Broadband matching for a broadband input in MWO or ADS?

Hello,

For broadband amplifiers, the Zoptimum is typically designed in a single frequency point and then match this single point say to a broadband output load...

However, since the PA Zopt response may differ in some other frequencies, it means the traditional way of matching a single frequency point at the input (normally at the midpoint of the broadband freq. range) to a broadband input is not too exact...

In MWO, you can determine the optimum load for a single freq. point....So if I try to get around 10 different frequency points ( 10 different Zoptimum ) and simulataneously broadband match this, how do you do it in MWO or ADS?

I think if Zopt points are close each other, it can considerably be easier.But if these points are spreaded, it must be tough..

Hello BigBoss,

How do you normally handle this broadband matching? Do you optimize the load at the midband frequency and then use matching softwares to get a certain broadband response?

The easiest way to increase the bandwidth of the matching circuit (or passive filters as well) is to add more poles. It will cost more components, but get good results. In this way, due to increasing of the loaded-Q, the insertion loss of the circuit is increasing insignificantly, of course for reasonable number of poles.

hello vfone,
thanks for your response, but practically, the significance of adding poles (LC) in broadband matching has an upper limit, say a 3-section LC low pass versus a 4-section or 5 section...the response improvement is getting insignificant as you increase further..

but my question is that since you determined the amplifier's optimum impedance at just a single point of frequency, normally at the center (example a broadband range of 1GHz to 1.5GHz, my Fcenter=1.25GHz), then you do broadband matching using this Fcenter...however, the amplifier Zoptimum would vary within the broadband range, right? so how would you account these Zoptimum variations in your broadband matching?

I have designed a X-band LNA. 1st optimize at the center freq point, then optimize with the desired bandwidth, al last need manually adjust.
So I got about 10% bandwidth.
So you can rely on the computer software and manually adjust both.
More pole of PA match is the same as the filter. If you have a LC filter, you only have one freq point pole, and the bandwidth is narrow. But more pole, there will be more resonant, so the bandwidth is increased. Over the whole pass band, the VSWR and IL are all good.

hi tony,

yes you are right, but the broadband match is only referenced to a single frequency; the Zoptimum you found in the center frequency. so you are saying that the manual adjustment will take care of the Zopts at the other frequency range? have you experienced when you see your output response with several dips and peaks within the broadband range? how do you normally resolve this?