Quadrature demodulator:I not locked at Q

I have a problem with a quadrature demodulator TDA8040T (see datasheet )
**broken link removed**
because the I channel is not locked at Q channel . I have tried to modify a capacitor at low pass filter ( see attachment Low-Pass Filter.doc ) of I channel from 220 pF to 60pF and I locked with Q channel but this is abnormal because two filters must to be equal .What do u think of this problem? What measures of test do you suggest me to do?

Regard
FOX

Do you mean what you says that phase difference between channels is not 90 degree?

If so , there is another problem because as I saw, the IC is completely symmetric.

There is one possibility, VCO signal does not drive 90 degree to the mixer du to low Q factor utilization.In this case you may not be obtained sufficient drive level.
Another possible problem may be high signal drive ( most possible). A high signal drives input amplifier into nonlinear region and amplifier disturbs phase between I and Q channels.

I recommend you that to use Bessel type filter for minimum group delay.
Otherwise you can be face to face signal delays between lower and upper frequency limit points.

GL

Do you mean what you says that phase difference between channels is not 90 degree?

Yes.When the phase difference between channels is not 90 degree them are not locked . The vco ( center frequency : 280Mhz , RFin:140Mhz ) of TDA 8040 is controlled from a demodulator controller and when the phase difference between channels is 90 degree them are locked and the frequency of vco is fixed at 280 Mhz . I see at open loop a correct voltage at vco of receiver difect .


There is one possibility, VCO signal does not drive 90 degree to the mixer du to low Q factor utilization.In this case you may not be obtained sufficient drive level.

What test misure do here ? Misure only level carrier 280 Mhz ?


Another possible problem may be high signal drive ( most possible). A high signal drives input amplifier into nonlinear region and amplifier disturbs phase between I and Q channels.

Are talking of amplifier internal at TDA 8040 ? If yes , output of low pass filter is a input for amplifier of TDA 8040 . If i calculated well the frequency response od filter , it is not flat and when reduce capacitor from 220 pF to 60pF become more flat change signal level and group delay of course.Perhaps an amplifier of the TDA 8040 has characteristics a little diverged from the specific ? I will try to change it again.

I recommend you that to use Bessel type filter for minimum group delay.

I thought at Bessel type filter too but this is not my design and is not prototype because is in production from at least two years.
The problem is only on some receiver , i need to search difect components on bad receiver . I here would like have of the suggestions to do a good diagnosis , search breakdowns .

Otherwise you can be face to face signal delays between lower and upper frequency limit points.

How to do this with instruments ?

I got some questions..

- Do you use any PLL for locking to 280MHz for VCO ?If you don't , you should be used. If yes , Loop Filter of PLL is calculated well. Because Phase Noise may affect the phase relations much..Additional to , you should use high Q factor at resonance circuit to obtain stable LO signal and low harmonics.. Because if harmonics' levels reach to certain level , it disturb operating of the mixer.

- I have meant that as input amplifier which has been at the input of the IC. This is a RF amplifier and I guess IP3 of it is not sufficient regarding to your input level.. Have you ever checked that related to the specifications ?

-If you can see I and Q channels before filter by applying X and Y channels of an oscilloscope , this is better for solve the problem..

-Does your IQ Modulator work well ? ( most stupid error source )

For the time being , that's it..

Rgrds

Do you use any PLL for locking to 280MHz for VCO ?

Yes and the vco works well . The control voltage of vco misured at open loop is correct and the vco shifts well the frequency if change control voltage from external .

I have meant that as input amplifier which has been at the input of the IC. This is a RF amplifier and I guess IP3 of it is not sufficient regarding to your input level.. Have you ever checked that related to the specifications ?

I understand now. The segnal at input of RF amplifier on IC has a correct level , i saw it with a spectrum analyzer .

Does your IQ Modulator work well ?
Yes, i m sure because i tried it with a other receiver and works fine and then i tried a new , good , IQ Modulator with the bad receiver but have same problem .

If you can see I and Q channels before filter by applying X and Y channels of an oscilloscope , this is better for solve the problem...

Is possible do this misure with I and Q not locked ? The phase difference between channels is not constant in this case.

And then can u explain this " Otherwise you can be face to face signal delays between lower and upper frequency limit points"

The story so far

The story so far:

CLUE: This demodulator has been used in a receiver that has a two-year history of successful performance in the hands of customers. DEDUCTION: The basic design is correct.

CLUE: Recently some of the receivers have failed test after manufacturing. DEDUCTION: There is a batch failure on one or more of the parts.

GAME PLAN: (1) Examine the construction records and see which component lots are common to the failed receivers. (2) Use a microscope to compare working and failed boards. What is different? (3) Perform the test described in note 4 of page 8 of the data sheet using frequencies and offsets that will represent the bit rate and IF frequency of your system. (Try different offset frequencies) Do the described sine waves exist? Are they the same amplitude and ¼ cycle out of phase? Are they clean and free of jitter?