Output impedance of the Vlog Pin of AD8309

Hi Avinash,

I'm guessing that you want to know the complex impedance of the VLOG output, that is, where is it on the Smith Chart. This is not really the correct way to think of this output. The output is more like an op-amp output (i.e. ideally 0 Ohms output resistance) than an RF Amplifier output (i.e. close to 50 ohms with some capacitance or inductance). So you should not be trying to do any kind of an impedance match between the AD8309 and the ADC. Since the AD8309 has such low output resistance, it should be possible to connect it directly to the ADC input if the ADC's input resistance is relatively high, say greater than 500 ohms. Alternatively, you could put an op-amp buffer between the two devices, particularily if the ADC input is a switching capacitive load. 

Best Regards

Eamon

Hi Avinash,

yes, that all makes sense. Your diff amp needs to do three things. It needs perform a single-ended to differential conversion. It needs to scale the AD8309 output voltage appropriately so that it maps to the input range of the ADC. And it need to set the common mode voltage of the differential signal so that it maps to the expected common mode input level of the ADC. ADA4945-1 looks to me like a good choice. (https://www.analog.com/en/products/ada4945-1.html). 

There are lots of other good options to consider also. Take a look at this selection table (Single-Ended to Differential Amplifiers | Analog Devices).

There is a calculator tool available for dimensioning the diff amp design (ADI-DiffAmpCalc | Analog Devices).

Best Regards

Eamon

hanks a lot Eamon, for correcting my understanding,

So we can just convert the signal output of the Vlog from the AD8309 to the ADC input range via a driver and connect it to the ADC. Since our ADC expects a differential input, we will need to connect a differential amplifier in between. Hence we wanted some understanding on the impedance.

From your above answer, we shall reduce the signal range of the AD8309 to match the ADC input and use a differential buffer in between.

Thanks for the answer.

Hi Avinash,

that sounds like a good plan. I don't think that you need a buffer amplifier between the AD8309 and the diff-amp. Just be cognizant of the current demand on AD8309. Your configuration may call for the AD8309 to sink current. The datasheet says that the part can sink 1mA and source around 50mA. So you may need to adjust the resistor sizes around the diffamp accordingly. Also I would check in the diff amp calculator or in LTSpice that the diff amp can indeed handle a closed loop gain of 0.5. It might be good to hit the diff amp with step response in LTSpice to ensure that it doesn't have excessive ringing. 

Best Regards

Eamon

Hi Avinash,

the (Log-Amp→Voltage Divider→G+1 Buffer→G+1Diff Amp) is conservative and completely un-problematic. Just select an op-amp that is unity gain stable that operates from a single supply (e.g. ADA4805-1). 

Eamon

Hi Eamon,

Thanks for the answer. We checked in the diff amp calc with the gain of the ADA4938 at 0.5, however, it suggested that the gain cannot be too less due to instability of the amplifier. We checked with LT-SPICE, but it wouldn't show anything on the stability.

Instead, to be on a safer side, we are thinking of using a voltage divider at the output of the AD8309 for equal division, to get the voltage range of the AD8309 to that of the ADC and then use the differential amplifier with a gain of 1. We can also use a buffer in between the voltage divider and the amplifier for stability.

Hence the chain will look as follows:

AD8309 (0.4V to 2.4V) --> Voltage Divider (0.2V to 1.2V) --> Non Inverting Amplifier (Gain = 1) -->Differential Amplifier(ADA4938-1) (Gain of 1) ---> ADC (LTC2256-14)

However, I do not know the implications of using this, especially wrt the stability/noise due to the use of the voltage divider. 

When we checked with the Differential amp that you had suggested (ADA4945-1), we were able to get the gain even below 0.5 to be stable in Diff Calc, but trying in the LT-SPICE ended nowhere as it requires various signals such as clamp and disable. We tried using signal ranges provided in the datasheet, but it didn't give us appropriate results due to which we are apprehensive to use the device for further analysis, owing to time shortage.

Request you to please provide your inputs on the chain.

Regards,

ARK