How to cascade BGA2869 amps

[daverajnauth]

I need help with BGA2869 MMIC amps.  I'm trying to cascade two amps for a 7 Mhz HF application to get more gain and running into a brick wall. Each amp works great on its own.  Tested with input of -107dbm, -73 dbm and -33 dbm from a 50R source. I get at least 28dB gain for each amp connected to 5V supply.  Each amp input/output is decoupled using 100nf caps.  The current draw to each amp is around 30mA.

I've created a PCB following power supply, ground plane and decoupling recommendations.  However, when I interconnect amps together (using  100nf coupling capacitor and a straight trace between input and output of respective amps) I run into problems.  On this PCB if I remove and bypass one amp and test the other I get the results noted above.

First I don't get the expected gain (i.e. 28db + 28 db). I measure less than 20db of gain for both amps.  Second, the current draw for the second amp drops to almost half of the spec (about 15mA) and the first amp is consuming about 30 mA.  Third, the noise floor comes up substantially more that for one single amp.  Would expect this if both amps were operating but based on current draw the second amp does not seem to be working well.  Again this ONLY happens when the amps are connected together. 

I have a 7805 regulator that is rated at 1 amp of current and i have various caps (100uf, 1uf, and .1uf) to control power supply noise from 7805 and power supply.  Also have a 470pf cap 7.4 mm from Vcc pin as recommended on data sheet.

Is there anything special that is needed to cascade these amps?

First time using MMIC amps and I'm lost.   Any help would be greatly appreciated.

[heikoleutbecher]

If you want to get highest gain of the cascade, you should run the first mmic with much lower current than the second one.

Maybe that helps.

[RA_RA]

Hi Dave,

Here the net gain of two devices is more than 50dB which may cause unwanted oscillations in the circuit.

Try by introducing a 3dB Pi-attenuator pad in between these two MMICs.

Pi-attenuator is implemented with the help of resistors. Lots of information is available on internet regarding implementation of pi-attenuator and required resistor values.

Yes 3dB attenuator will drop your overall PCB gain by 3dB but system will be become stable.

Regards,

Rajan

[daverajnauth]

Thanks Pavel and Rajan for the reply. 

One other observation is that, the current consumed by the second module drops significantly.  So this is consistent with Pavel's comment.

I have ordered some beads and in line attenuators.  I also have also ordered a low pass filter.   Hopefully one of these works.  I guess this is why many discrete amp designs has a bead at the base or gate of the semiconductor.

Thanks,

Dave

[LPP]

BGA2869 is a wideband amplifier. It can be used up to 2.2GHz. So, it can amplify or generate at very high frequencies.

BGA2869 quiescent current (if no or low input power) is about 24mA.  If you see different current or if your gain is lower than expected, the most probable reason is related to high frequency oscillation.

Please refer to layout in datasheet Figure 4.

- it is important to have enough ground vias underneath and around the MMICs ground pins. Use multiple vias as close as possible to the MMIC. This lowers the inductance to the ground plane.

- spread multiple vias all over the ground plane. It prevents board related resonances.

- use ferrite beads to feed the stages of a cascaded amplifier.

Have a great day,
Pavel
NXP TIC

-----------------------------------------------------------------------------------------------------------------------
Note: If this post answers your question, please click the Correct Answer button. Thank you!
-----------------------------------------------------------------------------------------------------------------------