CLASS AB Amplifier Vs CLASS D Amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,

Iam planning to build a two Way plate amplifer. For the LF Section iam planning to use a Class D Amplifier Module the power is 400W @ 4 Ohms.

For the HF sction which amplifier type would be best. The Wattage of the HF is 100W.

Some people say that CLASS AB Module is better for HF section and some people say CLASS D Module will be good.

So please advice me on this whether to use Class AB or Class D module for the HF Section.

Help me on this:confused::confused::confused:.

Thankx
 
Modern class D amps rival and surpass most class AB amps in sound quality at a much lower price and vastly lower power consumption.

Class AB amps are in my opinion a relic of the past, gone the same way as tube amps, vinyl records, and AM radio. Sure, some people will still swear to them but they're more an oddity having no value in themselves beyond nostalgism.
 
Modern class D amps rival and surpass most class AB amps in sound quality at a much lower price and vastly lower power consumption.

Class AB amps are in my opinion a relic of the past, gone the same way as tube amps, vinyl records, and AM radio. Sure, some people will still swear to them but they're more an oddity having no value in themselves beyond nostalgism.

I disagree, most people get headaches from listening to music through a class D amp!:cool:

Seriously, I still disagree to the categoric statement above and BTW it´s noting wrong to use a tube amp or vinyl records if you like it this way.
 
Ex-Moderator R.I.P.
Joined 2005
I would suggest to build two stereo amps, one each channel

or you could 3 smaller amps of equal design
bridge two of them for LF
and the other for HF
this way you get a high power and a low power amp running on the same voltage

though, a bridged amp into 4ohm may be a tough one
 

opc

Member
Joined 2004
Paid Member
Hi nkdecibels,

From a performance and sound quality standpoint, there's no question that a class AB, or better yet a class A amplifier would be your best choice. The only reason to use class D is high power density and better efficiency.

Class D pros:
1. Higher power density (both in size and weight)
2. Higher efficiency
3. Potentially lower cost

Class D cons:
1. Limited power Bandwidth
2. Lower SNR
3. Higher THD
4. Response that varies with load
5. Higher circuit complexity
6. EMI/EMC related issues

So in short, if you're building a powered monitor for portable PA use, then you'd be a fool to use anything but class D. For any application where sound quality is a primary concern, class A or class AB are the way to go.

I know there will be some who disagree, but I would like for them to show just one sound quality related area where a class D is better. I have yet to find one, and used to design class D amps for a living!

As a side note, are you sure you need 100W for your tweeter? That's a lot of power to be feeding into a tiny driver. Dropping that requirement would allow you to use something like an F5 which would really be the cat's meow.

Cheers,
Owen
 
opc, may I humbly suggest that there have been developments in Class D that elimiate any of the disadvantages you mention:

Class D cons:
1. Limited power Bandwidth
2. Lower SNR
3. Higher THD
4. Response that varies with load
5. Higher circuit complexity
6. EMI/EMC related issues

1 - Not that I know of. Perhaps I don't understand. Can you elaborate?
2 and 3 - Here are some measurements of a good, clean Tripath TA3020 amplifier, the 41Hz Amp15 which gives 0.008% THD+N at up to 74Wrms, and a noise floor below -120dBV. (And the IMD performance of good Class D is excellent too!) It is possible to reduce the distortion figure further, and increase power, with modifications / choice of chipset. I suspect that the Hypex NCore will then be significantly quieter and cleaner.
4 - Not likely to matter significantly in practice. Furthermore, Hypex UcD and NCore, and other designs, do away with this limitation entirely.
5 - Not likely to matter in practice. Secondly, Class D amps are much more efficient, and a populated board for a high-power Class D amp with very high sound quality is possibly going to be much cleaner and smaller than a similar Class A; Also more reliable due to less heat, and more quiet due to less stray inductance etc. in the signal path.
6 - If the design is done properly, then Class D amps don't need much shielding. Add minimal shielding and the environment is as good as EMI-free.
 
Last edited:
I test class D all day, measures fine w sines.
Horrible to listen to with real music. The way
it clips into real speakers with music of any
level, awful... Sorry engineers, compression
over the long time constant doesn't "fix" the
problem.

Can't wait to get home and listen to my tubes.
Yet, I don't think Class-D is beyond all salvage.
Try to tell a chip engineer to preemptively soft
clip before hard voltage and current limits is a
tough sell. That sort of processing is someone
else's job. I've had no luck presenting my ideas.

And that it behaves as a negative resistance
to the power supply, got those cutting in and
out too. No idea how to fix that aside from
setting stOOpidly high safety limits at the PS.
Maybe its just my Agilents? Overpriced crap...
 
Last edited:

opc

Member
Joined 2004
Paid Member
Hi kristleifur,

It might be more accurate to say that there have been some improvements that have helped to mitigate some of the disadvantages of class D amplifiers. Nothing groundbreaking has happened in the past 5 years in terms of class D amplifier design... marketing hype aside.

1. The LC filter at the output inherently limits your power bandwidth. There's no way around it. Post filter feedback can help a little at very low power levels, but not at higher power.

2-3. I would call those measurements acceptable, but not excellent. They've also done a fair amount of fudging to get there. They call it a 2x150W amp, but they've intentionally measured it under conditions that make it a 75W amp. Using lower voltage rails helps to lower the noise and also lowers distortion. They do state this, which is good, but these are not measurements from a high power class D amp. Also, that 0.008% is only at 10W and only below 1kHz. Above that, things degrade quickly.

Also note that a 2nd harmonic at -80dB is nothing to write home about. I've seen loudspeakers with distortion levels that low under the right conditions. There are also a lot of higher order harmonics in there. The 16th harmonic is clearly visible and that's with limited measurement BW.

If you want to see good measurements, check out the datasheet for the LME49830. I mocked up a little class AB amp last year with that IC and a pair of lateral FETS with 0.5A of bias current and 50V rails. It measured a factor of 10 better than what you linked to in every respect.

4. Of course it matters! Why would you invest in an amp that peaks by 2dB at 15kHz when a class AB doesn't? There are also phase issues involved that you don't get with a class AB. Nobody has actually solved this issue yet, and I have never seen a response graph with 2, 4, 6, 8, and 16 ohm responses that are all perfectly flat to 20kHz. Even a poor class AB can manage that.

5. You are correct that for some of the modular all in one IC's there is significant space saving (and none of the complexity) but for two comparable discrete designs, the class D amplifier will be significantly more complex to build and work with.

6. "If the design is done properly" - That's a big "if"! All switch mode amplifiers and supplies suffer from EMI/EMC issues, and it can be extremely difficult to deal with them! The integrated IC solutions obviously pose less of a risk if they're implemented as specified in the datasheet, but even then it can be an issue.

Overall though, what's the point in dealing with all of the above if you have a perfectly viable solution in a class AB amplifier that has none of those problems? The only two legitimate reasons are power density and efficiency, and depending on the application those can be very good reasons.

As an audiophile though, there's no good reason to use a lesser performing amplifier just because it's a little more efficient and affords you a little more power per litre/kilogram. The whole point is to get the best sound regardless of those other criteria. It you want the best sound, then Class D is not where it's at.

Cheers,
Owen
 

opc

Member
Joined 2004
Paid Member
Hi nkdecibels,

For mobile DJ/PA use, I'd say definitely go for class D throughout. In those conditions it's unlikely you'll notice a sound quality difference, and the weight savings alone make it well worth it. You actually do need high power for PA, and the best way to get there is class D in my opinion.

When it comes time to build something for your living room though, do yourself a favor and steer clear of the class D amps.

Cheers,
Owen
 
Hi,
After read this thread, i say...what type di D class you have listen?
Obvious that 80-100w A-B class is real good for HF driver but..some D class (serious develop) have not difference. i mean:linearity,thd,FR and response Vs. load. also 1,5dB from 4R to open load is very good.

Regards
 
A categoric statement usually don't start with: "In my opinion..." ;)

Indeed it can! The core of a categoric statment is the lack of the perception of uncertainty that we ought to account for if we want to be trusted.

A categoric statement could not be saved by "in my opinion" because if you maintain that your statement is by definition true it becomes obviously rather worthless to us as it proves noting. From Wikipedia;
" Tautology (logic), a technical notion in formal logic, universal unconditioned truth, always valid." Thus it´s hard to contradict such a statement.

A generalization like "all class D amps are better than all class AB amps has an intrinsic weakness; it demands to be more precise, in what respects? It is also hard to prove such a statement as neither you nor I have had the opportunity to compare all amps in the world.
 
Last edited:
None of the technicalities that opc mentions actually matter :) The sound quality does matter, and good Class D actually sounds great. There are numerous reviews around; I take issue with the portrayal of Class D as implausibly inferior sounding.

It's all good. I'm sure there are great Class-AB amps out there, I just haven't heard them yet.
 
When designed and build right, a Class D makes for a great power supply for a decent Class AB tho. Otherwise Class D has only very limited use in my books, may be ok to use with subs or servos and of course all battery driven devices. Other than that, Class D (and all the other ROHO's stuff for that matter) - No Thanks.

Until Class D signal reproducing issues are not resolved, I would not consider it fit for use. Not sue what some mfg's such as ADAM audio were aiming for when releasing there studio monitors equipped with this unfit Class D technology, it spoils the otherwise superb speaker.

But then it all depends, if quality is not so important and efficiency is requirement, then Class D has it's place, I guess.
 
To expand on what I said earlier. It's true that some class D amps aren't properly designed and therefore aren't terribly good but from person experience then a well built class D far exceeds the sound quality achievable by class AB or class A amps regardless of price.

So for a modern amp you so only consider class D if you care about sound quality at all, class AB is basically junk only suited for the scrapheap.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.