Circuit Board is not working properly at 2.4GHz

[pawalepravin92]

Hello,

I am using a circuit board that is working fine at 1.2GHz, But whenever I tried to work at 2.4GHz frequency its giving "extraa 10to13dBm loss". Is it because of PCB material that I have use (As I am using FR4 material).
Please help to find the exact reason.

Thanks,
Pravin.

 

[BigBoss]

-What kind of circuit ??
-Has this circuit been designed/developed for 1.2GHz and you would it to run @ 2.4GHz ??

 

[cyberrat]

"Please help to find the exact reason."

How long is a piece of string? I'm not going to show you the string, tell you what sort of string it is but you guess exactly how long it is? roflmao.

If you gave more to go on, such as a pdf of the payout etc. then people would be able to help you more.

 

[pawalepravin92]

Hello sir,

I have designed my circuit for the frequency range up to 2700MHz. I think there is no problem with the circuit. Is there a PCB material issue?

Thanks & Regards,
Pravin.

 

[pawalepravin92]

Hello,

Please find below details

L1 TOP 0.035 Finish Copper
0.150 2*1080 preprag
L2 GND 0.035 Inner layer
1.200 1\1 1.2 mm Laminate
L3 PWR 0.035 Inner layer
0.150 2*1080 preprag
L4 BOTTOM 0.035 Finish Copper

Total Thickness 1.64 +/-10% mm

 

[pancho_hideboo]

Show us thickness of RF4 and thickness of conductor.

If thickness between signal and ground is thin, loss will be increased.

 

[volker@muehlhaus]

I have designed my circuit for the frequency range up to 2700MHz. I think there is no problem with the circuit. Is there a PCB material issue?

FR4 dielectric loss is relevant and it increases with frequency. But that is a continuous effect and you will see dielectric loss at 1GHz also. High Q circuits are more sensitive to these losses than simple lines, so maybe your 1.2GHz was less sensitive to losses anyway.

 

[pawalepravin92]

Hello sir,

Thank you for your valuable reply. I am operating my circuit for these two frequencies (1.2GHz and 2.4GHz) only. As you have mentioned loss is relevant and it increases with frequency, then how much loss will be there for 2.4GHz. I am getting -10dBm for the frequency 2.4GHz extraa loss than 1.2GHz frequency.
For frequency 1.2GHz - loss is -8dBm
For frequency 2.4GHz - loss is -18dBm


Thanks and Regards,
Pravin.

 

[volker@muehlhaus]

Loss in dBm?

If you need an answer, give accurate information and answer questions.
As BigBoss asked in #2: what circuit?

 

[pancho_hideboo]

L1 TOP 0.035 Finish Copper
0.150 2*1080 preprag
L2 GND 0.035 Inner layer

It seems thickness between L1 and L2 is 0.15mm.

Don't you use relative large footprints in L1 for SMD components where L2 exists right under footprints ?

Show me pattern of PCB.

 

[pawalepravin92]

Loss in dBm?

If you need an answer, give accurate information and answer questions.
As BigBoss asked in #2: what circuit?

I have designed a Mixer board.

For 1.2GHz --> Input given - Frequency-1.2GHz Level- 0dBm
Output - Frequency-200MHz Level- -8dBm
For 2.4GHz --> Input given - Frequency-2.4GHz Level- 0dBm
Output - Frequency-200MHz Level- -18dBm

 

[pawalepravin92]

Thank you for your valuable reply.

I have used this thickness for another board and found okay at 2.4GHz.

 

[FvM]

Your vague answers are making the discussion almost useless. I agree with pancho_hideboo that 0.15 mm FR4 thickness is far from a reasonable choose for 2.4 GHz. 10 dB loss increase suggests however additional effects like mismatching because frequency depend attenuation caused by skin effect and dielectric losses will be rather ~√f up to ~f, 3 to 6dB for factor two frequency. But without seeing the schematic and layout, further guessing makes no sense.

 

[pancho_hideboo]

I have designed a Mixer board.

(1) RFin=0dBm@1.2GHz, LOin=1.4GHz or 1.0GHz, IFout=-8dBm@200MHz
(2) RFin=0dBm@2.4GHz, LOin=2.6GHz or 2.2GHz, IFout=-18dBm@200MHz

Show us followings.

Frequency and Level for LOin
Return loss values for RFin and LOin

I don't think your issues are PCB loss.

Simply RF missmatch for both RFin and LOin.

Especially, lack of LOin Level.

 

[pawalepravin92]

Getting extra loss at the input port only (Frequency - 2.5GHz,),

Hello,

I have taken a Bare PCB and just mounted a SMA connector at input. When I give 1.2GHz with power level 0 dBm using Signal generator, I am getting -1.9 dBm output (seen on Spectrum Analyzer) . But when I tried the same thing at 2.5GHz, I am getting -9.5 dBm.

I have checked cables and SMA connectors at 1.2 GHz and 2.5 GHz, no problem with it.

 

[FvM]

Merged post with previous same-topic thread. Same problem as before, effectively no info about circuit and layout.

"Bare PCB with just one SMA" makes no sense. How do you connect generator and analyser?

 

[pawalepravin92]

Merged post with previous same-topic thread. Same problem as before, effectively no info about circuit and layout.

"Bare PCB with just one SMA" makes no sense. How do you connect generator and analyser?

Input from signal generator is given to SMA. And to see the output on Spectrum Analyzer, I am using the RF Probe.
As I am facing problem at input side of my circuit only, how can anyone blame on circuit (good thing that I mentioned many time - "PCB is working fine at 1.2 GHz").
Also, I have taken all readings of RF Probe as well.

 

[volker@muehlhaus]

Input from signal generator is given to SMA. And to see the output on Spectrum Analyzer, I am using the RF Probe.

I agree on FvM's doubts.

What "RF Probe"? How and where is it connected? Is it high impedance and if yes, how did you terminate the 50 Ohm on the PCB?

 

[pawalepravin92]

I agree on FvM's doubts.

What "RF Probe"? How and where is it connected? Is it high impedance and if yes, how did you terminate the 50 Ohm on the PCB?

RF Probe that I am using to get spectrum is- RF Probe – 3 GHz – P20A (Auburn Technology).
dBm power readings :
20dB Attenuation: When Probing
50 ohm RF Circuits

 

[volker@muehlhaus]

RF Probe that I am using to get spectrum is- RF Probe – 3 GHz – P20A (Auburn Technology).
dBm power readings :
20dB Attenuation: When Probing
50 ohm RF Circuits

This looks a high impedance probe.
Above you wrote: "I have taken a Bare PCB and just mounted a SMA connector at input. "
It is no surprise that you measure nonsense results if you probe a line that is not properly terminated.

~~

Also, I can't believe the frequency rating of the probe. The ground path connection is way too long at these frequencies.
Edit: Ok, that's even documented here