Max Error in 8-bit ADC: 7.55mV

[asp_physics]

Homework Statement

An 8 bit ADC has a full scale input of 2.55V. If other cumulative errors are 2.55mV, determine the maximum error.

Options are: 10mV, 2.55mV, 12.55mV, 7.45mV

Homework Equations

I believe this is a conceptual question. The only equation I can come up with is that the quantization error in an ADC is at max, half of LSB.

The Attempt at a Solution

Assuming cumulative errors as mentioned in the question does not include quantization error (LSB/2),
maximum error
= Q error + C error
=5mV + 2.55mV
= 7.55mV

If cumulative errors include Q error, then max error = 2.55mV as the other errors that can creep into an ADC are non-linearity and aperture errors, both of which more data to calculate.

 

[berkeman]

Homework Statement

An 8 bit ADC has a full scale input of 2.55V. If other cumulative errors are 2.55mV, determine the maximum error.

Options are: 10mV, 2.55mV, 12.55mV, 7.45mV

Homework Equations

I believe this is a conceptual question. The only equation I can come up with is that the quantization error in an ADC is at max, half of LSB.

The Attempt at a Solution

Assuming cumulative errors as mentioned in the question does not include quantization error (LSB/2),
maximum error
= Q error + C error
=5mV + 2.55mV
= 7.55mV

If cumulative errors include Q error, then max error = 2.55mV as the other errors that can creep into an ADC are non-linearity and aperture errors, both of which more data to calculate.

I think your solution is correct. Half of an LSB plus the other error...

 

[biplabsahoo10]

There is no such ans like 7.55mv.
I think...
Quantization Error is (ε Max)= 2.55v/(2^8)=10 mv
so total error = 10mv+2.55mv = 12.55 mv(Correct Ans)

As per ur concept," if Qunts. error contain in 2.55mv other Cumulative error " which is not possible.

10-2.55 =7.45 mv(not the Ans)

 

[NascentOxygen]

Although you might think an ADC module should be designed to have a quantization error of ½LSB, in practice it may be the case that max quantization error = LSB. This comes about by max input (corresponding to output=11111111) being 127/128 Vref.