Hello.
Is there anyone who can explain the differences between delta-sigma DAC and MDAC multiplyngDAC?
Advantages and disadvantages for audio applications?
Is there anyone who can explain the differences between delta-sigma DAC and MDAC multiplyngDAC?
Advantages and disadvantages for audio applications?
Multiplying DACs have two inputs. One input is for the digital sample that is to be converted and output as it's equivalent analog value. The other input is an ANALOG input which can be utlized to scale (to multiply) the analog output value. For an simplified example, suppose that the digital sample value were 5, and the analog input were 1. Then, the analog output value would be; 5x1 = 5. Now, suppose the analog input value was increased to 2. Then, the the analog output output value would be; 5x2 = 10. Notice that the digital input value was not changed, only the analog input value.
While the MDAC architectures I recall seeing have not been Sigma-Delta, I don't know offhand that Sigma-Delta-Modulation is necessarily precluded from use in an MDAC.
While the MDAC architectures I recall seeing have not been Sigma-Delta, I don't know offhand that Sigma-Delta-Modulation is necessarily precluded from use in an MDAC.
Last edited:
I'm not sure what you mean by an MDAC multiplying DAC. Any DAC multiplies its input code by its reference and some scaling factor.
The advantage of delta-sigma DACs or sigma-delta DACs (two terms for the same thing) is that they can give good performance without ridiculous component matching requirements. The disadvantages are that they produce lots of ultrasonic noise that need to be filtered off somehow and that they are extremely sensitive for crosstalk from the sigma-delta modulate to the reference or to the clock, and from any clock signal that is not at an integer multiple of the sigma-delta clock frequency onto their reference.
Of course if you build a standard application of a sigma-delta DAC chip and follow the recommendations for PCB layout, then the manufacturer of the DAC chip has already solved these issues for you. Then again, if you build a standard application of a multibit DAC without sigma-delta modulation, the manufacturer has solved all matching problems for you.
If you are into valve circuits, then an advantage of sigma-delta technology is that the actual DAC circuit gets simple enough to make it with valves. I did that and needed no more than ten valves for the voltage reference (85A2), clock generator (ECC81), clock buffer (two times EF80) and two DACs (left three times E88CC and right three times E88CC) together.
The advantage of delta-sigma DACs or sigma-delta DACs (two terms for the same thing) is that they can give good performance without ridiculous component matching requirements. The disadvantages are that they produce lots of ultrasonic noise that need to be filtered off somehow and that they are extremely sensitive for crosstalk from the sigma-delta modulate to the reference or to the clock, and from any clock signal that is not at an integer multiple of the sigma-delta clock frequency onto their reference.
Of course if you build a standard application of a sigma-delta DAC chip and follow the recommendations for PCB layout, then the manufacturer of the DAC chip has already solved these issues for you. Then again, if you build a standard application of a multibit DAC without sigma-delta modulation, the manufacturer has solved all matching problems for you.
If you are into valve circuits, then an advantage of sigma-delta technology is that the actual DAC circuit gets simple enough to make it with valves. I did that and needed no more than ten valves for the voltage reference (85A2), clock generator (ECC81), clock buffer (two times EF80) and two DACs (left three times E88CC and right three times E88CC) together.
Last edited:
- Status
- Not open for further replies.