Very interesting thread. But Im not sure what the initial idea/architecture is. I understand you want to build an amp based on the TI PWM modulators. I think it would be a lot more exiting if you would include some more TI stuff. I checked the TI site and this gives you some really cool features (I'm no expert, I only checked the features and I'm no TI employee):
Something like this:
- Modular setup where every speaker has its own PWM modulator + power stage (you will need a host controller to update all the modulators). See picture below. The advantage is that you are not limited to any setup. You can buid it into a amplifier enclosure or into the speaker or sub. Or you can build multi room setups. Imaging that you build everything into the speaker. That would be so cool!
- Programmable DSP (they call it DAP): TI has software available for gain, EQ and crossover programming
- PSVC volume control: a combination of digital gain and power supply to control volume. This minimizes loss of resolution and keeps SNR high.
- Wireless speakers using a radio frequency link (uncompressed/lossless/error correction) you can have 1 speaker with a wireless master streamer and other speakers/subs with wireless slaves.
- Host controller: the CC85xx can be the I2C master and control the other devices (write default register setup to the modulators on system startup).
- Closed loop amps: TAS56xx series
Something like this:
Thanks for posting! I'm not necessarily geared towards TI but they seem to have interesting components.
I had as an architectural goal to minimize the length of the analogue path.
Can 10cm be reached? That was the only goal but a lot of things follow in order to achieve this. Clocking/synchronicity/jitter is one, level control is an other. These properties put high demand on the solution.
//
I had as an architectural goal to minimize the length of the analogue path.
Can 10cm be reached? That was the only goal but a lot of things follow in order to achieve this. Clocking/synchronicity/jitter is one, level control is an other. These properties put high demand on the solution.
//
Hello,
Just to update my thinkings on the topic, I finally arrived there http://www.diyaudio.com/forums/digi...ow-jitter-achieved-stm32-microcontroller.html
Those devices have some potential to solve some of our problems, like the ability to drive up to 4 stereo channels (8 speakers). It is not TI, but a competitor.
I'm interested in that option because of the nucleo boards that allow to go forward at low price without designing dedicated boards. This is one of my problems: wonderful chips are on the shelf, but designing a good dedicated board and assemble it is out of my capabilities. SMD are so small !
Nota: if you have identified the way to implement the wireless streamer slave in a arm MCU, I'm interested.
JMF
Just to update my thinkings on the topic, I finally arrived there http://www.diyaudio.com/forums/digi...ow-jitter-achieved-stm32-microcontroller.html
Those devices have some potential to solve some of our problems, like the ability to drive up to 4 stereo channels (8 speakers). It is not TI, but a competitor.
I'm interested in that option because of the nucleo boards that allow to go forward at low price without designing dedicated boards. This is one of my problems: wonderful chips are on the shelf, but designing a good dedicated board and assemble it is out of my capabilities. SMD are so small !
Nota: if you have identified the way to implement the wireless streamer slave in a arm MCU, I'm interested.
JMF
Last edited:
Very interesting thread. But Im not sure what the initial idea/architecture is. I understand you want to build an amp based on the TI PWM modulators. I think it would be a lot more exiting if you would include some more TI stuff. I checked the TI site and this gives you some really cool features (I'm no expert, I only checked the features and I'm no TI employee):
- Modular setup where every speaker has its own PWM modulator + power stage (you will need a host controller to update all the modulators). See picture below. The advantage is that you are not limited to any setup. You can buid it into a amplifier enclosure or into the speaker or sub. Or you can build multi room setups. Imaging that you build everything into the speaker. That would be so cool!
- Programmable DSP (they call it DAP): TI has software available for gain, EQ and crossover programming
- PSVC volume control: a combination of digital gain and power supply to control volume. This minimizes loss of resolution and keeps SNR high.
- Wireless speakers using a radio frequency link (uncompressed/lossless/error correction) you can have 1 speaker with a wireless master streamer and other speakers/subs with wireless slaves.
- Host controller: the CC85xx can be the I2C master and control the other devices (write default register setup to the modulators on system startup).
- Closed loop amps: TAS56xx series
Something like this:
What kind of delay span are you anticipating when you use WiFi to distribute several legs of what shall be a coherent music reproduction system?
I would not trust wifi as it is an open band with a lot of interference which could mean resending - and if, your synchronicity between L and R / sub goes down the drain.
The streaming is not via Wifi, but via RF. In the picture I use Wifi for instance for an Apple Airplay or Google Chromecast device. The picture would be more clear when I would have drawn them as an external device.
From the TI brochure for the CC8521 and CC8531 chips:
http://www.ti.com/lit/ml/swab002/swab002.pdf
From the TI brochure for the CC8521 and CC8531 chips:
http://www.ti.com/lit/ml/swab002/swab002.pdf
OK! Still, I would be reluctant to use an open band like 2,4 or 5 Ghz as interference can be so high and unpredictable. Re-sending is not an option - maybe if there was a a large delay budget with a general 1 sec delay for the purpose. Still, for high-end I will rely on cables - optical is my preferred choice and with so high bandwidth that no clock would be required at the D/A end while still preserve adequate jitter performance - 400Gbps? Why? Because in order to make a 10cm analog path it requires the DAC to be very close to the speaker driver which in turn means high levels of acoustic pressure which would effect any oscillator performance and thus create jitter - actually very close to the good ol' turntable feedback problem 🙂 Instead the clock (singularis!) could be placed far away in a dead silent place to create a global bit and frame clock distribution network which together with a control system sets delay to individual channels for time alignment. But maybe PurePath does this in some way - must admit I have not read up on this...
//
//
//
//
You could be right on this. I have no idea how this would work in the real world. On the other hand, in IT we have very reliable Wireless high speed communication over much longer distances. So it must be possible to do this between devices that are only a few feet apart. They have evaluation boards. I quess that would be the easiest way to try it.
I do not understand why you need a DAC if you use a PWM modulator and amplifier with PWM input. Maybe I misunderstood your setup.
I do not understand why you need a DAC if you use a PWM modulator and amplifier with PWM input. Maybe I misunderstood your setup.
- Status
- Not open for further replies.