Help for design in AD8138 or similar FDA as ADC Driver

Hi ChristianSE ,

We'll look into this.

Hi Gauzz 

Can you help Christian with the FDA query above?

Thanks,

Gio

Hi Christian,

Regards,
Goz

Hi Goz,
yes, the application was until now designed as a single-ended system.
The problem we see is now the output impedance of the VGA (AD8336), that's why we MUST introduce voltage-follower. The FDA provides a good method for converting single-end-input to differential-output.

You suggest to use the AC-coupled termination method, so we will use this.

My biggest question mark is the output impedance of the VGA (AD8336). I can't find anything in the datasheet.
The termination should be equal to the output impedance of the VGA, correct?

Can you confirm, that it is not possible to run in LT Spice TR-Analysis for the VGA (AD8336)?
Which termination value you would suggest to use?

Thanks,
Christian

Hi Christian,

Output impedance of pre-amp is in the TPC section.

Pre-Amp

VGA

Transient Analysis is possible for the part. May I know what LTspice simulator version you are on? It is working well on my end LTspice ver. 26.0.0.



Regards,
Goz

Hi Goz,

thanks for your reply.

Just like to note, there is a mismatch from the output in the schematic I have provided. Pins 4/5 must be swappe.

I have build up the simulation again with the latest version and now it's working.

I really don't know, why I haven't seen the output impedance curves. Thanks.

Ok, but know I got another question mark.

From the inital consideration and measurements I have done, the output impedance seems to be in the region near 1k.

When sourcing the single ended input of the ADC from a signal generator (50 Ohm output), we didn not see the problem in the signal.
But, when we try to amplify the output signal, around 300 mV, we are seeing this kind of effect and this happens only in direct connection between the VGA output and ADC input (in single-ended terminology)

VGA --> ADC(+) in single-ended

Can you explain what is happening in detail? Again, this only happens, when trying to increase the gain. Until a certain threshold, everything is fine.

We only tested with a 10 kHz signal, where the output impedance will be < 10 Ohm.

So, I'm really not sure anymore, if the FDA would really help. The output impedance would be the same.

I will build in now during next week the FDA solution between VGA and ADC (changing to the differenital input also).

I'm really happy for any advice and explanation, what is going on here.

Thank you,
Christian

Hi Christian,

Kindly include FDA in the middle of your circuit for your measurement. While the AD8336 is an excellent VGA, it is not optimized to drive a switched-capacitor ADC directly. The AD8138 isolates the VGA from the ADC's sampling 'kickback' and provides the low-impedance drive required for the ADC inputs.

Regards,
Goz