Cascading ADL5611, power supply issues

I want to cascade 3 RF gain stages using ADL5611. I power each ADL5611 directly from a laboratory power supply of 5V. Individually each IC draws a current of 85-90mA. But when I power up two of them together the current reaches 400 mA(170mA expected). I have used a microstrip design on Arlon AD600 substrate and used a layout similar to the evaluation board. This combination occasionally works but mostly gives unexpected results. Can you kindly help resolve this issue?

  • I have seen the same problem. We integrate two ADL5611's into the circuitry and notice that the ramping profile of two amps changes around 3V, it goes up to 360mA of total current consumption and then ramps down to 200mA at 5V. The solution seems to be using a regulator with a transient output of 500mA.

  • 0
    •  Analog Employees 
    on Apr 4, 2018 10:46 PM

    Hi,

    Can you share your schematic of the 3x cascaded circuit? 

    Thanks,

    David

  • As you know, it's very hard to simulate the ADL5611 model but here is the screen shot of the dynamic voltage response (Yellow) and the current(Pink).

    The circuit schematic is

    The right side is to simulate the transient response from the ADL5611 block. There are complex RLC response after the supply voltage passes ~3V. Prior to this threshold voltage, ADL5611 behaves more or less like a resistive load. The complext RLC response is simulated using a voltage controlled switch to turn it on and off.

  • 0
    •  Analog Employees 
    on Apr 14, 2018 12:26 AM

    The screenshot of the current waveform does not look right and the device looks like it is oscillating. I cannot tell from the simulated model but are you using DC blocking capacitor for RF input/output and bypass capacitors on the Vcc line? Figure 31 and Figure 32 from the datasheet shows the recommended layout and schematic for ADL5611.

  • The layout is strictly followed and the DC blocking caps are used in between the RF ports, like it in the original design. This oscillation is in the biasing circuitry, and it is called inrush current oscillation. Typically it occurs after a switch is turned on and the supply source sees a complex RLC load and adjust to this load. The hypothesis to this issue is that there is a parasitic diode (?) switch in the ADL5611 gain block. When the supply voltage passes the threshold region, around 3-4V, a switch turns on and the supply sees a different load.

    The easiest way to verify all the customers' complain is to DC voltage bias the sample circuit provided in your reply, measure the I-V characteristic of the gain block using an external power supply with 0.25V step, such as 2.75V, 3V, 3.25V, 3.5V, 3.75V, 4V, etc to 5V. You will see a large current draw around 3.5V and it drops down around 5V. 

    This is fine for external power supply, but for it to be useful in circuits, one needs to use a much bigger power circuitry to tolerate the current draw during initial bias voltage ramping through the 3.5V range. Or, like what I did, use a little inferior RF performance gain block as ADL5545.