Well, who's going to spec an SMPS not designed for audio?
If you buy a merchant supply off the shelf then all your points have merit. But, I think to do so is to undersell the power delivery capability of the technology.
It's easy to get an SMPSU to start into a very large cap load - you do it with the soft start circuitry. We did a PSU for a large Japanese Brand (yes, that one) when I was in Tokyo. 10 0000 uF no problemo.
But, as I said earlier, not my choice.
Everyone that does not recognise the problem of peak transient speaker current demand.
Which possibly describes 99% of our Membership.
The remaining 1% are aware, but most of them (including me) don't know how to specify an SMPS for that demanding transient duty.
Is it just fortunate that the conventional PSU with adequate smoothing does this, without the Membership having to think?
I don't understand the question. I'll have to leave it to the Forum experts.
dynamic power waveform is some power higher than rated steady state for some length of time or duty cycle.
I think most of the real Pro-audio class D,G,H amplifiers have figured out how to spec the smps sections to handle the massive power requirements that they are needing to handle. I'm not talking about the Guitar Center types of semi-pro equipment but the real professional models used daily in high powered PA applications. Perhaps that would be an area to investigate if someone wants to build a proper smps power supply for a power amplifier whether class A, AB, D or whatever?
Infinia,
I think one of the problems is that most smps power supplies are made for computer applications. I don't see where those would ever have the dynamic load variation that we would see in a power amplifier? For a preamp there are probably not the same issues as a power amp but that would appear to be the issue we are talking about. I don't know if just hanging more capacitance after the smps would take care of this, that is for the smart people here to answer.
I think one of the problems is that most smps power supplies are made for computer applications. I don't see where those would ever have the dynamic load variation that we would see in a power amplifier? For a preamp there are probably not the same issues as a power amp but that would appear to be the issue we are talking about. I don't know if just hanging more capacitance after the smps would take care of this, that is for the smart people here to answer.
SMPS are more than just "computer apps'
all I am just saying what if we could spec a SMPS just for audio what would it be, without arbitrarily crippling a regulated design with an unreasonable approach. trading off SMPS loop BW, stepped load response, current limit time shaping etc
all I am just saying what if we could spec a SMPS just for audio what would it be, without arbitrarily crippling a regulated design with an unreasonable approach. trading off SMPS loop BW, stepped load response, current limit time shaping etc
Last edited:
Fully agree Bob. Some SMPSs have a quite agressive fold-back current limiting and I have seen some that just shut down with a large reservoir cap and stay down. Not a good career move
Jan
Okay Infinia,
I have no intention of jumping onto the class D type amplifier at this moment in time. I am still not convinced that those applications have matured enough for a wideband application yet. Perhaps for power hungry subwoofers it is a good solution but that is where I will leave that implementation for now.
I have no intention of jumping onto the class D type amplifier at this moment in time. I am still not convinced that those applications have matured enough for a wideband application yet. Perhaps for power hungry subwoofers it is a good solution but that is where I will leave that implementation for now.
But most available off the shelf are.
If someone knows of a series of supplies that can tolerate traditional sized caps, please let everyone know. If the cost is less than a big honking transformer, maybe it is a way forward. Maybe a great project for someone; supply us with a scaleable design in say, 100W increments?
(Just wait, someone will build a class A amp and use a SMPS to claim they are going green)
Well, they can't drive worth a lick in the sunny dry weather here either! If I was still in Boulder, yea, spring t-shirt skiing. (The ladies dressed a bit differently
Read the bit on active PFC. It is looking at the problem from the SMPS input issue as to not cause issues back on the mains. Fine, but no help if we are building an amplifier other than in understanding the cheap PS in my computer does not have such devices in it and is putting those harmonics back onto the same branch as one of my amps which is why my computer strip is fed from a big multi-stage pi filter. So this is causing a harmonic noise problem, rather than the phase problem motors cause.
Other than the need for a decent input filter, what does this mean to building a power amp? Big inefficient E-cores still rule? Does the active PFC effect the OUTPUT of the SMPS if we are using it for an amp? The paper does not suggest this.
You're right; the biggest reason for active PFC is the mains interface, first in terms of how the load may adversely affect the mains (good neighbor, governmental regulations, etc), the second how well the load can extract power from the mains.
The effect of PFC on what mains garbage gets into the amplifier, and what garbage from the local SMPS gets into the amplifier is much more subtle and much more dependent on the design of the PFC/SMPS.
In an amplifier with PFC, apart from the mains EMI filter, the PFC is the first line of defense from noise from the mains. It has the potential to attenuate some mains EMI. For example, right after the raw bridge rectifier on the mains, the power passes through an inductor that is an integral part of the dynamic boost converter of the PFC. This inductor can block some EMI, especially if it is designed with that in mind.
The second thing that PFC does is to provide a fairly clean REGULATED DC voltage to the switching DC-DC converter. This converter can be designed to behave better when it does not have to operate over a wide voltage range. Note also that you can put all the LPFing you want on the intermediate DC bus.
But we still have the mains Neutral with a fairly straight path to the isolation transformer. But in an SMPS, the transformer is much smaller, and has less capacitance between primary and secondary, so it is likely less mains EMI will get in this way.
At the end of the day, however, the two switchers - the PFC converter and the isolated converter - need to be designed and laid out so that they create a minimal amount of EMI themselves that will get into the amp.
Cheers,
Bob
capacitors designed to operate with steady, constant load, are not the same cap design for a load that is for use in high pulses or bursts apps. A cap which says it is a Computer cap apps is NOt what you want.... that is a steady constant load. Designed for pulsed power caps are what you want. In some cases a large value film cap in parallel helps in pulsed power apps.
THx-RNMarsh
THx-RNMarsh
Note to doug:
To be fair handed and fair minded, means - to not be so insulting.
I'm about 50 pages in and find the level of insult leveled at the 50% of human race that dares to use the right hemisphere of their brain..well..that's down right illiterate and out of bounds on level that is totally unacceptable.
It's horses for courses.
we have breeding programs for horses. We don't take plough horses and enter them in the triple crown of horse racing.
Breeding counts.
Likewise, we each, as humans, have skill sets and cognitive levels that are different. We even have grading of scholastics, heaven forbid.
In the same way -the EXACT same way- we have different levels of hearing skill and different levels of capacity in the cognitive sense, that is tied to those hearing skills.
And, to top it off, the brain is plastic. If it was not... there would be no humans as we know them and no need for schools, or for this planet.
So grow up and stop being so limited in reach and so insulting to people who can hear those differences.
I mean, **** me Doug..I'm TRYING to read your book and get something out of it...and the crap thrown at half the freaking human race who is not entirely objectivist is ...sickening, at best.
To be fair handed and fair minded, means - to not be so insulting.
I'm about 50 pages in and find the level of insult leveled at the 50% of human race that dares to use the right hemisphere of their brain..well..that's down right illiterate and out of bounds on level that is totally unacceptable.
It's horses for courses.
we have breeding programs for horses. We don't take plough horses and enter them in the triple crown of horse racing.
Breeding counts.
Likewise, we each, as humans, have skill sets and cognitive levels that are different. We even have grading of scholastics, heaven forbid.
In the same way -the EXACT same way- we have different levels of hearing skill and different levels of capacity in the cognitive sense, that is tied to those hearing skills.
And, to top it off, the brain is plastic. If it was not... there would be no humans as we know them and no need for schools, or for this planet.
So grow up and stop being so limited in reach and so insulting to people who can hear those differences.
I mean, **** me Doug..I'm TRYING to read your book and get something out of it...and the crap thrown at half the freaking human race who is not entirely objectivist is ...sickening, at best.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.