Talk about RF layout experiences about mobile phone.

[loucy]

Hi::

Is (are) any EDA software that's particular useful/necessary for what you are doing?

 

[xinxin]

Which tools to simulate the line radiation or any EMI problem?

 

[mckinson]

hi,activewei:
switching spectrum transmitter output spectrum due to PA switching transients

 

[xinxin]

I think the switching spectrum fail was almost due to the T/R switch , you could tune the capacitor around the switch.

 

[K'unI]

due to switch fail,
usually is PA Ramping up/down not tuning balance(smooth)
not T/R switch, beacuse the T/R switch is a passive component,
so it not affected transmission spectrum.
or
maybe is power supply trace or decoupe capacitor placement issue.

 

[wirelesscomm]

I have a problem in TE design, which is not mibile used,
but so familiar with the MP

I connect the vco and the modulatar with a buffer amp
and it is direct upconverter.

I find the output of the modulator is oK, even in the state of output power is 10dbm.
the phase error is less 2 degree, evm is 4%,
but when i use the pa to make the output power up to 40 dbm
the phase error is degraded , and the ACP is worest!

can someone tell me how to do with it ?

 

[activewei]

mckinson: Power ramp is due to the slow pick up time of Power Amplifier as the data is in burst mode. But I do not have experience in the question addressed by you.

Could you try to test your circuit at lower power where the PA is not operate near P1dB? I believe you will see some improvement. DO let me know for any new discoveries

 

[activewei]

TO wirelesscomm: 1 of the problems of poor ACPR is due to non-linear characterisitic of the PA. There is always trade-off between efficiency of PA and the performance.

The other reason is the PA degrades the phase noise of the LO.

Test at lower input power to the PA, ACPR/phase error will be improved. I am sure! :wink:

 

[jackyqiu]

1. Rx and Tx is separated to different layer
2. use microstrip
3. use shielding house in Rx and TX circuit
4. Crystal line is separated by the ground plane, don't infect RF signal
5. PLL circiut is layouted to same layer with Rx circuit

 

[mckinson]

I've made some matching from PA to FEM. I found the swiching spectrum improve great.
I think the PA output no-matching cause the switching spectrum fialed.

 

[activewei]

mckinson: what is the P1dB, gain and VSWR of the PA before and after performing matching?

 

[iambbcat]

1,Carefully placement ,keep RX far from TX and PA
2,use microstrip trace and calculate its impedance.
3,RX trace as short as possible.
4,reduce capacitance load of small signal
5,carefully select and place decouple CAP
6,mount a shield

 

[mckinson]

Befor matching ,the VSWR is about 1.75.After matching , about 1.16.But ,I found other problem . When transmiting , the current is more 10mA .And the Gsm harmonic (about 1.8G) is very big . I think the harmonic couple to the DCS track , because the matching I made is for the DCS track .
Activewei , do you think so ?

 

[activewei]

mckinson:
1) You performed matching for DCS path, and you found that the harmonic for GSM path is greater? Are you operating both amplifiers(GSM/DCS) at the same time?
->If yes, it is common phenomenon to have RF coupling as you are trasmitting two >1W RF power.

2) The current is 10 mA higher when you perform matching-> this is normal condition as the output power is slightly higher now. (Am I right?)

3) How do your perform matching for your PA? Load pull? or just paste a copper tape at the output and solder when max power is measured?

4)is the VSWR measured under small signal condition? If yes, most likely you will not see the additional 10 mA drawn by the amplifier.

 

[dandan]

i think amplifier is driven to saturation ~

 

[pewang]

Be aware with your match network banwidth,may be it also match the 2nd harmonic well.

 

[Wilson_yu_chen]

P1dB is basically not affected by output PA matching, it's a linearity parameter instead of changeable index except bias supply changes.

SWR/gain can be improved after your matching job done.

With Regards,

 

[Wilson_yu_chen]

Poor layout of supply pushing effect on VCO would degrde phase error sand
even frequency error. Lower VCO bias or attenuation adding before PA would
reducing load pulling effect on VCO.

With Regards,

 

[foxpro]

hi i need your hand man

i have a broject about TETRA
can any body help me??
i need a simulation and E-Books about TETRA . if you want Give me your hand send it :?: to me on :
zetfmss@hotmail.com