AnsweredAssumed Answered

AD5142A as T-attenuator or voltage divider

Question asked by fuchs on Mar 8, 2016



I just designed a moduled AGC-circuit. Amongst other things there is a fixed gain module and an adjustable attenuation moduel (AD5142A).


I got problems with the attenaution module. It's just a AD5142A connected as follows...

Signal goes into R_A, comes out over R_W and R_B is connected to GND (don't wonder, its german):AD5142A-Circuit_contactSupport.jpg

-> I would call it a T-attenuator and I expect a voltage dividing effect set by  R_A / R_B


I'm wondering why my attenuation decreases with increasing frequency when I measure the circuit with a PicoScope (power is supplied by a laboratory power supply)

The datasheet makes me think that attenuation is as expected til some Mhz (the graph is just horizontal til ~3MHz) but it just does t very low frequencies. At some kHz the amplitude starts to decrease with the frequency sweepAD5142_2k5_lowF2(1k-10M).jpgAD5142_2k5_xhighF2(1k-10M).jpg

The network analyzer shows off some exponential impedance decrease (Z11 and Z22) too what makes me pretty sure that there is something wrong. Maybe the AD5142A can't be used as a attenuator / voltage divider?


I wont be able to use this as an attenuator in my AGC with this behavior.

Has anyone an idea where I could be wrong or what I could improve?



Thanks and Greetings