Chopper amplifier offset measurement

[deba_fire]

Hi All,

How should one measure the offset of chopper amplifier?

I have made an circuit with gain of 50 with this chopper amplifier.And I have shorted the inputs,on the differential output I see chopped output at chopper frequency.How can we estimate offset from this waveform?

Thanks

 

[FvM]

In my understanding, a chopper amplifier is comprised of a modulator, AC amplifier and a demodulator. All parameters are usually determinded for the complete chain, applying a DC input voltage and measuring the DC output voltage. Because the demodulator is phase sensitive, you can't "see" the related DC level from the AC waveform.

 

[deba_fire]

Hi FvM,
The chopper amplifier which I am using has the full chain,modulator,amplifier and demodulator.Can we relate the AC waveform at the output to DC input to determine the offset?Or are there other ways of knowing offset?The amplifier is supposed to have 1uV offset.

 

[FvM]

I don't understand, why you are looking for an AC waveform. An input offset should involve an output DC voltage, isn't it? You didn't tell about the amplifier design. Is it an OP type ("infinite" gain) or a fixed gain amplifier? If it's of the OP type, you need a feedback network to measure the input offset voltage. You can basically refer to the offset measurement methods for OPs. If it's fixed gain, just short the input and measure the output DC voltage.

 

[dick_freebird]

Are you looking for the natural offset that is being chopped
away, or are you looking for the in-application residual offset?
The latter just wants a filter that can roll off the chop. The
former wants you to stop the chop in both of the states and
take half of the difference (if state is unknown) as an abs val
or force it to a known "right side up" state and direct measure.

In production you probably want to throw away amps that
are really crappy but chop to a "good enough" Vio. Just on
the theory that something's wrong, you don't know what,
so don't ship it.

Net Vio depends on the DC Vio, and the chop clock duty
cycle symmetry. Only at 50/50 do you get a true cancellation
and if there is any output current asymmetry you can get
some second order chop effects (duty same, slew rate not,
"effective duty" != 50%).

 

[deba_fire]

This is the block diagram of the chopper amplifier.Crosses denote chopper.I have access to differential output and the single ended output after the last stage.I haven't put the low pass filter.I have read somewhere(unable to find reference) that the rms ripple on the output is the offset.That's why I asked for the mapping of offset to the output AC waveform.

 

[deba_fire]

Hi All,

Can someone help me how rms ripple is related to chopper input ofset?

 

[FvM]

You didn't yet tell, which offset you are looking for, please refer to the question posted by dick_freebird.

Can someone help me how rms ripple is related to chopper input ofset?

What's the reason, you bring in rms ripple as a new term to the discussion? How do you measure it?

 

[deba_fire]

I have one doubt:why do we define two offsets one in-application and other in two different chop states?is there any relation between them?I think the in-application chopper offset is more important parameter to measure.
The rms ripple can be measured by oscilloscope.

 

[dick_freebird]

What's important depends on who's asking. The circuit guy
probably just cares about error components. The quality
guy wants to know that the circuit isn't "hiding its flaws"
that could continue to worsen with aging. That's why we
ask -you- what -you- are after.

Unfiltered, your RMS ripple will be about 2X unchopped Vio.
Filtering pushes this down, at the cost of usable BW.

 

[deba_fire]

My main criteria is offset in chopping action,in-application