Colon
Member level 3
When I left work today 4 of us were convinced that the Smith Chart measurement on our Agilent E5061B (100k-3GHz VNA) was telling porkies.
Here's the low down. First we were trying to find out why the input match of our new 1-3GHz amp was poor. Having slowly removed all possible candidates we ended up with just a series 10pF 0402 capacitor and an 0603 51R resistor to ground. At this point out smith Chart impedance was not reading 51 ohms of real impedance! It was actually reading around 33.
This got even more hard to believe when we lowered the frequency to 30MHz. At this point our measured impedance was about 33-1666j ohms. We checked the resistor, it was 51R with a DMM. We placed it directly across the SMA input to the board and it correctly measured it as approx 50R. But only at lower f, above 1GHz it got lower, towards the 30's, hmm...
So basically, our VNA thinks our 51R resistor is actually much lower, around 33R. At around 1.5GHz, the SRF of our capacitor, it's reactance is zero, but it still thinks our resistor is 33R.
how can this be!!!! Just to reiterate, thi is an SMA input, a section of track a few mm long, a series 10pF cap, then 51R to ground, showing as 33-1666j ohms at 30MHz. what the hell is going on? 3 very much more experienced RF engineers than myself could not explain this. maybe one of them will help figure it ou tomorrow, to me, this makes no sense. We might have to contact Agilent if we can't work it out seen as though we base a lot of our matching on these readings! Oh, the VNA in question is not broken, we tried it on two others the same model.
any help/ideas greatly appreciated!
Cheers
James
Here's the low down. First we were trying to find out why the input match of our new 1-3GHz amp was poor. Having slowly removed all possible candidates we ended up with just a series 10pF 0402 capacitor and an 0603 51R resistor to ground. At this point out smith Chart impedance was not reading 51 ohms of real impedance! It was actually reading around 33.
This got even more hard to believe when we lowered the frequency to 30MHz. At this point our measured impedance was about 33-1666j ohms. We checked the resistor, it was 51R with a DMM. We placed it directly across the SMA input to the board and it correctly measured it as approx 50R. But only at lower f, above 1GHz it got lower, towards the 30's, hmm...
So basically, our VNA thinks our 51R resistor is actually much lower, around 33R. At around 1.5GHz, the SRF of our capacitor, it's reactance is zero, but it still thinks our resistor is 33R.
how can this be!!!! Just to reiterate, thi is an SMA input, a section of track a few mm long, a series 10pF cap, then 51R to ground, showing as 33-1666j ohms at 30MHz. what the hell is going on? 3 very much more experienced RF engineers than myself could not explain this. maybe one of them will help figure it ou tomorrow, to me, this makes no sense. We might have to contact Agilent if we can't work it out seen as though we base a lot of our matching on these readings! Oh, the VNA in question is not broken, we tried it on two others the same model.
any help/ideas greatly appreciated!
Cheers
James
