PA design with custom IC

[Lathas]

I am trying to get a PA working (using GALI24+ chip) with the recommended circuit given in the datasheet: https://www.minicircuits.com/pdfs/GALI-24+.pdf

I have used Cblock, Cbypass of 10nF each, RFC 150uH, Rbias of 47ohm and Vd =9.4V. But I am not getting any positive gain. What could be the problem here? How can I start debugging the circuit?

 

[FvM]

Did you check the DC operation point (voltage drop across the 50 ohm resistor)?

How do you measure the gain?

 

[Lathas]

didn't check the voltages, I will check it out..

I have given a signal from an oscillator as input and measured output power with spectrum analyzer.

 

[BigBoss]

Pay attention the input of this amplifier.If you apply a high RF power, the amplifier may be easily burned.( Emitter followers cannot resist to high input levels.)
Instead, increase the RF level incremently and observe the SA.

 

[volker@muehlhaus]

RFC 150uH

That is a large inductance value. Have you checked the inductor's self resonance frequency? What is your operating frequency?

Same for the 10nF series caps: have you checked the self resonance frequency?

 

[vfone]

The simplest way to find if an MMIC is still alive, measure the DC voltage referenced to the ground on PIN 1, and the voltage should be somewhere between 1.5V and 2V.

 

[Lathas]

Ok, I applied a 9.4V as Vd, and got 5.01V at Pin3. I tried to add another 50ohm parallel to reduce the drop across R, but still the Pin3 voltage could go upto 5.15V only. The device operating voltage is 5.4 to 6.2V, means the device is not having the right Vd. If I try to increase the Vd, the current drawn increases, but not the voltage at pin 3. What could be the problem?

 

[ALERTLINKS]

The device operating voltage is 5.4 to 6.2V, I applied 9.4V

Vd should be within operating voltage range.

 

[Lathas]

I used the same value of resistor as given in the recommended ckt in datasheet and was able to get the amplifier working, but it is giving positive value of S11 at certain band, what could be the issue? Measurement was done on VNA after calibration.

 

[FvM]

Did you connect pin2 and 4 directly to a continuous ground plane?

 

[Lathas]

Yes, I have the same pad(copper trace) connecting pin2 and pin4, which is connected to the gnd underneath through via hole.

 

[FvM]

Yes, I have the same pad(copper trace) connecting pin2 and pin4, which is connected to the gnd underneath through via hole.

um. I would rather expect 3 or 4 vias than just one. Ground inductance can definitely cause negative input impedance (S11 outside the unit circle in smith chart).

 

[Lathas]

Thanks FvM! I have fabricated another board with 3 via holes and now S11 at the said frequency is -6.9dB. I was expecting a value lower than -10dB. Anyway, the new issue is that I am getting a S11 phase zero cross over every 100MHz, and the s2p on datasheet shows phase cross over only at 2GHz interval. Would it be the input and output transmission lines that I have added that generates those multiple zero phase cross overs? How can I fix this?

 

[Lathas]

Here is the chart https://obrazki.elektroda.pl/4661424500_1439142028.jpg Any suggestions on what could be the cause for the sub-phase variations?

 

[D.A.(Tony)Stewart]

Get rid of the 100uH RFC . you're probably oscillating at 160MHz due to SRF and causing 360 deg cycles of phase shift of interference every 160MHz on the sweep. THis may allow pin 3 to rise

 

[Lathas]

I already replaced the 150uH with 43nH inductor and the chart is from the 43nH setup.

 

[Lathas]

After changing the dc block capacitors at input and output to a lower value of 1pF, S11/S22 phase variation is settled, but S21 phase variation is still an issue. Any suggestions to fix the S21 phase?

- - - Updated - - -

I checked the amp ckt to a spectrum analyser and couldn't find any oscillations.