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.

Talk about RF layout experiences about mobile phone.

Status
Not open for further replies.

xinxin

Full Member level 3
Joined
Oct 8, 2003
Messages
184
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Activity points
1,238
Could u talk about the RF layout experiences about mobile phone , especially RF line , Crystal line , PA layout , switch and VCO layout.
 

rfic

Full Member level 2
Joined
Dec 14, 2003
Messages
138
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,298
Activity points
784
hi, Could you tell me what's you doing?
 

jupitorcuu

Member level 5
Joined
Apr 24, 2003
Messages
90
Helped
3
Reputation
6
Reaction score
1
Trophy points
1,288
Location
chinaBJ
Activity points
823
PA you should consider more about frequency for output match and DC block.
VCO should have good power supply.
Other is 50Ohm required
 

egemini

Full Member level 2
Joined
Apr 2, 2002
Messages
132
Helped
8
Reputation
16
Reaction score
5
Trophy points
1,298
Location
GOC
Activity points
990
Hello
could you please share your expereince in VCO and power amplifier layout in mobile phones.
Any tips for LO phase noise improvement.
Some body designing TETRA radios
:)
 

jupitorcuu

Member level 5
Joined
Apr 24, 2003
Messages
90
Helped
3
Reputation
6
Reaction score
1
Trophy points
1,288
Location
chinaBJ
Activity points
823
I think for VCO is less voltage fluctuation, so need some lowpass filter( LC or RC) to remove high frequency interference
 

activewei

Full Member level 2
Joined
Sep 19, 2002
Messages
125
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Location
UK
Activity points
1,433
Egemini:
1.If you want to achieve good phase noise, I would not recommend you to use VCO unless you are designing frequency synthesizer. Generally, VCO phase noise is not as good as Dielectric Resonator Oscillator (DRO). If you need fixed LO source, DRO is a better option.
2. Consider pulling and pushing effect of the VCO. For high frequency modulator, there is impact if you connect your frequency synthesizer directly to your modulator, the EVM and phase error will be significantly degraded.
3. For PA, minimize the loss to the biasing circuitry. There are many more to consider in PA design.
 

egemini

Full Member level 2
Joined
Apr 2, 2002
Messages
132
Helped
8
Reputation
16
Reaction score
5
Trophy points
1,298
Location
GOC
Activity points
990
activewei:
We are designing a fractional synthesizer not VCO alone. More over our frequency is in UHF range, so I dont think DRO will do.

regarding 2nd point
We are connecting synthesizer output directly to modulator,
Do you have any alternatives

In PA, the biasing circuitary is being taken care of
Is there any issues regarding grounding.
Since I am designing an 8 layer PCB for mobile phone for the first time, your expereinces in this field would be helpful

thank you
 

Olxx

Member level 1
Joined
Nov 29, 2002
Messages
35
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
286
2activewei

activewei said:
Egemini:
2. Consider pulling and pushing effect of the VCO. For high frequency modulator, there is impact if you connect your frequency synthesizer directly to your modulator, the EVM and phase error will be significantly degraded.

May I ask you to clarify your statement about direct connection between modulator and synthesizer? Why I should expact any degradation in this case?
 

bbscom

Newbie level 3
Joined
Apr 9, 2003
Messages
4
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
12
Does anyone can provide a refrence design of a cdma mobile phone?
 

mckinson

Full Member level 2
Joined
Sep 3, 2003
Messages
126
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,298
Activity points
1,034
For PA supply ,the track must be large and short ,the PA gnd must be directly to the main GND layer inclue decoupling capacitor
 

mckinson

Full Member level 2
Joined
Sep 3, 2003
Messages
126
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,298
Activity points
1,034
the TX track must be far away from RX track .the RF track must be far away from the control tracks .
 

eecly

Member level 2
Joined
Feb 25, 2004
Messages
47
Helped
2
Reputation
4
Reaction score
0
Trophy points
1,286
Activity points
356
Hi,

1. Isolation between PA and VCO (Prevent VCO pulling)
2. RF decoupling in every connection pin
3. Good match between PA and tranceiver
4. Good shielding
5. Steady power supply

eecly
 

activewei

Full Member level 2
Joined
Sep 19, 2002
Messages
125
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Location
UK
Activity points
1,433
egemini: Sorry for the late reply as I just found your questions.

I am currently working on this and I can share it with you. It is really not recommended to connect synthesizer output directly to modulator. To minimize the pulling effect, add attenuator (match to 50 ohm) as close as possible to the VCO output. For isolation, you can add buffer amplifier after attenuator to minimize the effect. My recent investigation shows that isolation of at least 30dB is necessary between the VCO to other circuit (modulator in your case). I think for Mobile phone application, data rate is quite low (less than 4Mbps). So isolation is very important to minimize the pulling effect. WHen that effect take place, you will see degradation of phase noise when it is connected to modulator. Phase noise degraded significantly when the data rate is low (i.e. 10kbps). Even good shielding is useless if isolation between VCO and modulator is insufficient.

Grounding is the most critical issue on stable PA design. I think you also realized bias network is another consideration. I presume you have the knowledge in grouding and bias network design. It is quite hard to explain the fundamental here. It will be easier if you tell me what is your actual PA problem.

Cheer~~
 

activewei

Full Member level 2
Joined
Sep 19, 2002
Messages
125
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Location
UK
Activity points
1,433
Olxx: The degradation can be found after modulation. In vector analysis, Error vector magnitude (EVM), phase error, frequency offset are some of the parameters used to evaluate the quality of modulation. Direct connection between modulator and synthesizer WITHOUT SUFFICIENT ISOLATION will degrade above parameters due to non-linear characteristic of the modulator.
 

egemini

Full Member level 2
Joined
Apr 2, 2002
Messages
132
Helped
8
Reputation
16
Reaction score
5
Trophy points
1,298
Location
GOC
Activity points
990
Thank you activewei
My actual problem is VCO phase noise is very badly affected when PA is ON.
Note that we are generating VCO frequency which is twice the required.
there is a divider in modulator. I have applied sufficient isolation even then it did not improve much. The consequence of this problem is that ACP requirements are difficult to meet.

I have one more question. Is there any concrete relation between EVM and BER. EVM can be measured using a VSA but if it can be mapped to correspoding BER it would have been very useful..

thank u friends
 

activewei

Full Member level 2
Joined
Sep 19, 2002
Messages
125
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Location
UK
Activity points
1,433
Hi Egemini,
What is the tx transmit power? is the frequency synthesizer board located near to the PA? Let me know this before I comment further.
I am not very sure about the relationship between EVM and BER. But what I can tell you is:

EVM-> to show how "linear" is your circuit in vector analysis.
BER-> if EVM is poor, the BER will be worse.

I have gone through vector analysis, I am going to try out BER testing next. =)
 

mckinson

Full Member level 2
Joined
Sep 3, 2003
Messages
126
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,298
Activity points
1,034
HI,activewei:
Recently, I 've found the phase error be lower if I changed the raising edge shape of the PA power ramp .
Could you please tell me the reason?
 

mckinson

Full Member level 2
Joined
Sep 3, 2003
Messages
126
Helped
3
Reputation
6
Reaction score
3
Trophy points
1,298
Activity points
1,034
But ,I found the switching spectrum is failed . Could you tell me the relation between the shape of power ramp and the switching spectrum .and how can I do ?
 

activewei

Full Member level 2
Joined
Sep 19, 2002
Messages
125
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Location
UK
Activity points
1,433
Hi mckinson,
"if I changed the raising edge shape of the PA power ramp "-> could you send me the waveform captured for further analysis? you can email me at activewei@yahoo.co.uk

"I found the switching spectrum is failed "-> what do you mean by switching spectrum?
 

xinxin

Full Member level 3
Joined
Oct 8, 2003
Messages
184
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,298
Activity points
1,238
You can adjust the burst time , such as up and down , and also reduce the power , your switch spectrum can improve.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top