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.

Register Log in

How to get rid of DC offset from VGA output?

Status
Not open for further replies.
W

wee_liang

Junior Member level 3
Joined
Jan 4, 2005
Messages
26
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,283
Activity points
258
VGA dc offset problem

If DC offset is a problem in my VGA output, how should I get rid of it? (Calibration methods avoided please). Does putting ac coupling caps help? How do I bias the next stage? What value of caps should I use if I have a 4MHz signal?
 

Re: VGA dc offset problem

You can use AC coupling to remove the propagation of DC offset.
Use a cap value that gives you the desired corner frequency in your design (watch out for settling time problem).

G.L.
 

Re: VGA dc offset problem

DonJ said:
You can use AC coupling to remove the propagation of DC offset.
Use a cap value that gives you the desired corner frequency in your design (watch out for settling time problem).

G.L.
Click to expand...

How to solve the settling time problem? Do you know what't the settling time
requirement of 802.11a?

Thanks,
 

Re: VGA dc offset problem

In 802.11ag you have 10*800ns for all AGC, offset correction, frequency offset estimation and symbol correlation or synchronisation. So it depend on the capabilities of the BB how much time remain. Typical is about 1-2 short 800ns preamble symbols. They call the short symbol P16.
 

Re: VGA dc offset problem

rfsystem said:
In 802.11ag you have 10*800ns for all AGC, offset correction, frequency offset estimation and symbol correlation or synchronisation. So it depend on the capabilities of the BB how much time remain. Typical is about 1-2 short 800ns preamble symbols. They call the short symbol P16.
Click to expand...

10*800ns for so much stuff. If we allot 2 symbols for AGC settling. the settling time
must be smaller than 1.6us. But, the AGC shooting is not possibly finished in the
first time, so we assume 5 shootings to obtain the correct AGC gain setting.
And, each shooting must be settled within 1.6us/5 which is around 300ns.

For 1% settling accuracy, the time constant should be smaller than 60ns. It means
the corner of high pass filter should be less than 1/(2*pi*60ns)~2.6MHz.
Is my calculation right? Do it hurt the signal quality in so large corner freq?
 

Re: VGA dc offset problem

IEEE 802.11ag was not intended to implemented with ZeroIF radios. The issue is that there is a frequency difference between the zero subcarrier in the OFDM signal and the radio zero frequency. This frequency mismatch impact also the OFDM signal if the lower corner frequency is low.

The practical solution is to use switchable BWs and an additional filter in the BB. That is from what I know used everywhere except Broadcom. They using a course frequency correction applied to the LO in the radio. So the zero of the OFDM signal and the radio zero are equal. If you compare both solution with optimized parameters you will end up by a small dB EVM advantage for Broadcom. They have a patent on that. All others using the mixed analog switchable and digital filter approach.
 

Re: VGA dc offset problem

rfsystem said:
IEEE 802.11ag was not intended to implemented with ZeroIF radios. The issue is that there is a frequency difference between the zero subcarrier in the OFDM signal and the radio zero frequency. This frequency mismatch impact also the OFDM signal if the lower corner frequency is low.

The practical solution is to use switchable BWs and an additional filter in the BB. That is from what I know used everywhere except Broadcom. They using a course frequency correction applied to the LO in the radio. So the zero of the OFDM signal and the radio zero are equal. If you compare both solution with optimized parameters you will end up by a small dB EVM advantage for Broadcom. They have a patent on that. All others using the mixed analog switchable and digital filter approach.
Click to expand...

I roughly read the several related issue publicated by Teresa Meng of Stanford before. By the way, I can't figure out that "The practical solution is to use
switchable BWs and an additional filter in the BB."
What is additional filter? And, what's the purpose?
 

Re: VGA dc offset problem

Switchable BW mean different discrete selectable lower corner frequency highpass filter to reject the DC offset.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top