-on about all of that, for so long!
Cheers,
Tom
Last edited:
Sorry Tom, we usually agree on everything, but here I draw a line.
metre is a unit of Length. eg. kilometre is 1000 metres.
meter is an instrument that measures an effect. eg, voltmeter measures voltage.
Input signal does not just stop !!! The input signal circuit has the same rules as all other circuits. Where your power current circuit example listed the route out (flow) and back (return) to the power supply, she source similarly has a flow route and a return route. For current to flow in the signal circuit there must be a return route. This is usually the signal ground. Current must return all the way back to the source.
One of us is having a bad hair day.
Last edited:
Gootee,
read what I said, last sentence...
You cannot design a power supply by ears alone, that is what my point was, I am quite aware of the problems with power supplies...and as I've stated before, the power supply isw the MOST important part of any electronics design, get it right and the rest will go OK, get it wrong and nothing else is gonna work.
BUT when designing a PSU you need to look at the output with a scope, is the rf present, whhat ripple is there etc etc
You cannot hear DC, that is a fact, you can hear noise intereference etc iterposed on the DC but DC is silent, clean and wonderfully unobtainable, even using a battery cannot guarantee noise free DC, so PSU's are designed to WORK in the equipement they are designed for, if they dont as I said before its a BAD design.
My pet peeve is people designing power supplies with their ears...
Again with Andrew, signal has a return path
read what I said, last sentence...
You cannot design a power supply by ears alone, that is what my point was, I am quite aware of the problems with power supplies...and as I've stated before, the power supply isw the MOST important part of any electronics design, get it right and the rest will go OK, get it wrong and nothing else is gonna work.
BUT when designing a PSU you need to look at the output with a scope, is the rf present, whhat ripple is there etc etc
You cannot hear DC, that is a fact, you can hear noise intereference etc iterposed on the DC but DC is silent, clean and wonderfully unobtainable, even using a battery cannot guarantee noise free DC, so PSU's are designed to WORK in the equipement they are designed for, if they dont as I said before its a BAD design.
My pet peeve is people designing power supplies with their ears...
Again with Andrew, signal has a return path
Well , I'm afraid it does a little bit yes....
From Wikipedia:
Metre is used as the standard spelling of the metric unit for length in all English-speaking nations except the USA, which uses meter.[16]
The most recent official brochure, written in 2006, about the International System of Units (SI), Bureau international des poids et mesures, was written in French by the International Bureau of Weights and Measures. An English translation (using the spelling: metre) is included to make the SI standard "more widely accessible".[17]
In 2008, the U.S. English translation published by the U.S. National Institute of Standards and Technology chose to use meter in accordance with the United States Government Printing Office Style Manual.[18]
Measuring devices (such as parking meter, speedometer) are traditionally spelt "...meter" in all countries.[19] The word "meter", signifying any such device, has the same derivation as the word "metre", denoting the unit of length this article is about.
I think "signal path" is a more useful concept with a restricted definition
while "everything" affects the amplifier performance I dislike the claims that PS Caps are "in the signal path" - when typical amp PSRR # can be better than 60 dB - ie only 1 part singal influence to 1000 parts of PS V change
to claim something with 1 part in a thousand sensitivity to "signal" is "in the signal path" makes the concept have no edxplainatory power, doesn't draw useful distinctions
while "everything" affects the amplifier performance I dislike the claims that PS Caps are "in the signal path" - when typical amp PSRR # can be better than 60 dB - ie only 1 part singal influence to 1000 parts of PS V change
to claim something with 1 part in a thousand sensitivity to "signal" is "in the signal path" makes the concept have no edxplainatory power, doesn't draw useful distinctions
Last edited:
Perhaps I'm missing something here (many pages to read again), but why on earth is the above information/link taken into consideration, or considered fanatastic emperical data not requiring any further discussion ?
This test is between 300 kHz and 100 MHz (....), just what is the relevance of this test to audio equipment? The pics may look nice, but relevant? Also, large power caps behave much worse below 20 kHz than the small caps tested in the above test, like I measured in this thread. There is much more to caps then only impedance, such as knowing the capacitor's own resonance frequency (where its ESR and impedance equals), before adding a bypass (or not, that I leave in the middle).
Just my 5 eurocents (for as long they still exist).
Again, you're only thinking about VOLTAGE, which is what the PSRR is concerned with. That's all pretty-much irrelevant in this context.
The SIGNAL that eventually drives the speakers is CURRENT, straight out of (or into) the PSU (and then back). And I believe that much or even most of that current has to come out of capacitors, correct? (or else they are being topped up at the same time as the rectifiers are trying to supply it directly). The local bypass/decoupling capacitors should be properly specified to perform much of the job, especially for transients but depending on how everything was spec'd out and implemented, large portions of the current could come from the main PSU caps, or (I imagine) even directly from the rectifier bridge at times.
NO the SIGNAL is actualy voltage. An amp is a voltage device, x volts in gives y volts out regardless of the frequency or load. And y volts out no matter what the current draw.
Try feeding your speakers with a 100ma sine wave and sweep the freq, you will not get a constant SPL output like you should from a 100mv sine wave.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.