I am facing some trouble while using the RF multiplier ADL5391, where our aim is to use it as a squarer (same input over Xp, Yp).
Main issue is the fact that we observe reflections over the input Yp.
By following the datasheet we placed R2=56.2 ohm (over Yp), but still we cannot get rid of those relfections.
The only modification that we haven't done is to remove R9, R10 since the are connected to green test loop (YP_DC, YM_DC) and we thought they were disconnected....
Someone has experienced the same behaviour? Do you think we should remove R9 and R10?
Thanks for your cooperation :-)
If you still have R1 (56.2 Ohms) installed on the board and you have YP and XP shorted together, the equivalent resistance looking into that node is 28 Ohms. If you are driving from a 50 Ohm source then your reflection coefficient will be about -.282, your return loss is about -11 dB, and your SWR will be about 0.56.
Why don't you remove R2 and see if that makes things any better.
Hope this helps,
We do have 2 ADL5391, one with R2 open (evaluation board untouched) the other one with R2=56.2 ohm. In both cases we observe reflections from Yp (while Xp seems always matched).
Attached you'll find the setup we used.
Do you have any suggestions?
How are you determining there are reflections? And how big are the reflections?
Are you not getting the expected levels at the output and thus concluding the there are reflections on the input (S11 is poor)?
Attached you can see the setup I used in order to determine the reflections (input pure sinusoid with power always 0 dBm from signal generator).
- First experiment I made, I used a 50 ohm termination instead of the Xp input of the ADL5391. So one branch of power divider into the ESA and one branch terminated in order to avoid reflections. In this configuration power of the harmonic into the ESA was -6.2 dBm confirming the IL of the power divider and confirming that no reflections happened.
- Second experiment I used one branch of the power divider into the Xp input of the ADL5391. At that time looking at the ESA trace still the power observed was -6.2 dBm or close to it, confirming that no reflections took place.
- Third experiment I used one branch into the Yp input. At that time the ESA trace was characterized by a peak power ranging from -3dBm to -11 dBm. This peak power is variable according to the input frequency (instead in the previous experiments was more or less the same).
Thanks for you feedback.
For the third test you mentioned on the Y input, did you have the 56 ohm resistor installed across the differential Y input? I'm just wondering it the root of the problem is in the Y input.