Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Newbie: mixers, IIP3 and it's measurement via Sginal Generator and Spectrum Analyzer

Status
Not open for further replies.

Lucky Ali

New Member
hello all,
for the moment i am measuring some characteristics of a mixer with RF, LO & IF ports. I have read some article about taking measurements but not every thing is so clear enough. I have few questions.
1. very basic question. since we give LO and IF to a mixer and measure RF at output (up-conversion). when we say input power vs output power. what does input refers to, is it only the IF, or only the LO, or both of them (IF and LO) are considered as input signals.

2. what is 1dB compression point. and how do we measure it, (i am working in a lab with signal generator and spectrum analyzer so one can explain practically instead of some simulations, or what ever is easy)

3. IIP3 or 3rd order input intercept point... I have to calculate this thing. and to present it graphically what are the components of the graph.. and same question. LO and IF should both be considered as input signal. if yes then do we vary the power level of input signal at the same time, I mean the starting point of both signal should be at same input level or should be both at same output level.
like if LO= IF= 0dBm and then change both at the same time to LO= IF = 3 dBm or how, I don't get it.

many thanks in advance
best regards
Ali
 
Last edited:
Input frequencies and products should all be in the mixer output bandpass.

Third order will create two side products at the spacing of the two input tones on either side. You may also see 5th order products when you get high enough input levels. One db compression is not accurate to make IP3 projection.

The 3rd order produces will rise 3 db for each 1 db increase of the the two tone as displayed on spectrum analyser. Lets say you input -40 dbm for each of the two tones and observe the 3rd order produces to be each be -30 db down on the I.F. output. The IP3 will be -40 dbm + (1/2 the relative db down of the two products) or -40 dbm + 30db/2 = -40 dbm +15 db = -25 dbm.

You have to make sure there is enough isolation between the two generators so they don't create IM products in their ALC output circuitry. (automatic level control).

You can just look at the one of the products in a narrow band receiver but any RF filter and input matching circuit may effect the input levels of the two test tones and the different impedance at offset freq may effect mixer IP3 result.
 
Last edited:
Last edited:
1) i/p refers to power at input, o/p power at output. For down-conversion i/p at RF port, o/p at IF port. For up-conversion i/p at IF port, o/p at RF port. The difference between i/p and o/p is called conversion loss or gain; depending on mixer technology. LO is always an input signal, but it should not vary (i.e. power level should be kept constant); that is to say the LO is only use to translate the IF into RF (for up-conversion), or RF into IF (for down-conversion).

2) 1dB compression - look up this term for amplifiers. It is the point at which the gain deviates by 1dB from the expected gain. That is to say, the gain should be constant at low input power levels; gain at -40dBm input level should be the same as gain at -39dBm, -38dBm, etc. Keep increasing the input signal level until the measure gain decreases by 1dB. That input level is the input 1dB compression point, that output level is the output 1dB compression point. For linear systems you want to operate below this level (considerably lower) to avoid distortion.

3) As for IIP3 or 3rd order input intercept point, it is a measure of non-linearity. Add two tones at the input of a non-linear device at a power which is significantly lower than the 1dB compression point. Measure the unwanted tones at the output and compare their levels to the wanted tones at the same output. Suggest you read Introduction to Radio Frequency Design by Wes Hayward, or RF Design Guide from Pete Vizmuller.
 
thanks every one for your nice comments, @ Mikebits the links are very useful but too much reading from different article with different way of explanation increase my confusion
lets say i am doing upconversion so IF= LO +/- RF, I choose RF = 100 Mhz with power varied from -6 dbm to -12 dBm and LO= 16 Ghz (12 dBm fixed)
the ouput measured at spectrum analyzer are

for
RF= -12 dBm ==> (LO - RF)= -61.40 dBm (LO - 2*RF)= -38.38 dBm
RF= -9 dbm ==> (LO - RF)= -56.90 dBm (LO - 2*RF)= -35.38 dBm
RF= -6 dBm ==> (LO - RF)= -53.32 dBm (LO - 2*RF)= -32.00 dBm

how can i Calculate IP3 for this mixer from these values...
 
thanks every one for your nice comments, @ Mikebits the links are very useful but too much reading from different article with different way of explanation increase my confusion
lets say i am doing upconversion so IF= LO +/- RF, I choose RF = 100 Mhz with power varied from -6 dbm to -12 dBm and LO= 16 Ghz (12 dBm fixed)
the ouput measured at spectrum analyzer are

for
RF= -12 dBm ==> (LO - RF)= -61.40 dBm (LO - 2*RF)= -38.38 dBm
RF= -9 dbm ==> (LO - RF)= -56.90 dBm (LO - 2*RF)= -35.38 dBm
RF= -6 dBm ==> (LO - RF)= -53.32 dBm (LO - 2*RF)= -32.00 dBm

how can i Calculate IP3 for this mixer from these values...

You can't, you have to use two signal generators and setup equipments as I linked above. You should get a display as shown in image.

Side note: From your numbers, your insertion loss is huge, you have a problem there.
View attachment 67326
 
Last edited:
so that means at RF I must input a multi tone signal... Is there any other way to make the setup, because I am not having any equipments other then signal generator and spectrum anlyzer
 
I don't know of any other way. What is this for, what mixer are you using? You may be able to infer IIP3 based on design, using component data sheets.
 
hey lucky ali,

I think there is misunderstanding of the input vs output spectrum..... you see the sequence follows like this
IF input(IF_IN)--->amp,mixer,bpf--->transmitter output(TX_OUT)--->receiver input(RX_IN)--->IF output(IF_OUT)

So your first question of what is input vs output is IF_IN vs IF_OUT, which does not include the local oscillator input. Whereas when you use LO into consideration, the plot would be TX_OUT vs IF_OUT

you can see the **broken link removed**
 
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top