The TPA3116 has a full bridge configuration to get 50W into 4 ohms and to drive a 2 ohm load, they can be put in parallel.
If each output channel were half bridges then yes, you would need two amplifiers rated at 25W into 2 ohms but this isn't the case.
Where are you getting these currents? The class D amplifier acts to switch the load so the minimum impedance should look like the load itself. The current ripple should be handled by the filter capacitors... I'm pretty sure they include headroom in the power handling capability of the TPA3116 as they do with all ratings.
If each output channel were half bridges then yes, you would need two amplifiers rated at 25W into 2 ohms but this isn't the case.
Where are you getting these currents? The class D amplifier acts to switch the load so the minimum impedance should look like the load itself. The current ripple should be handled by the filter capacitors... I'm pretty sure they include headroom in the power handling capability of the TPA3116 as they do with all ratings.
Last edited:
I might have read it wrong. But did you actually take a look at the data sheet?? Or you are so confident that there is no need. "AndewT is never wrong". By the way, I did not "read" anything, I just look at the graph that TI posted in the data sheet and took the data out of it. I could not find any note saying that it is one channel driven. Please go take a look before you start to insult people. If you do not like what is reported there, then explain how TI screwed up and misinformed so that "ignorant" people like me can learn something. You will gain a lot more respect (I do not think you care).
For people like me, audio electronic is my hobby and not my day job, I participate in the forum to share my common interest with fellow diyers and pick up new knowledge and information. I suspected that your reason to participate is to pick on people that are less knowledgeable than yourself since you enjoy it so much.
Looking at the TDA7492MV datasheet and taking Ohm's law and voltage drops into account, I'd say you can expect about 40WPC RMS before heavy clipping. With a 24V supply and 4 ohm load. Nothing unusual, that's about right for this type of amp.
you are missing the point I am making.
24V cannot give 50W into 4ohms.
If you want 50W into 4ohms and must stay with a 24V supply, then you need to bridge TWO amplifiers each of which is rated for 25W into 2ohms.
Once you have decided to adopt the bridged arrangement, you must consider the extra current demands made by the bridging topology on the PSU.
If Ti/ST make this easy by providing all the information to bridge, then go ahead, but remember you will be using TWO amplifiers.
From the very first page of DS:
This amp is already bridged.
With some stereo chips it is possible to parallel BTL the outputs.
Maybe with some (over the top) engineering it is possible to parallel two chips.
Look at the the datasheets of TAS5630 or the newer TPA3251.
Those can be used with 4 SE outputs, 2 BTL outputs or one single PBTL output (2.1 with 2 SE and 1 BTL is possible, too).
A completely different story is how much power is delivered with tolerable distortion.
I experienced those TDA749x boards to sound nice, but to distort heavy long before what i interprete (from listening) as actual clipping.
Compared to the TI-chips that effect is more ostensible/audible in my subjective experience.
I think you are missing the point. The following are your words:
Quote:
Originally Posted by Lo_Tse View Post
Well, TI claimed that number.
AndrewT
You are proving that you cannot read and interpret a datasheet.
Quote:
In the data sheet, TI showed that at 24 V, into a load of 4 ohm (resistive I presume), the chip output can get to 75 W at 10% THD + noise. Or 60 W at 1% THD + noise.
AndrewT
As I showed in my earlier post these numbers are impossible.
The focus point here is - did you read the data sheet before accusing me of "cannot read and interpret a data sheet"? I even mentioned in the first message that the chip is in BTL (bridged tied load) configuration already. You should read the data sheet before you jumped the gun. I really felt being insulted.
No, I did not read the datasheet. See my next post.
But you should have read and realised that what I did say and still say is: that the simple statement that you NEED two amplifiers in Bridge topology to achieve those power claims is still true.
Once you adopt bridged topology, you must take account of the special conditions that ensue.
The Manufacturer may not spell out the special operational conditions for fear of losing potential customers.
You as a Builder should be able to read the datasheet and extract the necessary information that allows a good implementation.
Or go in blindly and accept a substandard quality of reproduction.
But you should have read and realised that what I did say and still say is: that the simple statement that you NEED two amplifiers in Bridge topology to achieve those power claims is still true.
Once you adopt bridged topology, you must take account of the special conditions that ensue.
The Manufacturer may not spell out the special operational conditions for fear of losing potential customers.
You as a Builder should be able to read the datasheet and extract the necessary information that allows a good implementation.
Or go in blindly and accept a substandard quality of reproduction.
Last edited:
Last night I looked at two of the datasheets recently mentioned on this Thread.
Both clearly show that the amp is NOT already bridged.
What I do see is that the datasheet shows implementations and schematics to guide the Builder/Designer to bridge the separate amplifiers contained inside the chip.
Go and look for yourself. The datasheet SHOWS the separate inputs and separate outputs of several amplifiers, inside. the one IC chip.
The supplementary information shows how to combine these separate amplifiers into a bridged topology.
Andrew. Please stop. You are quite simply wrong. Check Ohm's law. Read the datasheet. Learn about bridged amps. Stop hassling those who understand. Stop.
The amps are not bridged in case of being paralleled, but bridged in the way that the load is connected into a full bridge rather than a half bridge. That's why 24V in BTL is giving you 17Vrms. (24/sqrt(2)).
- Status
- Not open for further replies.