Hi, There is surely very much more to say, but I just wanted to explain why IIP3 alone is totally useless. It must always be given in relation to something. There is a lot of literature but the original question was a simple one (seemingly.)
I am sorry I failed to be clear enough. The point of my message was that although an attenuator does not help receiver performance (dynamic range) it will change IIP3. I mentioned this in an attempt to explain the uselessness of IIP3 when given as an isolated number without relation to sensitivity. The close range IP3 that you refer to is something that I think receives too much attention. Some receivers produce much worse IP3 at wider separations than they do at close. (Typically ill designed H-mode mixers) Designers may place the bad spots (produced by filters or amplifiers loaded by filters) at frequency separations where measurements are not done in standard procedures and that is cheating to me..... The dynamic range (IP3, blocking or reciprocal mixing) should always be specified at its worst points if there is a significant variation with frequency. Surely close range suffers from very bad IP3 performance in some designs, but a narrow filter in the first IF is not necessesarily the best solution, such filters can not be used together with good noise blankers. A very good dynamic range through the first IF is a better solution - but more difficult to design. Dynamic range is a big subject on which there is very much to say. IP3 in itself is a very simple concept and it is useful not only as a measure of receiver performance (when compared to a noise floor). IIP3 and OIP3 (input and output IP3) are also useful when building a chain of amplifiers and mixers, Both for tx and rx. The OIP3 of an amplifier should match the IIP3 of the next amplifier in a chain. By "match" I do not mean they should be equal, it is always good to make any of them higher, but it is waste of resources to make one much higher than the other, that would not improve IP3 of the amplifier chain. 73 Leif / SM5BSZ > IP3 is a little more complex than stated below. It is an important part of > the > overall dynamic range of a receiver. This is important if you are attempting > to > receive a weak signal when a strong local signal is in the receiver passband. > An > attenuator will not be helpful under such conditions. > > If you look at some of the newest transceiver designs, many are emphasizing > narrow band roofing filters in the first IF just to improve that IP3 number. > The > ARRL added close in IM measurements a while back for the same reason. > > > > Leif Asbrink wrote: > > Hello Frank, > > > >> I am trying to understand the IIP3 concept a bit better. I get the > >> general idea that it is a non-linearity measure of mixer that takes > >> one tone, applies a non linearity and it will interfere with another > >> one of different frequency. According to this graph > >> http://en.wikipedia.org/wiki/Image:Interceptpoint.png > >> am I supposed to look for a large positive dBm to get good performance? > > > > IIP3 on its own has no meaning. Put a 30 dB attenuator in front of > > the radio and enjoy a 30 dB better IIP3! > > > > IIP3 is the upper limit of the permissible signal range (One of them. > > Other limits may be lower.) It is meaningful only in relation to > > the lower limit of the permissible signal range which is the noise floor. > > The noise floor can be specified in many different ways. One convenient > > way is to look at IIP3-NF when comparing systems. > > > > (Or you may adjust the gain/attenuation of an imaginary pre-amplifier > > that you add in front of each radio to give it a standardized > > sensitivity and then compare IIP3 with receivers having the same > > noise floor.) > > > > 73 > > > > Leif / SM5BSZ > > > >> Looking at the below figures: > >> AS8347: +11.5dBm > >> softrock40: above 30dBm > >> would this mean that the H mode FET switch mixers are far superior to > >> analog mixers in the IIP3 spec? > >> > > > > > > Yahoo! Groups Links > > >
