ADA4938 common mode setting

Hi Harry,

thanks for your reply. So far so good.

so I'm assuming there is a 20k/20k divider internally.

Thats exactly what i want to know. Because, as you mentioned, this would load my external divider. So for calculating the right external values i need to know the internals. But your 20k/20k assumption might be wrong because i tried R1=20k (R2=open). With your assumption this should lead to Vocm = 1.67V but actually i get only 0.4V!

WRT "We tried different external resistors but it seems that the common mode voltage is different with every ADA4938."

How different?  1mV??  100mV? 1V??

At first we tried R1=10k and R2=2k87 on a test PCB and got Vocm~0.9V. Great. Now on a final AD converter PCB we used the same values and suddenly got Vocm~0.27V. With the same values of R1 and R2 and same dual supply (+/- 5V) voltages. How can that be?

It would be great if some AD experts could reveal the mystery about the internal wiring of ADA4938...

Julian

Hi Jino,

thanks for your participation. The schematic is as follows:

I am confused about your suggestion:

Vocm is driven by low impedance voltage source, for example, using low value resistors to set the Vocm.

Because in the datasheet it is mentioned (page 21):

"In cases where more accurate control of the output common-mode level is required, it is recommended that an external source or resistor divider (10 kΩ or greater resistors) be used."

In my opinion 10kΩ is not low impedance. So what should i do now: low impedance or >10kΩ resistor devider??

Best regards

Julian

Hello Julian!

Two options:

1. Remove R15(2k87) and set R14 to 3k9. With the 10k input impedance of the ADA4938 and AVDD of 1.8V, you will expect a VOCM of 1.295V. or

2. Implement a resistor divider setting VOCM as an ideal voltage source, means that the impedance driving the VOCM should be low( 100's of ohms low).

For debugging, did you use the same resistor divider in your prototype and your final board for VOCM? Please remove R5 and R9, then measure the VOCM.

Regards.

Hello Jino,

sorry for the late reply. We've found the correct wiring for our application.

1. We removed R15 and left R14 at 10k. Now we get almost AVDD/2 on VOCM (~0.9V). But for a redesign i would rather use an additional OpAmp (impedance converter) to set VOCM to AVDD/2 independently of the exact value of AVDD.

2. We found out that our test PCB is not exactly the same as our final ADC PCB. On the test PCB we supplied AVDD with 5V and that is why we got different values for VOCM.

Thats it! I'm so sorry for the unnecessary excitement but you helped me a lot! Thank you!

Best regards,

Julian