Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

results do not match with datasheet

Status
Not open for further replies.

prateek3790

Full Member level 2
Joined
Feb 8, 2016
Messages
131
Helped
1
Reputation
2
Reaction score
1
Trophy points
18
Activity points
1,190
i have a board in which we are using vmk 3803. now when we checked for the overall receiver matching it was very bad. so in order to check whether the lna is matched or not checked the s parameters of the lna . they were not matching with the datasheet. is this expected. what can be done for thi rx_lna_s22.jpegrx_lna_s21.jpegrx_lna_s12.jpegrx_lna_s11.jpegView attachment AV02-2920EN_DS_VMMK-3803_2014-02-100.pdfs.
 

Most obvious point is S21 dropping to about zero gain at 10 GHz compared to 18 dB with the reference design. Suggests unsuitable layout and/or bias T. Show your design.
 

can i not try with simply playing with the matching components that have been used in the developement board design, i am using the same matching components.
 

In my understanding, the development board doesn't have "matching components" just a bias network designed to have little effect in the 3 to 10 GHz range.

I don't know if the layout or the components make 10 GHz gain drop to zero in your design.
 

S22 looks much worse than S11 on top of the S21 mentioned by FvM. I would be very suspicious of something on the output (assuming your line is good 50ohms for 10GHz).
 

i had used a poor cable(connected to board sma from lna ) in my previous measurement, so just to make sure that nothing is wrong with the measurement going to check with rf cables.
 

i checked my results again this time using a better cable to sma, now both the port response do not match with the datasheet response.since i used the same board i was not expecting the port 1 performance to get effected. since there is no change at port 1 side.RX_LNA_S22.jpegRX_LNA_S21.jpegRX_LNA_S12.jpegRX_LNA_S11.jpeg
 

The new results are in some regard even worse, showing now a S21 drop to nearly zero dB around 5 GHz.

My post #2 question is still pending.

The "poor cable" point makes me also wonder if the VNA setup has been regularly calibrated.
 

we have a uwb chip which is connected to this lna (along with the referred matching as given in the datasheet ) and then it is given directly to the antenna. no other components.

so on your post 2 regarding the "drop in the gain" since no other components were used in this path, we were not expecting the performance to be this much bad in the recommended frequency range.
also since we have started testing the board now i do not have sufficient data, so i thought s parameter can give pretty good idea. now i am thinking to measure the received power after this lna just to make sure that the problem is in lna block (matching or layout, if it exist even after changing the components than can be more sure that it is happening because of the layout thing as you said). although this same as s para measuement

also pretty sure that the network analyzer was calibrated because repeated the experiment twice from start just to make sure and got the same results also tried also lifting the board and the cable of sna whether any loose connection but the response was still the same, so yeah pretty sure about the measurement.

one thing which i am still puzzled about is that why the s11 got changed as nothing was done on it used the same board just with different cable.(already checked for any short).
can you recommend any measurement or test to confirm the layout or matching issue to be sure why the performance is so poor.
 

we have a uwb chip which is connected to this lna (along with the referred matching as given in the datasheet ) and then it is given directly to the antenna. no other components.
The "UWB chip" (I presume the Johanson 3-10 GHz device mentioned in a previous thread) must be disconnected in LNA S-parameter measurements. In so far I don't understand the relation of this comment. As previously mentioned, there's no actual "matching network" in the LNA reference design, just a bias-T with hopefully small effect in the frequency range of interest.
 

yes it is disconnected sorry for making it confusing.
the matching network i am referring to is the one they use in their developement board. i am measuring the s parameters by keeping those elements(s parameters across those). also they should not impact that large bandwidth.

is there any other test which you can recommend
 

Both input and output return loss is very poor above 8GHz. I would not expect good but what bias voltage are you using? Looking at the data sheet, return loss goes to crap below 3V. Sure your line widths and connectors are correct?
 

measured both vpd and vdd, both are above 3.03v
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top