ADL5324 Datasheet

Hi Rog,

                the ADL5324 is not internally matched to 50 ohms. This impedance is 6 ohms or so and must be externally matched to 50 ohms. L2 & C2 work with the internal impedance of the ADL5324 to create a 50 ohm match for the ADL5324. The value for L2 is correct and has been tested.

Regards,

Jim B

Hi Jim,

Ok, yes I see the S21 is 5.6 ohms and needs to be matched to 50.

I am also trying to simulate this, but I cannot get it to match to 50.

Do you have any spice or ADS simulation files on this device, I could then see how I am going wrong.

Best regards,

Roger.

Hi Rog,

             On the web, we have an S-Parameter file you can download and use with ADS. I generally use a smith chart program to get a basic match at the center frequency of the band. I then use ideal components and tune function to get the desired gain and match across the band. I then substitute real components in ADS and adjust as needed to keep my gain and match. I then build and test. One point, I found at the higher freqs there is an error in the s2p file and my actual circuit is off by 250 MHz when tested. SO I offset the simulation freq and rematch in that case. I have never had to do more than 2 iterations to correct for the offset in frequency that may or may not occur.

Regards,

Jim B

Hi Rog,

             On the web, we have an S-Parameter file you can download and use with ADS. I generally use a smith chart program to get a basic match at the center frequency of the band. I then use ideal components and tune function to get the desired gain and match across the band. I then substitute real components in ADS and adjust as needed to keep my gain and match. I then build and test. One point, I found at the higher freqs there is an error in the s2p file and my actual circuit is off by 250 MHz when tested. SO I offset the simulation freq and rematch in that case. I have never had to do more than 2 iterations to correct for the offset in frequency that may or may not occur.

Regards,

Jim B

Hi Jim,

Yes I have the S parameters file, am using it in ADS.

I have tried placing in substitution vales for the reactive components - without any success in matching it.

I still believe the inductor is too high a value. Would you be able to check you results, for the 3500-3700 MHz band.

Best regards,

Roger.

Hi Rog,

             I have limited data for the ADL5324 at 3500-3700 MHz. Attached is a scalar gain plot. Gain in the band is ~ 12 dB.

Regards,

Jim B

PDF

Hi Jim,

Thanks for that.

This is the gain I was hoping for, however I do not get close to that.

What is the matching network associated with that gain?

Cheers,

Rog.

Hi Roger,

                   The network is the one given in the data sheet. Are you using the exact components specified for C1, C2, C3 and L2,(Labeled as R2 in the table and on the board)

Regards,

Jim B

PDF

Hi Jim,

Yes, exactly as per that suggested layout.

The 4.7nH inductor is in series with the output (only an ac coupling cap after it)

It has a reactance of 103 ohms at 3.5 GHz.

But the output impedance is designed to be 50 ohms.

This is I believe is an issue.

Best regards

Rog.

Hi Rog,

Is C2, the 0.75 pF cap placed? it is located 125 mils from the output pin. It works in conjunction with the inductor to match the output.

Regards,

Jim B

Hi Jim,

Yes. Exactly to the datasheet.

Does this act as a T network?

Best regards,

Roger.

HI Rog,

                 Yes, this would act like T network.

Regards,

Jim B

Hi Jim,

I can get the S11 and S22 to dip at 3.5 and peak power at 3.5 but the shunt capacitor on the output is 2.0 pF - not 0.75... This is using ADS S parameters simulations.

I am going to try out a range of cap values on my board and see what gives the best output.

On a related not but slightly different, I am trying to model load pull.

To do this I am using ADS harmonic balance.

I therefore require a model for the ADL5324 that is similar to an actual RF transistor ie: not just it's S params.

Do you have this?

Best regards,

Roger.

PS: am getting there slowly...

Hi Rog,

                Analog devices does not offer a transistor based models for RF amplifiers. There may be some error between ADS and built matching circuits. I have noticed this at some of the higher frequencies. So that may be why you have a different value inductor.  My trick when that happens, is to shift the frequency by the delta between the ADS freq and the freq the part is actually tuned to in real life. After 2nd simulation with the offset, I find the tuning is spot on.

Regards,

Jim B