Performance ranking of CLASS D power amplifier MOSFET.

A few off the top of my head - simplicity, cost, compactness, reliability, efficiency.
Sorry, I can't agree.
Same performance. Power amplifier. A single power supply requires two chip btls. Therefore, its cost is higher and more complex. The cost will also be doubled, and the reliability will be doubled, because it requires two amplifiers. I don't think BTL will be more efficient. Because BTL cannot increase current output capacity.
If the power supply is only half. The output voltage is the same.
We cannot compare different types of amplifiers.
For example, TDA2030A, single power supply TDA2030A BTL. Obviously, TDA2030A BTL is more expensive, larger and more complex. And the performance will not increase.
 
I view a true digital amplifier as a single bit power DAC, probably at too low clock and with all the idle tone noise issues that causes.
A true class-D has continuously variable duty cycle. Unfortunately real power mosfets take significant gate charge to switch and generate a lot of ringing noise on all the pins. This makes the real PWM - voltage curve ragged. NFB helps, but does not cure as at the detail level you may even get negative gain and therefore positive feedback. Actual implementation and layout is far more critical than with a Class AB
Glad to hear from you.
I think your reply is very close to what I imagined. Because my English is poor.
It's not easy for me to explain clearly as you do.
I think different MOSFET QG and QSW have obvious influence on the shape of PWM. High power and low THD will have the opposite performance.
Because high power requires higher supply voltage. It will make MOSFET heating more serious. Therefore, irs2092 mode. The performance under high power is better than that of individual chips, such as TI products.
This is a bit like the difference between lm3886 and lm4702. Lm4702 is easier to achieve higher voltage. Higher power. And thd is smaller.
The MOSFET is separated from the drive circuit. High voltage can be avoided. The influence of high calorific value on the driving chip irs2092.
 
Single rail does not necessitate BTL, a coupling cap will do if a single-ended amp is powerful enough.
If BTL is not used. Then the performance of OTL power amplifier is far lower than that of OCL amplifier. No manufacturer will use OTL amplifiers in the 20th century.
OTL was mainly used before 1975. It's something to be eliminated.
of course. DIYer many people like things that are eliminated.😛
 
You can write to Infineon. Tell them. Your PDF document is wrong.
Digital amplifiers should not be used. Irs2092 should be written as an analog amplifier.
I think they may not reply to you.
Ohh, I often write to manufacturers about errors in datasheets.
Most modern datasheets have a link somewhere where you can report errors.
Not just names, sometimes graphs are clearly wrong.
They always reply and thank me.
Everybody makes mistakes, in this case everybody can see it is wrong.
Note that the guy who is tasked to write a datasheet is often not the guy who designed the part.
He is busy designing more parts ;-).

Example: https://www.ti.com/feedbackform/techdocfeedback?litnum=SBOS901B&partnum=OPA1655

Jan
 

Attachments

  • docfb.PNG
    docfb.PNG
    8.2 KB · Views: 97
Last edited:
Everybody makes mistakes, in this case everybody can see it is wrong.

I suspect in this case its really just marketing calling an amp a 'Digital Audio Amp'. Just like some headphones get marketed as 'Digital Audio HPs'. Of course it can be wrong as well as be marketing but in this case I doubt they'll acknowledge it as a mistake. The first bullet point acknowledges the device is 'analog input'.
 
Last edited:
  • Like
Reactions: jan.didden
I thought 'OTL' meant 'output transformer less' and was only applied to valves?

The principle is the same.
Why do you need to add a transformer to the output. Because the output of a single power supply will have half of the DC voltage.
Capacitance is also OK.
JLH1969 is OTL AMP Unless there is really no way to get a dual voltage amplifier. Otherwise, I don't think we should use single power amplifier. It has too many defects.
 

Attachments

  • 单电源3886.jpg
    单电源3886.jpg
    62.1 KB · Views: 139
I thought 'OTL' meant 'output transformer less' and was only applied to valves?

I view a true digital amplifier as a single bit power DAC, probably at too low clock and with all the idle tone noise issues that causes.
A true class-D has continuously variable duty cycle. Unfortunately real power mosfets take significant gate charge to switch and generate a lot of ringing noise on all the pins. This makes the real PWM - voltage curve ragged. NFB helps, but does not cure as at the detail level you may even get negative gain and therefore positive feedback. Actual implementation and layout is far more critical than with a Class AB

We can agree to differ. I find it works fine.
yes. A car with three wheels is actually very easy to use.
But if the price is the same. Why not buy a car with four wheels.
In fact, if the same structure is adopted. Dual voltage amplifiers will be cheaper. Better performance.
Let me make an analogy.
Lm3886 should be a chip that everyone knows.
If you use the first scheme, use an lm3886 chip. Use + - 30V voltage.
The second scheme uses two lm3886 chips. Use + 30V GND voltage BTL.
Which effect do you think is better. Which is cheaper. Obviously.
The first option is cheaper. And the effect is better.
 
The important thing is how it sounds. If it sounds fine, who cares whether it is OTL or OCL or whatever?
You confuse performance, cost, complexity.

Jan
Let me make an analogy.
Lm3886 should be a chip that everyone knows.
If you use the first scheme, use an lm3886 chip. Use +30V GND - 30V voltage.
The second scheme uses two lm3886 chips. Use + 30V GND voltage BTL.
Which effect do you think is better. Which is cheaper. Obviously.
The first option is cheaper. And the effect is better.
And their functions are the same.
 
Hm, what those defects would be?
Seems that designers like Nelson Pass are not aware of those defects as some of their designs use single rail PS, latest being First Watt F8. 🙂

First Watt F8
There are some OTL amplifiers in the world.
For example, jlh1969. Or some single power amplifier. For example, quad303.
It's just that OTL circuits are not used from quad405-606-707-909.
OTL circuits were mainly used before 1980. Few manufacturers used it after 1980.
Specific advantages. Disadvantages. I think it can be in Baidu. Or you can find it on Google.
I don't really want to explain these things. Because it's just like someone said that there is any difference between a three wheeled car and a four wheeled car.
Maybe Lamborghini's tricycle is better than four-wheel R


I think many people here don't understand real analog amplifiers.
In fact, hifi, hiend products. Almost no OTL analog amplifier is still in production.
So many people don't understand why Amplifiers need dual power supplies.
But in hiend manufacturers, they will not consider this problem. Because this is like saying that Porsche will design a three wheel car in the future.
I don't think it's necessary to consider this issue at all.
Digital amplifiers do have many of these chips. They were born. It's going to be on mobile devices. Car 。
It is not considered to be used in home hifi sound.
 
Last edited:
Let me make an analogy.
Lm3886 should be a chip that everyone knows.
If you use the first scheme, use an lm3886 chip. Use +30V GND - 30V voltage.
The second scheme uses two lm3886 chips. Use + 30V GND voltage BTL.
Which effect do you think is better. Which is cheaper. Obviously.
The first option is cheaper. And the effect is better.
And their functions are the same.
What do you mean 'effect'? Lower THD? Depends on the rest of the circuit.
Cheaper? Don't know, two chips but simpler power supply; an extra high value electrolytic is probably more expensive than an extra 3668.
Total price depends on, well, the total circuit.
Better sound? Maybe for you, maybe not for someone else, very subjective.
So I really don't know what your point is.

Jan
 
I think many people here don't understand real analog amplifiers.
Ahhh yes, that must be it! We bow for your superior knowledge!
🙏

BTW Did you check how many of your hi-end amplifiers use class D modules? Measuring and sounding better than class A?
Check the tests and rankings at ASR.
You really have to educate yourself, you seem to live in the last century.

Jan
 
What do you mean 'effect'? Lower THD? Depends on the rest of the circuit.
Cheaper? Don't know, two chips but simpler power supply; an extra high value electrolytic is probably more expensive than an extra 3668.
Total price depends on, well, the total circuit.
Better sound? Maybe for you, maybe not for someone else, very subjective.
So I really don't know what your point is.

Jan

You can say single power amplifier. It is easier to buy industrial power supply. Such as charger. This is really true.
But that doesn't mean it's better. Although they can also be used. And the effect is good.
But you can't say three wheel car. It will be better than four wheels. At least I won't tell others like this.
Because I think it's good or bad. It has nothing to do with your personal preferences.

In my cognitive range. Iraudamp4a and iraudamp5a are the best amplifiers.
Then iraudamp7s.
I used these in 2010. In addition. I used Jeff Rowland 250
ARC DSI200
They all cost more than $3000. And I have fm115 here. Class B amplifier.
The price is $150000
 
Sure. You use Jeff Rowland 250?
Class D and SMPS.
Have fun, I have better things todo.

Jan
I don't use Jeff Rowland 250
I just compared this machine with iraudamp7s.
The result is that iraudamp7s is much better.
JEFF ROWLAND 250 use ICEPOWER 250
If only for convenience. Easy. Can use notebook charger power supply to make power amplifier.
Then I recommend tda2005. Analog amplifier. It has better effect than tpa31163225.
The price is cheaper. Although the calorific value will be a little higher.