I wonder Control Voltage using HMC346AMS8GE?

Hi,

V1 voltage controls the series devices and V2 voltage controls the parallel device. To set the attenuation, V1 is adjusted between 0 to -4V. However, to ensure 50ohm impedance for all V1 voltage, V2 needs to be adjusted. Otherwise RF ports will not show good return loss.

There is also a reference attenuator inside the part. The transistor sizing and resistors are 10x scaled and they are also connected to V1 and V2. Thus the external loop reads the input impedance from pin I, and tries to adjust the input impedance to 500ohm via an external opamp loop and drive V2 voltage.

Thus, V1 is driven only as single drive but V2-I pins are needed to be connected to an external opamp loop.

Best Regards

HI.

i`m Test HMC348AMS8GE Results.

1. V2 Voltage adjust and V1 Voltage Fixed 

2. V1 Voltage adjust and V2 Voltage Fixed 

The Test result Correct?

it seems that the test result is slightly different from the data sheet.

I Wonder if the test method is Right or Wrong.

If want to adjust the Voltage with the master, Please let me know if it is like a test method.

Hi Sang,

HMC346A is a GaAs based VVA and we would expect part to part variance on the gate pinch-off of series and shunt devices on the attenuator. Therefore, it is not meaningful to create a look-up table and apply to every part. 

The external I-V2 loops ensures the reference attenuator's input impedance to be 500ohm and main attenuator's input impedance to be 50ohms. If this is not set, mismatch losses would add on the attenuation, and this creates uncertainties on the end-to-end attenuation.

For experimenting purposes, if you still want to drive V1 and V2 separately, I suggest you monitor input impedance of I pin. You may use an SMU or fixed current source to apply 10uA and monitor the voltage on it to calculate the input impedance.

Thanks

Sinan 

Hi Sang,

HMC346A is a GaAs based VVA and we would expect part to part variance on the gate pinch-off of series and shunt devices on the attenuator. Therefore, it is not meaningful to create a look-up table and apply to every part. 

The external I-V2 loops ensures the reference attenuator's input impedance to be 500ohm and main attenuator's input impedance to be 50ohms. If this is not set, mismatch losses would add on the attenuation, and this creates uncertainties on the end-to-end attenuation.

For experimenting purposes, if you still want to drive V1 and V2 separately, I suggest you monitor input impedance of I pin. You may use an SMU or fixed current source to apply 10uA and monitor the voltage on it to calculate the input impedance.

Thanks

Sinan 

Hi Sinan,

A follow up question on these replies please...

Can negative input voltage to the op amp (AD8675 in the datasheet) and V1 (as control voltage of the HMC346) be ganged and supplied from the same -ve voltage source? Such that when V1 changes (between -4 and 0V), this change is reflected at the op amp which is then reflected at V2, and then adjust the input impedance accordingly. Is that the idea behind the mode of operating this component?

The opamp needs to take the input from I pin, and reflect to V2 pin always. The voltage of I pin changes with respect to V1 voltage. You should follow the external opamp configuration exacly the same.

The opamp supply voltage should be at least -5V, so that it's rail can cover -5V.

Hi, how important is the power supply noise and ripple on the -5V rail coupled to the inverting input? is a filtered DC-DC converter with a few capacitors enough or should I use low-noise LDO like the LT3032 Series? Thx

Hi AGU,

I would recommend including an LDO because any ripple on the V1 or V2 will modulate the attenuation and reflect to the output as additional spurs.

Thanks