Hi,
this is my first post here and i hope that my question is not too silly and i use the correct forum.
For a project i need to playback ultrasound (40-50 kHz ...) with a dsp. This ultrasonic frequency needs to be sent to a grid of ultrasonic transmitters.
Now i have been searching for an affordable evaluation board and found the texas instruments TMS320C6748 DSP Development Kit (LCDK).
The data sheet tells me that there is a McASP in use that can handle a sample rate of 192 kHz. Does this mean that i can generate a maximum frequency of 87 kHz at the output or are there other limiting factors like a low pass filter with cutoff at 20 kHz or something else?
What is the maximum analog frequency that can be achieved at the outputs?
Do you perhaps have an other good idea how this can be done?
I would be very grateful for any help.
Thank you in advance!
this is my first post here and i hope that my question is not too silly and i use the correct forum.
For a project i need to playback ultrasound (40-50 kHz ...) with a dsp. This ultrasonic frequency needs to be sent to a grid of ultrasonic transmitters.
Now i have been searching for an affordable evaluation board and found the texas instruments TMS320C6748 DSP Development Kit (LCDK).
The data sheet tells me that there is a McASP in use that can handle a sample rate of 192 kHz. Does this mean that i can generate a maximum frequency of 87 kHz at the output or are there other limiting factors like a low pass filter with cutoff at 20 kHz or something else?
What is the maximum analog frequency that can be achieved at the outputs?
Do you perhaps have an other good idea how this can be done?
I would be very grateful for any help.
Thank you in advance!
Hi Funkandsoda,
That's an interesting project you have here.
As you know, at the digital level there's no problem with handling higher frequency components.
Some problems might appear at the digital to analogue conversion, at 2 levels.
First level is the digital reconstruction filter. The only way for you to know is looking at the DAC datasheet. You may find a spec like 'Passband (-0.1 dB) is 0.47 x sampling frequency', which would be good for you, or 'Passband (-0.1 dB) is 0.24 x sampling frequency' - which is not good.
The second level that might cause issues is the analogue output buffer, which usually includes a low order low-pass filter as well. This one doesn't depend on the sampling frequency: it's an analogue filter with fixed specs. Either your documentation includes a description, or you'll have to simulate that in software.
Best and good luck
Nick
That's an interesting project you have here.
As you know, at the digital level there's no problem with handling higher frequency components.
Some problems might appear at the digital to analogue conversion, at 2 levels.
First level is the digital reconstruction filter. The only way for you to know is looking at the DAC datasheet. You may find a spec like 'Passband (-0.1 dB) is 0.47 x sampling frequency', which would be good for you, or 'Passband (-0.1 dB) is 0.24 x sampling frequency' - which is not good.
The second level that might cause issues is the analogue output buffer, which usually includes a low order low-pass filter as well. This one doesn't depend on the sampling frequency: it's an analogue filter with fixed specs. Either your documentation includes a description, or you'll have to simulate that in software.
Best and good luck
Nick
Sorry if it's naive, but is it a "must" that your signal is digital?
You mention doing DSP on it, so I guess it *is* a must, I mean it's not just a simple thing at 50KHz, like sinusoidal, that you could generate with analog circuitry, right?
-
Thank you all for replying and helping me.
Yes, unfortunately i need to process the input signal (music) with a psychoacoustic effect algorhythm on the dsp. Afterwards the processed signal needs to be amplitude modulated with the ultrasonic carrier and then amplified to send it to the transmitter grid.
As a workaround, i'm thinking about processing the psychoacustic effect with the dsp in a first step, then doing the amplitude modulation with the ultrasonic carrier in an analog IC and then in the end amplyfing it with another IC and send it to the grid.
I unfortunately couldn't find something like "Passband" or "analog output buffer" in the datasheets so i wrote to the texas instruments suport and asked them for help...
Yes, unfortunately i need to process the input signal (music) with a psychoacoustic effect algorhythm on the dsp. Afterwards the processed signal needs to be amplitude modulated with the ultrasonic carrier and then amplified to send it to the transmitter grid.
As a workaround, i'm thinking about processing the psychoacustic effect with the dsp in a first step, then doing the amplitude modulation with the ultrasonic carrier in an analog IC and then in the end amplyfing it with another IC and send it to the grid.
I unfortunately couldn't find something like "Passband" or "analog output buffer" in the datasheets so i wrote to the texas instruments suport and asked them for help...
The McASP just copies audio from memory and blasts it out as a serial I2S/TDM/whatever stream... and vice versa. There's no audio filtering or anything that happens in the peripheral, and there's no hard limit that says the port is limited to a sample rate of 192KHz.
Technically the port can run a serial clock of 40MHz. With 16 bit stereo I2S, you're good for a sampling rate of 2.5MS/s. You'll be fine.
Technically the port can run a serial clock of 40MHz. With 16 bit stereo I2S, you're good for a sampling rate of 2.5MS/s. You'll be fine.
Hey gmarsh, thank you very much for your reply.
So if i understood you correct, 40-50 kHz won't be a problem as long there is no analog filter after the McASP on the board, right?
I'm sorry, i'm still a noob. I'm just at the beginning and haven't worked with DSPs before.... I tried to ask Texas Instruments but they don't answer my question since i'm just a student. They sent me a mail and told me that they don't answer people without a reference to a company...
Is there some modification needed to get the higher sample rates that you mentioned on the McASP or can this sample rates easily be configured on this evaluation boards?
@ Julf: You are right. There is no special reason to do the amplitude modulation in the digital domain. I only thought that it would be more convenient to have all the processing done on the evaluation board without help from other ICs. If this doesn't work, doing the amplitude modulation in the analog domain with a multiplier IC would be my workaround.
The effect i try to work on, can be done with different types of modulation as AM, FM or PWM. It also would be easier to check which modulation type works best if i could do this right on the DSP in software without the help of different ICs.
Thank you
So if i understood you correct, 40-50 kHz won't be a problem as long there is no analog filter after the McASP on the board, right?
I'm sorry, i'm still a noob. I'm just at the beginning and haven't worked with DSPs before.... I tried to ask Texas Instruments but they don't answer my question since i'm just a student. They sent me a mail and told me that they don't answer people without a reference to a company...
Is there some modification needed to get the higher sample rates that you mentioned on the McASP or can this sample rates easily be configured on this evaluation boards?
@ Julf: You are right. There is no special reason to do the amplitude modulation in the digital domain. I only thought that it would be more convenient to have all the processing done on the evaluation board without help from other ICs. If this doesn't work, doing the amplitude modulation in the analog domain with a multiplier IC would be my workaround.
The effect i try to work on, can be done with different types of modulation as AM, FM or PWM. It also would be easier to check which modulation type works best if i could do this right on the DSP in software without the help of different ICs.
Thank you
Oh, okay.
So is it possible to route the digital output of the McAsp to one of the non audio outputs of the evaluation board to use a different DAC or is this impossible and i have to be satisfied with the one included on the board?
The TMS320C6748 DSP Development Kit for example has these interfaces:
• USB serial port
• Fast Ethernet port (10/100 Mbps)
• USB host port (USB 1.1)
• USB OTG port (USB 2.0)
• SATA port (3 Gbps)
• VGA port (15-pin D-SUB)
• LCD port (Beagleboard-XM connectors)
• 3 audio ports
• 1 line in
• 1 line out
• 1 MIC in
• Composite in (RCA jack)
• Leopard Imaging camera sensor input (32-
pin ZIP connector)
• Authentic fingerprint sensor
So is it possible to route the digital output of the McAsp to one of the non audio outputs of the evaluation board to use a different DAC or is this impossible and i have to be satisfied with the one included on the board?
The TMS320C6748 DSP Development Kit for example has these interfaces:
• USB serial port
• Fast Ethernet port (10/100 Mbps)
• USB host port (USB 1.1)
• USB OTG port (USB 2.0)
• SATA port (3 Gbps)
• VGA port (15-pin D-SUB)
• LCD port (Beagleboard-XM connectors)
• 3 audio ports
• 1 line in
• 1 line out
• 1 MIC in
• Composite in (RCA jack)
• Leopard Imaging camera sensor input (32-
pin ZIP connector)
• Authentic fingerprint sensor
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.