Differential signal queries

[Ardni]

Hi,
I´m having some problems understand a circuit. We are using a a differntial ADC driver, the ADA4937 before the input to the ADC (AD9640).

Please find attached our circuit and the data sheet of the differential driver. We would be very grateful if someone could shed some light on the following questions:

1) We have a differntial signal applied to the inputs of the differential driver. If these inputs go negative, what will happen given that the ADA4937 does not have a -ve supply? Must the applied differential signal have a DC bias ideally in the mid point of the power supply? (The power supply is 3.3V)

2)It does not seem to be specified in the datasheet what the maximum voltage levels can be on the inputs to the differential driver. I was expecting it to say something like the limit would be the from gnd to Vcc but I cannot find any reference to it.

3) Can someone explain what a differential analog signal would look like in relation to a single ended signal?

Many thanks for any help

 

[ericgibbs]

Hi,
I´m having some problems understand a circuit. We are using a a differntial ADC driver, the ADA4937 before the input to the ADC (AD9640).

Please find attached our circuit and the data sheet of the differential driver. We would be very grateful if someone could shed some light on the following questions:

1) We have a differntial signal applied to the inputs of the differential driver. If these inputs go negative, what will happen given that the ADA4937 does not have a -ve supply? Must the applied differential signal have a DC bias ideally in the mid point of the power supply? (The power supply is 3.3V)

2)It does not seem to be specified in the datasheet what the maximum voltage levels can be on the inputs to the differential driver. I was expecting it to say something like the limit would be the from gnd to Vcc but I cannot find any reference to it.

3) Can someone explain what a differential analog signal would look like in relation to a single ended signal?

Many thanks for any help

hi,
Look at this image for single supply input CM range.

Ref Q3: do you mean the input or output signals.??

 

[ericgibbs]

hi,
This is the diffamp from your datasheet.
Using LTspice sim.
A appreciate the test freq is low.

 

[Ardni]

Hi Eric,
Many thanks for both of your replies. Things are more clear now. I should have spotted the range given in datasheet before.

It says in the datasheet that the ADA4937 is optimized for level-shifting ground-referenced input signals but that the input range of a 3.3V supply is between +0.3V and 1.9V. Maybe I am not understanding what is meant by ground referenced, but I thought it was a signal centered on 0V that would go negative? If possible could you please clarify what it is I have misinterpreted?

Also in your simulation (and thank you for taking the time to post it) I noticed that the input signal was between 100mV and -100mV but the chip was powered between +5V and Gnd. I don´t fully understand why you can send a signal to the inputs which goes negative when the chip does not have a negative supply.

Many thanks again for your help. I´m very grateful

 

[ericgibbs]

Hi Eric,
Many thanks for both of your replies. Things are more clear now. I should have spotted the range given in datasheet before.

It says in the datasheet that the ADA4937 is optimized for level-shifting ground-referenced input signals but that the input range of a 3.3V supply is between +0.3V and 1.9V. Maybe I am not understanding what is meant by ground referenced, but I thought it was a signal centered on 0V that would go negative? If possible could you please clarify what it is I have misinterpreted?

Also in your simulation (and thank you for taking the time to post it) I noticed that the input signal was between 100mV and -100mV but the chip was powered between +5V and Gnd. I don´t fully understand why you can send a signal to the inputs which goes negative when the chip does not have a negative supply.

Many thanks again for your help. I´m very grateful

hi,
Some OPA's operating from a single supply will accept a input signal less than 0V to about -0.5Vish.

Ground ref in this context means the input signal must not swing below +0.3V when the single supply is +3.3V

Example: say you have a peak to peak input signal of say 1Vac, [0.5Vpk] that would have to superimposed on a 'dc' level of +0.8V.
So the excursion of the 'ac' part would swing from +0.3V [min] to +1.3V [max]
in order to meet the requirement of this OPA.

EDIT:
Look at these diff amp application sheets.

 

[Ardni]

Hi again Eric,
Thank you for all your help so far. We have been discussing this issue amoungst ourselves and have become confused once more. One member of our team tells us that in our configuration (in the circuit attached in the first post), the ground from the source of the differential signal is not connected to the ground supply of the ADA4937 therefore applying lets say a voltage of -1V with only induce a negative current and will not get clipped.

Is this true? that if the grounds are not related and a differential signal is applied then it won´t clip when it goes negative if there is no negative supply.

Is the paragraph in the datasheet "INPUT COMMON-MODE VOLTAGE RANGE IN SINGLE-SUPPLY APPLICATIONS" referring to a differential configuration or single ended config or both. My own interpretation is that it refers to both, but some of my colleges disagree.

Once again thank you so much for your help

 

[ericgibbs]

hi,
Are you saying signal source is floating or the grounds are not common to the source and OPA.??

Any chance you could post a drawing showing how the source is connected to the OPA.??


EDIT:
is this what you have in mind for the isolated source.?
Note its using a LTC1992 for simulation.

 

[Ardni]

Hi Eric,
Yes the signal source has a different ground. Unfortunatley I don´t have much info on it, as it is another board containing RF circuitary which passes a differential signal. The differential signal is connected through a connector between the 2 boards and both boards do not have a common ground.

I think that the signal that enters is a differntial signal centered on 0V

Unfortunately I dont have a drawing of the circuitary on the other board, so unfortunatley I can´t provide any further info other than the circuit attached in the first post.



Please let me know if you nhave any further ideas or need any further info.

Thanks again

 

[ericgibbs]

Please let me know if you nhave any further ideas or need any further info.

Thanks again

hi,
Using the vmid to the OPA of 3.3V/2 should centre the diff input signals at approx 1.65V.
I would set up a test link with the diff source and the ADA4937 and try it.

 

[crutschow]

Yes the signal source has a different ground. Unfortunatley I don´t have much info on it, as it is another board containing RF circuitary which passes a differential signal. The differential signal is connected through a connector between the 2 boards and both boards do not have a common ground.

Standard differential receivers will not work reliably if the two grounds are not tied together at some point. If the two ground potentials differ by more than the common-mode voltage range of the input, the receiver will not work.

If the two grounds can not be tied together for some reason, then you will need to use an isolation type amp that can transfer a signal between grounds that have a significant voltage difference.