Class D audio power amplifier research

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I am currently in my final year at college studying Product Design Engineering(from an electronic engineering background). For my final year project i have decided to design and build a working Class D audio power amplifier. However since my course incorparates business related subjects, my project must be designed and developed as if it was by a "real world" company. Therefore, marketing and customer needs must be employed. So before jumping into the technical end (circuit design ..etc) i need to show that there is a need for my Class D Stereo Power Amplifier. I have done some research and i can see that alot of the focus on the Class D topologies are mostly based on mobile applications (phones, portable audio products..etc). The power efficiency of the designs and less heat dissipation makes it appealing for small low powered devices. I have not yet found any information regarding the advantages of Class D audio amplifier design in power amplifiers with respect to high-end audio.

Was wondering if anyone could shed some light on this?
 
Some thoughts....

Some thoughts....

Fist of all why classD ? I would think the main reason would be to improve efficiency. Less loss of power ( as heat ) per watt output. Very important with battery operated equipment ( portables and non portable ) AND even more so now in EVERY energy consuming product on Earth. We need to conserve energy use urgently and classD is a good way of doing it.
Now coming to classD being used in high-end audio. It all depends on if it can sound as good as the best available standard equipment.
This subject is being debated right now.
I have used some classD amps and I think they do some things amazingly well. From a price point they are winners. BUT come to meaurements and they are poor in some areas.

How much of the measurements matter in real life is being debated hotly by some people.
But then again amplifiers sound different due to many reasons and you will find hundreds of opinions on the subject.

MY opinion would be that they should refine classD ( which is very young right now ) to a stage when it can outclass the best tube and solid state equivalents. So ( design wise ) you have a challenge right there !
Do a search on this forum and you will find plenty of information. Or just go to the classD section and look under the posts it has. Will keep you occupied for a while !
Wish you all the best. Enjoy being on this forum.
;)
 
It seems many of the same reasons for class D in portable audio still apply for high end audio. Heat is certeinly a factor. Big class A solid state and tube amps can run very hot, which act as a space heaters. Energy savings is always attractive from both an environmental and marketing perspective. Price is also a consideration. Class D amplifiers generally run cool and do not require large, expensive heatsinks and enclosures. The cost to manufacture drops as a result. Aesthetics can be a factor as well. Class D amplifiers can generally be smaller and less obtrusive. This fits into the modern trend of stylish, smaller devices (WAF factor) that allows it to have broader appeal.
 
I would think you could do your project a favor by getting your hands on one of the low-cost, low-power tripath 2024 boards out there and listening to it. There's a ton of them on eBay and in other threads on this forum. They really do have a unique sonic quality, and other benefits that others have just noted.

Good product design and marketing come from passion about ideas, and those are best when fueled and tempered by real life. Too many products are thought up by guys sitting in cubicles, trying to use new technology or looking to fill "holes" in the market.

--Buckapound
 
Well, i had other ideas for my final year project including an Analog Synth, and a tube pre-amp, both of which my lecturers weren't really feeling. Yeah of course i chose the Class D with regards to efficiency. Less power, smaller heat sinks, smaller form factor..etc

ashok said:
Some thoughts....
MY opinion would be that they should refine classD ( which is very young right now ) to a stage when it can outclass the best tube and solid state equivalents. So ( design wise ) you have a challenge right there !
;)

Probably why my lecturers let me do this project, there is still a lot of on going development within the field of class D design.

dweekie said:
It seems many of the same reasons for class D in portable audio still apply for high end audio. Heat is certeinly a factor. Big class A solid state and tube amps can run very hot, which act as a space heaters. Energy savings is always attractive from both an environmental and marketing perspective. Price is also a consideration. Class D amplifiers generally run cool and do not require large, expensive heatsinks and enclosures. The cost to manufacture drops as a result. Aesthetics can be a factor as well. Class D amplifiers can generally be smaller and less obtrusive. This fits into the modern trend of stylish, smaller devices (WAF factor) that allows it to have broader appeal.

Exactly what i was looking for!..

dweekie said:
Aesthetics can be a factor as well. Class D amplifiers can generally be smaller and less obtrusive. This fits into the modern trend of stylish, smaller devices (WAF factor) that allows it to have broader appeal.

I have two main personal goals from this project too. My main interests are circuit design and industrial design. Aesthetics i find to be an important part of product design. One thought i had regarding the use of class D is the smaller form factor.


Thanks very much to the both of yous (ashok & dweekie). Both made very valid points.
 
I think that the advantages of class D are severalfold.

Form Factor - Small is beautiful
Flexability - Power suplies need not be bulky or expensive
Low cost -- You can either use this to make a cheap project with cheap parts throughout, or a mid range product with high end components inside.

Option 1: Since your project is aiming to be commercial in nature, I would either take your Class D amp in one of two directions.

Esoteric Audio, perhaps in a creative designer case, the small form factor would allow for creative solutions here.

http://www.audioinnovations.co.uk/main.htm

It is possible to produce a very nice sounding amplifier with money left over for the finish. Trouble is the mid range audio market is pretty much saturated, so to produce a competitive product, you must either have a technically excellent product, or a suitably desirable design.

Option 2: Cheap and innovative. The Ipod aftermarket is huge, Class D can be used to make portable, the streamed media thing is getting bigger all the time as well. Trouble is there are a multitude of products competing here and accessories are generally cheap and chearfull.

A product that is innovative, yet not too costly can become the next best thing in this market in a short space of time.

The third option is a blend of the two, high end amps for high end streaming sources, or high status products for Ipod owners could perhaps survive by selling exclusively and at a premium.

I am sure others may have other ideas

Oh check out the maxim 9744 digital amplifier IC. It may give you ideas for the second option!
 
Buckapound said:
Good product design and marketing come from passion about ideas, and those are best when fueled and tempered by real life. Too many products are thought up by guys sitting in cubicles, trying to use new technology or looking to fill "holes" in the market.

--Buckapound

I couldn't agree more, Il check out the tripath 2024 boards. Thanks
 
justblair said:
I think that the advantages of class D are severalfold.

Form Factor - Small is beautiful
Flexability - Power suplies need not be bulky or expensive
Low cost -- You can either use this to make a cheap project with cheap parts throughout, or a mid range product with high end components inside.

the streamed media thing is getting bigger all the time as well.

A product that is innovative, yet not too costly can become the next best thing in this market in a short space of time.


Sonos is a leader in the streaming market, with Olive http://www.olive.us/home.html aiming for the higher end. Logitech just introduced the SqueezeBox Boom, adding an amp and speakers to their streamer.

A compromise product that has the streamer and Class D amp, but allows external speakers for better separation and sound, at a price closer to the SqueezeBox than the Sonos, might work.

Here's my attempt a such a compromise: an Omnifi DMS-1 streamer with an added Sure Electronics amp:

http://www.chicksolutions.com/dms1/100_1040.JPG
http://www.chicksolutions.com/dms1/100_1041.JPG
 

Attachments

  • 100_1040.jpg
    100_1040.jpg
    85.2 KB · Views: 1,264
It's very difficult to make a technical business case for entering the manufacture of hi-fi amplifiers, other than perhaps class-D and other amplifiers benefiting from improvements in materials technology and the understanding of digital systems.

These development of these amplifiers does at least have the justification that they are increasingly efficient.

Hi-fi equipment is however a boutique market, bearing more resemblence to clothing fashion than perhaps any other electronic product.

Amplifiers of all classes entirely adequate to the task of hi-fi reproduction abound.

Given that wasting energy is going out of fashion fast, the case for your business looks better than it otherwise might, but very large companies are spending lots of money trying to make your efforts redundant.

It could be of course, that you feel that your talents lie more in the direction of sales than engineering; in either case, good luck.

w
 
chicks said:


Sonos is a leader in the streaming market, with Olive http://www.olive.us/home.html aiming for the higher end. Logitech just introduced the SqueezeBox Boom, adding an amp and speakers to their streamer.

A compromise product that has the streamer and Class D amp, but allows external speakers for better separation and sound, at a price closer to the SqueezeBox than the Sonos, might work.

Here's my attempt a such a compromise: an Omnifi DMS-1 streamer with an added Sure Electronics amp:

http://www.chicksolutions.com/dms1/100_1040.JPG
http://www.chicksolutions.com/dms1/100_1041.JPG


Chick, could you contact me through either pm here or through my contact form on my site?

Regards


Blair
 
chicks said:

A compromise product that has the streamer and Class D amp, but allows external speakers for better separation and sound, at a price closer to the SqueezeBox than the Sonos, might work.

Thats a great idea, however i don't have much time to complete this project (college year including holidays is only about 26 weeks (approx), also i have other projects & assignments to complete during the year as well as exams at both christmas & summer, as expected i have deadlines to meet!

wakibaki said:

It could be of course, that you feel that your talents lie more in the direction of sales than engineering


No thanks, id take engineering any day ! :)
 
Re: Some thoughts....

ashok said:
I have used some classD amps and I think they do some things amazingly well. From a price point they are winners. BUT come to meaurements and they are poor in some areas.

...

MY opinion would be that they should refine classD ( which is very young right now ) to a stage when it can outclass the best tube and solid state equivalents. So ( design wise ) you have a challenge right there !

I wonder about a class-D amp using "feedforward". Quad did that in some of their amps http://www.quadesl.org/Album/InterviewsReviews/MikeAlbinson/mikealbinson.html and I recall one of the oriental companies advertising it as "their" latest great advance, mid-80s.

One way to make an amp more linear is to apply negative feedback. The basic notion of feedforward is that instead you take a sample of the main amp's output, compare it to the input (delayed to take account of the amp's delay), compute a difference signal, invert that, amplify it a bit and mix it back into output.

In theory, this is great. Say your main amp is 200 watts and 5% distortion. The error signal is just 10 watts, and even if your small amp gives 1% distortion, that's only .1 watt, or .005% of the 200-watt signal.

Of course, in practice it is far more complicated. That's why we need engineers!

Either or both amps in such a system could be class D. The overall design leaves the constraints on the big amp fairly loose. For the small one, a Tripath or other class D chip, or for that matter an op-amp might do.
 
I am currently working on identifying customer needs, which is in the form of a Questionnaire. I am trying to come up with questions that will define Product Specification. The majority of questions are technical, regarding the specifications. Some of the questions include:

- How many audio channels would you require in an audio amplifier?
(2, 4 or 4+)
- What power (watts) would you require in each channel of the amplifier (assuming 8Ώ load)?
(50,100 or 200)

...etc

I have other questions regrading specifications like acceptable THD and desirable Frequency Response, S/N ratio, type of speaker connections...etc. I am sure there will also be subjective questions regarding the actual design & form of the amplifier. Although I was wondering if someone was to build you a Class D stereo power amplifier, what other specifications/details would you really look out for?
 
rob mullen said:

- How many audio channels would you require in an audio amplifier?
(2, 4 or 4+)
- What power (watts) would you require in each channel of the amplifier (assuming 8� load)?
(50,100 or 200)

...etc

I have other questions ...

Can you define the application? The target market?

An "audio amplifier" is dreadfully vague. Even within a single system, say a home theater setup, requirements are different for main speakers, surround & sub. There are also "audio amplifiers" in cars, theaters, electric guitars, ...

Even for a single application, what's your market? High volume mass-market? Cost-no-object perfection? Best price/performance trade-off? Sounds better than anything else in price range X, or is cheaper than anything else with performance Y?

The power amp I want most would be three channels, to drive the front three of a home theater system. There are lots of choices in 2-channel amps to drive the surrounds. The main design goals for the three-channel would be that it be be moderately priced and extremely clean.

At least 70 watts because I consider 110 Db per speaker a design goal for adequate loudness, and if sensitivity is 92 Db for 1 w 1m, you need 64 W to get that. More power is good, but not entirely necessary.

Three Coldamp class-D modules and a power supply would give me 240 W/channel for about 500 Euros in parts, plus case, etc. There are also quite a few 5-channel amps available with various specs and prices. To be really desirable compared to those, the three-channel would need to be under $500.
 
Sorry Pashley, I should have been more specific. I guess I am looking at Best price/performance trade-off. I was thinking about the Cost-no-object perfection, but given the nature of my subject I think the Best price/performance trade-off is more suitable. My target market is the HIFI/audiophile market, not home theatre, not automotive, not musical instruments...

The Questionnaire is not finished yet,but when it is i will be looking for people from these forums to fill it out.
 
I'm not sure how hard it would be to design in, but extra flexibility for bridging would be a nice feature.

Say you are designing an amp whose basic spec is 100 W into 8 ohms. Typically, these would also be happy driving a 4 ohm load, or bridged for ~ 200 W into 8. Can you extend this? An amp that tolerates 4 ohm speakers when bridged? Or that can bridge four modules for a ~400 W amp?

If those are possible, can you reduce the basic amp spec? Make a 50 W module bridgeable for 100 or 200? This gives a lot of flexibility, hence a broad market. Depending on room size, speaker choice, budget, etc. one might choose 50, 100 or 200 for stereo. Home theater might use them all -- 100 for mains, 50 for surround, 200 for sub.
 
Two other things that Class D amps seem to do particularly well are (1) very high power and (2) driving very low impedances.

Seems like these in combination might be used to implement some otherwise impractical loudspeaker types. Such as a direct drive ribbon type speaker (very low impedance, and not usually very efficient unless the magnet structure is huge).

Though designing an amp AND a speaker makes for much more work, given your limited time budget. Maybe the amp could focus on those attributes, directed only to a hypothetical ribbon loudspeaker. Some planar magnetic speakers similar to ribbons are discussed in the Loudspeaker section of the DIY forums.
--
Class D also has possible advantages in terms of low level detail. Unlike class B or AB, there is nothing special about small signals in the region where positive and negative voltage outputs meet (a.k.a., no crossover distortion mechanism). In that way, it is more like Class A, except Class D also makes no particular distinction between large excursion and small excursion (even in Class A, the linear gain changes with signal level, better Class D types like UCD have about the same distortion at any level until clipping is approached).
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.