Author Topic: DC biasing AC signal for ADC  (Read 4488 times)

0 Members and 1 Guest are viewing this topic.

Offline initTopic starter

  • Contributor
  • Posts: 26
  • Country: au
DC biasing AC signal for ADC
« on: April 11, 2016, 03:32:13 pm »
Hi guys,

I have an AC signal source that I am measuring and amplifying to be eventually read by an ADC. I haven't really toyed with this stuff very much before and I thought I would ask some basics. I want to boost my AC signal to half of the ADC max (likely 3.3 V or 5 V) such that the signal is between 0 - V(max). What is the best method to bias an AC signal for ADC? I know I could use a simple voltage divider with a DC blocking cap connected to the AC feed but I've seen more advanced topologies used. What types are there and what are the advantages/disadvantages?
 

Offline tszaboo

  • Super Contributor
  • ***
  • Posts: 7307
  • Country: nl
  • Current job: ATEX product design
Re: DC biasing AC signal for ADC
« Reply #1 on: April 11, 2016, 04:17:18 pm »
For example a difference amplifier with a Vmax/2 reference as the second input.
That amplifier can be a the ADC driver itself.
 

Offline danadak

  • Super Contributor
  • ***
  • Posts: 1875
  • Country: us
  • Reactor Operator SSN-583, Retired EE
Re: DC biasing AC signal for ADC
« Reply #2 on: April 11, 2016, 04:43:21 pm »
In this zip file is a tool to calculate using two R's that will allow
offsetting a signal and DC couple it to boot. Or how to develop
a - V to allow OpAmps to perform level shifting.

Regards, Dana.
Love Cypress PSOC, ATTiny, Bit Slice, OpAmps, Oscilloscopes, and Analog Gurus like Pease, Miller, Widlar, Dobkin, obsessed with being an engineer
 
The following users thanked this post: ianxtreem

Offline Alex Trofimov

  • Contributor
  • Posts: 43
  • Country: ru
Re: DC biasing AC signal for ADC
« Reply #3 on: April 12, 2016, 11:10:11 am »
I think, basic capacitor coupling is good enough for most applications under 100MHz if the source impedance is low enough. If it's high, use solution suggested by NANDBlog.
There's a place for topology complication. If use buffer, I'd add a cap to Vref/2 divider, so it works as a Lo-Pass filter and don't pick up garbage. If impedance matching is in demant there's another reason for topology to be evolved.
Comparing the two suggested schemes, DC-cap add some loss, ADC input currents may cause errors on the divider resistors, if you need to work with low frequencies, a cap may be just to big, etc) Op-amp solution is free of these disadvantages, but it's basically more pricey and takes more room on pcb and in the power budget as well.
There's a lot of subtleties, and I don't even know do you need them. :-)
And if you need to decoule the AC signal from the signal which contains both DC and AC, then there's not much of a choice but using decoupling cap anyway. Of course, there is (or I better say must be) some other clever ways, but I myself didn't give them much of a thought.
Sorry for the rant style) Dont really have time to edit and compile all this.
 

Offline broz

  • Regular Contributor
  • *
  • Posts: 50
  • Country: ca
  • Pronounced like the "bros." in "Super Mario bros."
    • My Wordpress Page
Re: DC biasing AC signal for ADC
« Reply #4 on: April 12, 2016, 03:43:42 pm »
Not too sure if this would work well or not, but what about using an op-amp as a voltage follower while having your desired bias voltage fed into the "offset null" pin?
Slowly but surely making my way through EE school
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16272
  • Country: za
Re: DC biasing AC signal for ADC
« Reply #5 on: April 12, 2016, 05:44:44 pm »
Not too sure if this would work well or not, but what about using an op-amp as a voltage follower while having your desired bias voltage fed into the "offset null" pin?

Nulling pins are meant only to reduce the opamp offset voltage alone, not to provide any voltage shift. internally they are a tap on the input long tailed pair emitter resistors that are there to compensate for any slight difference in the value of these resistors. The offset null method you want to do will be absolutely guaranteed to both drift with time, temperature and the input voltage, as well as it will likely saturate the opamp internally. Most offset nulling pins call for a resistor of around 1-10k from each pin to the correct supply rail as per datasheet, and the resistance in the opamp is in the region of 100R to 1k in all cases.

To provide a DC offset you would need an analogue adder, and this is most commonly done using a virtual ground to add the signal and an offset current, and having a second inverting amplifier to get a positive voltage for the ADC. this needs a second negative supply, or you will have to use a virtual ground on the 5V rail and accept the reduces signal swing into the ADC if you only have a single 5V rail.
 
The following users thanked this post: broz


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf