Put the modsel pin (pin 1) to logic high instead of connected to ground, ie. just connect it to the shutdown and fault pins beside it but make sure the ground connection is broken.
Last edited:
Lower switching losses. Slightly higher THD. Better sound quality at low listening levels, same at higher levels. But you draw your own conclusions. And please be aware that it requires very tight tolerances of the output filter to even consider it and it's better suited for ferrite bead output filters.
Last edited:
Any minute advantages, either real or perceived can only be discussed if you all have the same sound source, speakers, room acoustics and ears?
Speakers do not feature at all in any of the discussions yet they are what colours a system the most, not amplifiers.
Any minute advantages, either real or perceived can only be discussed if you all have the same sound source, speakers, room acoustics and ears?
Speakers do not feature at all in any of the discussions yet they are what colours a system the most, not amplifiers.
suppose we should stop discussing amplifiers then
Any minute advantages, either real or perceived can only be discussed if you all have the same sound source, speakers, room acoustics and ears?
Speakers do not feature at all in any of the discussions yet they are what colours a system the most, not amplifiers.
I sort of assume that when you compare differences you only change one element, ie. the amplifier or a specific amplifier setting, and that everyone uses the same speakers, in the same position they are accustomed to otherwise listening tests are meaningless.
It is also meaningless. And I do mean completely and utterly irrelevant what speakers you listen to. Period. As long as you listen to different settings using your normal set-up that you are used to and listen to the same music you are used, everything else is irrelevant.
Note, that no one here claims any definitive knowledge on listening experience. That some or most of us don't write "in my experience" or "to me personally" or "on my system" in every post is because it's implied that it is all each our own subjective opinion that we post.
As any rational person is aware, even if my listening experience in one particular detail is shared by other people, doesn't make it an absolute fact. Every one here knows this, or should know this, and applies his/her own dose of salt to the mix.
Last edited:
I usually find something worthwhile in hearing others impressions of what makes a difference in their system even if its quite different from my own. And I especially pay attention when someone talks about micro details, downward dynamic range, and other aspects that indicate that the equipment is getting out of the way and allowing the subtle details of a recording to come through.
Hi all!
I am secretly following this topic for quite a while.
I am currently waiting for 4 versions of the TPA3116 amps:
- Yuan-Jing 2.1 (with tone control and a volume pot)
- Yuan-Jing 2.0 (no tone control, just a volume pot)
- Feixiang M-DIY-2.1 Rev 3 (on/off switch, tone control, bass frequency control, RCA inputs)
- I believe also Feixiang 2.0
The last one seems more promising. (10uH inductors, good input caps, good output caps, no potmeter and it's a rather small pcb)
The above is not actually what I wanted to share with you, it's that I believe that it would be a great idea to create some sort of an open-source TPA3116 PCB.
I believe if we all work together on this, we could create a decent design. Maybe even a universal design. (for example: use SMD and through-hole components and set BTL/PBTL aka stereo/mono)
I did make a start(made the TPA3116 part of my eagle library).
That's as far as I went. (mostly because of the limited time I have)
If we succeed, I think we should all 'sample' the TPA3116 from TI, get the components we want for our amp and produce the board at Seeedstudio.
Anyways, I would like to hear your opinion on this before I start a thread! (and put too much effort into this...)
Greetings,
Loek
I am secretly following this topic for quite a while.
I am currently waiting for 4 versions of the TPA3116 amps:
- Yuan-Jing 2.1 (with tone control and a volume pot)
- Yuan-Jing 2.0 (no tone control, just a volume pot)
- Feixiang M-DIY-2.1 Rev 3 (on/off switch, tone control, bass frequency control, RCA inputs)
- I believe also Feixiang 2.0
The last one seems more promising. (10uH inductors, good input caps, good output caps, no potmeter and it's a rather small pcb)
The above is not actually what I wanted to share with you, it's that I believe that it would be a great idea to create some sort of an open-source TPA3116 PCB.
I believe if we all work together on this, we could create a decent design. Maybe even a universal design. (for example: use SMD and through-hole components and set BTL/PBTL aka stereo/mono)
I did make a start(made the TPA3116 part of my eagle library).
That's as far as I went. (mostly because of the limited time I have)
If we succeed, I think we should all 'sample' the TPA3116 from TI, get the components we want for our amp and produce the board at Seeedstudio.
Anyways, I would like to hear your opinion on this before I start a thread! (and put too much effort into this...)
Greetings,
Loek
The answer is not that simple. As mentioned, based on datasheet, it should not. But as I found DC input offset is unexpectedly lowering while rising power voltage there could be any side effects with Plimit too.
It's a great idea!
IIRC, someone here on DIYaudio has already designed a board for the TPA3116D2 chip (I don't remember the thread name, but should be the one started by XRK about the "dead bug" wiring project)
Maybe we should get in touch with him and take it from there.
Judging by the listening impressions of everyone here and the shortcomings of the YJ boards, it would be only logical to make properly designed board for this chip with the right inductors sizes, better caps, optional passive crossover for bi-amping, etc....
Last edited:
If you just look at the opensource Arduino platform, you can see that the chinese are producing them by the thousands!
Why not start an open-source audio platform, right?
It could start just right here!
The good thing about open-source is that if it get's popular, prices will be very low. (a lot of competitors trying to be the lowest, look at Android for instance)
Just like the arduino platform, we could make 'shields' or 'modules'(for tone control, volume control, pre-amp etc.) Just think of the possibilities! Everyone could attach their own favorite pre-amp/tone-volume control. No hassle.
The final thing would be that we go to a manufacturer to produce them. (maybe Yuan-Jing or that Feixiang company)
I need more opinions on this of course!
Why not start an open-source audio platform, right?
It could start just right here!
The good thing about open-source is that if it get's popular, prices will be very low. (a lot of competitors trying to be the lowest, look at Android for instance)
Just like the arduino platform, we could make 'shields' or 'modules'(for tone control, volume control, pre-amp etc.) Just think of the possibilities! Everyone could attach their own favorite pre-amp/tone-volume control. No hassle.
The final thing would be that we go to a manufacturer to produce them. (maybe Yuan-Jing or that Feixiang company)
I need more opinions on this of course!
I fear that there will be too many varying opinion on the design choices.
Starting from the output. Ferrite bead or LC outputs filter? The first being the better and cheaper option but then speaker cable has to be less than 1 meter, so that's only suitable for active speakers.
Chip settings. There will be different opinions on whether on not it is better to have one or the other setting in relation to clock frequency, modulation mode, mono/stereo mode and gain. And there will certainly be arguments against a solution that is customizable as that severely limits optimum board layout considerations which are vital in class D.
Input section. Buffers to feed the differential inputs directly or standard? Active/passive filters?
Power supply. On-board or off-board? Variable?
Cooling. Use the TPA3118 and use PCB bottom ground cooling (basically have the whole bottom side made into a ground plane so that it can be attached to any metal surface with double adhesive thermal tape for cooling) or use TPA3116 and have a heat sink as normal, or attach the chip to the bottom so that it can be fixed to an aluminium cabinet for cooling, a heat sink or just a sheet of aluminium.
So many choices and I have only named a few.
Starting from the output. Ferrite bead or LC outputs filter? The first being the better and cheaper option but then speaker cable has to be less than 1 meter, so that's only suitable for active speakers.
Chip settings. There will be different opinions on whether on not it is better to have one or the other setting in relation to clock frequency, modulation mode, mono/stereo mode and gain. And there will certainly be arguments against a solution that is customizable as that severely limits optimum board layout considerations which are vital in class D.
Input section. Buffers to feed the differential inputs directly or standard? Active/passive filters?
Power supply. On-board or off-board? Variable?
Cooling. Use the TPA3118 and use PCB bottom ground cooling (basically have the whole bottom side made into a ground plane so that it can be attached to any metal surface with double adhesive thermal tape for cooling) or use TPA3116 and have a heat sink as normal, or attach the chip to the bottom so that it can be fixed to an aluminium cabinet for cooling, a heat sink or just a sheet of aluminium.
So many choices and I have only named a few.
Last edited:
In looking at that list, it strikes me that in the end most people are probably more interested in a better version of the standard boards as designed/built by TI and Yuanjing, than a radical redesign.
So far, I have not seen too many people writing about active speakers or a need to change clock frequency or modulation mode, onboard power supplies or active filters. Changing gain does not require anything more than parts choices (as I read the TI datasheet).
The TPA3118 has not attracted nearly as much attention as the TPA3116, so it seems that the dinky little heatsinks used so far seem to be working for people. Mine never even get warm.
In other words, the idea has great possibility for those happy with the basic board, but wanting better (maybe even legitimate) parts than Yuanjing puts on their boards.
So far, I have not seen too many people writing about active speakers or a need to change clock frequency or modulation mode, onboard power supplies or active filters. Changing gain does not require anything more than parts choices (as I read the TI datasheet).
The TPA3118 has not attracted nearly as much attention as the TPA3116, so it seems that the dinky little heatsinks used so far seem to be working for people. Mine never even get warm.
In other words, the idea has great possibility for those happy with the basic board, but wanting better (maybe even legitimate) parts than Yuanjing puts on their boards.
Isn't that simple? Just implement both option. (ability to solder both, you'll be able to decide for yourself or the manufacturer could produce both versions)
I believe all these options are settings. Just like the jumpers used on the EVM from TI we would be able to solder different resistors/different connection.
I also believe this is a setting. (connect grounds together or not). Inputcaps for the L and R inputs could be replaced by a 0 ohm resistor or simple wire.
I would say off-board for sure. A good (linear) power supply is just way to expensive. Think about the variety in voltages.(some want to use 5V, other 9V,12V or 24V etc.) This should be the component that people could select for themselves based on their own requirements of sound quality.
But... and "on-board power supply" could be sold in the form of a module or shield.
The TPA3118 is pin-to-pin compatible with the TPA3116 I believe. And also uses the exact same components and corresponding values etc. Only minor changes have to be made to use the TPA3118 or TPA3130. (create cooling pad under chip)
Looking at the EVM, TI placed a cooling pad under the TPA3116. Since the TPA3116 has the pad on the top, the board is probably made in such a way to be compatible with both the TPA3118 and TPA3130 to enable cooling.
It is very good of you to look over this with a skeptical way. But you must also see the possibilities!
Ask yourself, what would you rather like: a yuan-jing board or "our future board"?
I am aware of the effort and work that needs to be done prior to designing the boards.
The options listed by Saturnus are just the beginning of what could become an impossible headache to manage.
So it might be a good idea to set some fixed general guidelines, and if I were to choose, I'd go for something that's fit for a wider audience with some potential for optimization:
- LC filter with proper inductors matched to speaker impedance (more adaptable to a wider audience)
- Off board PS (the voltages are quite generic, and current draw fairly easy to get with an SMPS or battery)
- Chip settings: minimal distortion for a certain output level (I have no idea what the different options are and am counting on Saturnus's input here )
- Input gain: selectable through dip switches (it's just a couple of resistors) or different component value options
- PRICE! this shouldn't cost an arm and a leg to build, a cost of around 50$-60$/board would be a good target to meet if good quality components are used.
Had I the proper skills, this is how I would build one of these amps
The options listed by Saturnus are just the beginning of what could become an impossible headache to manage.
So it might be a good idea to set some fixed general guidelines, and if I were to choose, I'd go for something that's fit for a wider audience with some potential for optimization:
- LC filter with proper inductors matched to speaker impedance (more adaptable to a wider audience)
- Off board PS (the voltages are quite generic, and current draw fairly easy to get with an SMPS or battery)
- Chip settings: minimal distortion for a certain output level (I have no idea what the different options are and am counting on Saturnus's input here
)- Input gain: selectable through dip switches (it's just a couple of resistors) or different component value options
- PRICE! this shouldn't cost an arm and a leg to build, a cost of around 50$-60$/board would be a good target to meet if good quality components are used.
Had I the proper skills, this is how I would build one of these amps
There are some aspects of the YJ boards that we all agree should be optimized (LC filtering, Input topology and gain, etc...)
You have added to that the potential chip settings which most of us are oblivious to.
Wouldn't it make sense in your opinion to address these while keeping a minimal adjustability built into the setup?
If you had the skills, do actually think that you could make one for $50-60 bucks? I can't!!
It would have to be a super deal, just to get the blank pcb for $50
Sure I'd release my portable audio platform to the world, maybe I'd get a hand writing some software, since that is a bigger challenge to me than the hardware, as that is done already. BTW it uses a TPA3110, as TPA3116 was not released at that time.
A few years back a designer friend of mine suggested an Arduino platform, at the time I was designing. I am glad I designed my own hardware, using a Amtel xm128a3.
Makes no sense to make the audio components modular, as the costs are too much. I put a complete radio/media player on a 30 sq in pcb. It would not take too much work, or increase the pcb size to make it 2.1 channels.
Anyone interested to fund/support/assist in my next revision? I can also include the new tuner design using a Si47770. Send me a PM.