FWIW I would find a couple-three (1 channel per driver) switchmode chip amps probably on eBay, power output scaled (if reasonable) to the actual needs of the respective drivers, and stitch them together with a MiniDSP.
OTOH, I have already have three commercial examples of that in service - the JBL Pro LSR 308s. They cost me far less than the parts for a comparable set of parts to design my own. I've looked inside the box, and they are pretty close to what I'd do if it were up to me.
The time to build them will be reinvested where the economics go the other way, namely the subwoofers.
Well I received a first response from MiniDsp today and though they will gladly talk to me the basic answer is they really don't do design work anymore. I guess they are busy selling their commercial offerings and don't need to do that anymore. I will run my ideas by them anyway, perhaps they will see something new in what I am asking that they could add to their offerings.
RNM,
When I initially looked for smps inline power supplies the output power output was rather low. It appears that has somewhat changed today and some higher wattage power supplied with higher voltage output seem to be available. I'm not competent enough to spec a smps that would be applicable to audio and not something like a steady state device like a computer, the demands aren't really the same for instantaneous power. It would make things much easier, they would already be UL, CE approved as a power supply and it would keep my internal space inside the enclosure free for Air rather than a transformer and large caps, a real advantage to me. So just the actual active circuitry would be inside and mounted hanging off that heatsink, easy to design a simple mechanically sound mounting system that would make servicing easy as you would just remove the heatsink as an assembly with everything attached. As a few have pointed out the circuit designer would have to be adept at wireless design and the standards needed for a legal product, don't forget this won't be inside a metallic enclosure so no shielding to speak off though I could have at a cost the interior of the enclosure metal sprayed. I'd rather not do that though as it does get expensive to have to metal spray a plastic part, been there and done that.
I know I am not trying to reinvent anything really, just put all the parts together.
Richard that has been a great thing you have done for that family. If people could actually do that and give money directly to people rather than through most charities who gobble up the majority of the money to operate, not much actually gets to where we want it to. Enjoy that families successes and hopefully they will have a way to put all that effort to their own benefits and help others around them at the same time from your generosity.
RNM,
When I initially looked for smps inline power supplies the output power output was rather low. It appears that has somewhat changed today and some higher wattage power supplied with higher voltage output seem to be available. I'm not competent enough to spec a smps that would be applicable to audio and not something like a steady state device like a computer, the demands aren't really the same for instantaneous power. It would make things much easier, they would already be UL, CE approved as a power supply and it would keep my internal space inside the enclosure free for Air rather than a transformer and large caps, a real advantage to me. So just the actual active circuitry would be inside and mounted hanging off that heatsink, easy to design a simple mechanically sound mounting system that would make servicing easy as you would just remove the heatsink as an assembly with everything attached. As a few have pointed out the circuit designer would have to be adept at wireless design and the standards needed for a legal product, don't forget this won't be inside a metallic enclosure so no shielding to speak off though I could have at a cost the interior of the enclosure metal sprayed. I'd rather not do that though as it does get expensive to have to metal spray a plastic part, been there and done that.
I know I am not trying to reinvent anything really, just put all the parts together.
Richard that has been a great thing you have done for that family. If people could actually do that and give money directly to people rather than through most charities who gobble up the majority of the money to operate, not much actually gets to where we want it to. Enjoy that families successes and hopefully they will have a way to put all that effort to their own benefits and help others around them at the same time from your generosity.
The quality and performance of smps's has improved markedly. Early incarnations were not ready for prime time in a big way. But for some time now smaller current supplies have been working quite well in products and now higher current supplies are working well for large power amps.
Finding one off the shelf for your needs of course is full of the usual trials and tribulations. But Newark, Digikey, Arrow and Mouser are full of potential resources.
Alan
Finding one off the shelf for your needs of course is full of the usual trials and tribulations. But Newark, Digikey, Arrow and Mouser are full of potential resources.
Alan
Waltzingbear,
Yes I remember years ago looking for an inline power supply to do an amplifier and the maximum wattage was rather low. Now I've been told that in-order to work well in an audio amplifier situation added capacitance may be needed for the instantaneous needs driving a power amplifier. Perhaps that is just bs, I don't know enough to argue the point.
I'm thinking since there are some real great colleges here in California perhaps I can find a graduate student who would want to take this design project on. I see at UCSD they have a high level DSP program. Thinking of contacting one of the professors to see if they have a favorite graduate student that may be interested.
Yes I remember years ago looking for an inline power supply to do an amplifier and the maximum wattage was rather low. Now I've been told that in-order to work well in an audio amplifier situation added capacitance may be needed for the instantaneous needs driving a power amplifier. Perhaps that is just bs, I don't know enough to argue the point.
I'm thinking since there are some real great colleges here in California perhaps I can find a graduate student who would want to take this design project on. I see at UCSD they have a high level DSP program. Thinking of contacting one of the professors to see if they have a favorite graduate student that may be interested.
If you are using single-ended amps the rail pumping effects can be alleviated with added capacitance. If bridge-tied load amps you may not require it.
I believe Hypex has a switchmode supply with some extra C on the output, or possibly something more than that.
Thanks Brad.
I'll take a look at Hypex power supplies and see what they have. Not sure about their pricing model though.
I just tried to send a letter to a Prof. at UCSD and of course the listed email address is bouncing back as un-deliverable! Now I have to hunt down his correct email address.
I'll take a look at Hypex power supplies and see what they have. Not sure about their pricing model though.
I just tried to send a letter to a Prof. at UCSD and of course the listed email address is bouncing back as un-deliverable! Now I have to hunt down his correct email address.
Last edited:
Derfnofred,
That would be very nice and helpful. I just sent a letter to a Professor named Dan Sievenpiper. Perhaps there is someone else that would be a better contact for this type of inquiry? You can send me a PM and we can take this out of John's thread though it has been a great discussion so far. Hopefully others have gotten something out of this discussion.
Stcven
That would be very nice and helpful. I just sent a letter to a Professor named Dan Sievenpiper. Perhaps there is someone else that would be a better contact for this type of inquiry? You can send me a PM and we can take this out of John's thread though it has been a great discussion so far. Hopefully others have gotten something out of this discussion.
Stcven
I recall seeing a blog from Bruno where he added a cap multiplier to one of his smps designs and thought it was slightly more musical. I think it was for his DLCP, a line level dsp card.
I have a pair of the Hypex NC400 amps setup with their smps supply, I see no need for added capacitance on that supply.
One of the advantages of smps supplies is that they can inherently supply large peak currents without sweating. I believe this is one of the reasons we often see positive evaluations of LF from amps with smps supplies. The bugaboo may be in the hf area where depending on lots of factors there is more potential for interaction, ie change of signal. The ability for these supplies to work well at higher frequencies has been one of the big improvements overall.
Alan
I have a pair of the Hypex NC400 amps setup with their smps supply, I see no need for added capacitance on that supply.
One of the advantages of smps supplies is that they can inherently supply large peak currents without sweating. I believe this is one of the reasons we often see positive evaluations of LF from amps with smps supplies. The bugaboo may be in the hf area where depending on lots of factors there is more potential for interaction, ie change of signal. The ability for these supplies to work well at higher frequencies has been one of the big improvements overall.
Alan
Last edited:
As said before, i knew Levitins book and i knew the material he referenced, but unfortunately there is no study included that examined the relationship between the degree of a difference and the ability to transfer the information to long term memory.
Of course it depends on the hypothesis that is under test, but if one is interested in results of practical relevance it is hard to justify the need for rapid switching/comparison when lets say doing a test with amplifiers or cd-players.
Maybe maybe not; but what about the emotional reaction to a complex stimulus (aka music)?
Is it independent of the sample length?
Given that you take the hypothesis that comparison between two stimuli only takes place in STM/working memory and the information begins to fade away within a couple of seconds (neglecting the influence of awareness and different memory paths for the moment), you have to choose a sample length of ~5s .
????
IME it is painfully easy to do a practical test of the potential efficacy of rapid switching. Come up with small audible difference, and try to detect it with fast versus slow switching. Of course you have to account for the effects of learning on detection. So do it with fast switching first, and then last.
Making audio files with small audible differences is painfully easy if you know how to use audio editing software such as the freeware Audacity. They all have built in equalizers, whose equalization can be adjusted as to frequency, bandwidth, and magnitude. Just pick some reasonable frequency and bandwidth, and vary the magnitude from audible to inaudible.
This same sort of thing can be done for frequency response, noise, phase response, nonlinear distortion, polarity, jitter, you name it. And if you have the .wav file for some recording of interest, there's your raw material.
The tool I used way back when for this was a software ABX comparator named PCABX.EXE that I wrote back around Y2K. It had a slider for varying switch over time. The results of varying the position of the slider is IME hard to ignore.
PCABX.EXE can still be downloaded from images of my old PCABX.COM web site on the Internet Archive site. It was written in VB60 so anybody who wants to repeat this experiment has to line up the runtime support for VB60 as well.
OK subjectivists. Nit pick away. I know that nothing is more evil in your sight than an audio issue that is well understood and somewhat settled. ;-)
Meanwhile here's some relevant reading: https://web.archive.org/web/20150627022419/http://www.nousaine.com/pdfs/Flying%20Blind.pdf
Last edited:
I have seen the comment to run two class-D amplifiers in parallel a few times. Are they talking about two separate IC's or are we talking about a single chip with four output used in a single ended configuration? Can two separate class-D chip amps be used in a push pull configuration with one of the chips working in inverse polarity? I have the convenience of picking the speaker impedance and can make it either 4 or 8 ohm, or even higher if there is an advantage.
That's a good trick even with old fashioned AB amps.
Allows use of both rails at once.
Quite a bit more efficient.
I first seen it in Fosgate car amps.
Yes, it doubles the ripple frequency seen by the filter caps IF low frequencies are of equal level in both channels---which they will almost always be, unless there was a mistake in the mixdown.
- Status
- Not open for further replies.