The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> IIP3 @ 10MHz offset for WCDMA
https://designers-guide.org/forum/YaBB.pl?num=1123785193

Message started by boshiouke on Aug 11^th, 2005, 11:33am

Title: IIP3 @ 10MHz offset for WCDMA
Post by boshiouke on Aug 11^th, 2005, 11:33am

The inband IIP3 for WCDMA is specified @ 10MHz offset. So I ran a 2-tone test w/ following parameters,

RF1=2150MHz
RF2=2160MHz
LO=2140MHz

First of all, should I have used the 2 tones @ 10/20.5MHz so that IM3(0.5MHz) falls within the BW(5MHz)?

Secondly, due the LPF @ mixer output w/ cutoff of 8MHz, the output @ 10MHz is much bigger than @ 20MHz. In this case, how do I calculate IIP3? Does "Pin+1/2 x (delta between fundamental and IM3)" still apply? Which 2 frequencies should I take the delta from? Assuming the Pin is in linear region of the mixer.

Thank you so much.

Title: Re: IIP3 @ 10MHz offset for WCDMA
Post by pancho_hideboo on Aug 13^th, 2005, 8:22am

Hi.

>The inband IIP3 for WCDMA is specified @ 10MHz offset.

IIP3 is component's specification not WCDMA specification.
Instead of IIP3, IM sensitivity requirement is specified in WCDMA,
Interferences are defined like followings.
U1 ; Non-Modulated Interference +10MHz apart from Desired Signal
U2 ; Modulated Interference +20MHz apart from Desired Signal

For component level design, you have to translate IM sensitivity to IIP3.
If you have Filter limiting IF bandwidth, you must compensate this effect
for tranlating IM sensitivity to IIP3.

> So I ran a 2-tone test w/ following parameters,
>RF1=2150MHz
>RF2=2160MHz
>LO=2140MHz
>First of all, should I have used the 2 tones @ 10/20.5MHz
> so that IM3(0.5MHz) falls within the BW(5MHz)?

I can't understand your question.
What does it mean by "2 tones @ 10/20.5MHz" ?
You mean RF1=LO+10MHz, RF2=LO+20.5MHz, IM3=0.5MHz and 31MHz ?

>Secondly, due the LPF @ mixer output w/ cutoff of 8MHz,
> the output @ 10MHz is much bigger than @ 20MHz.

I can't understand your sentence. ???

With LPF of fc=8MHz, RF1=LO+10MHz, RF2=LO+20.5MHz,
resulting frequencies at IF port are IF1=10MHz, IF2=20.5MHz, IM3=0.5MHz.
You mean IF1 and IF2 are out of pass band ? ???

> In this case, how do I calculate IIP3?
> Does "Pin+1/2 x (delta between fundamental and IM3)" still apply?
> Which 2 frequencies should I take the delta from?
> Assuming the Pin is in linear region of the mixer.

(1) Determine substitute RF1, RF2 and IIP3 where all IF1, IF2 and IM3 are with in in-band.
(2) Apply loss due to LPF for IF1 and IF2.

There are other methods.

Which methods you use is your own task. :-/ :-/ :-/

Title: Re: IIP3 @ 10MHz offset for WCDMA
Post by Frank Ke on Aug 13^th, 2005, 3:09pm

Thanks for the reply. I'm sorry that I wasn't clear on my description. I was trying to simulate the IP3 for the Rx Mixer, with a 2-tone test. Since the inband IP3 for WCDMA is specified @ 10MHz offset, the 2 tones will fall out of the signal bandwidth given the fact that WCDMA's BW is only 5MHz. Therefore, I am not sure about my selection of the 2 tones. Should I use LO+10MHz and LO+20.5MHz so that one of the IM3(0.5MHz) will fall within the BW? There are papers that claimed they used LO+10MHz and LO+20MHz. I can't figure out how that would make sense.

Secondly, since the cutoff at mixer output is only 8MHz, natually IF1(10MHz) and IF2(20.5MHz) will have different amplitude. So the question is, does the equation "IIP3=Pin+0.5(Pout-IM3)" still apply if IF1 and IF2 are different? If yes, should I use IF1 or IF2 in this case?

Lastly, does anyone know that in a WCDMA's datasheet when they specify IIP3, do they specify it at a particular RF input level?

Thank you so much.

Title: Re: IIP3 @ 10MHz offset for WCDMA
Post by jefkat on Aug 15^th, 2005, 3:58am

sniff something fishy here .....why are u applying tones at the LO port ...shouldn't they be at RF port ...???

If you are interested in just the IP3 of mixer itself
You can do it with one tone as well .~(1dB compression+10dB).

And the equation is still valid ...Just consider the filter and mixer as two cascaded blocks... If the filter is attenuating it is just as good ....
hope it helps.
shafkat.