How to manage sync noise in PI

[revmaster]

Hi,

I have another problem for discussion.
If freq of PI is not synchronized with main AC then it causes a lot of periodic noise in the circuit. One has to manually adjust every time PI is used to get optimum results, as i think this is the main sourse of noise when using detector in room. Is there any way through which we can adjust automatically.

I tried one method but wasnt fool proof. At power ON calib read peroidic devitations thourgh PIC micro and keep freq band changing & at the end put PI to best freq which has lowest noise, but this option sometimes worked and sometimes didnt. Might be due to bad program code.

But do we have any easier way out ? has anybody worked on it ? I think with better sync one can exploit more sensitivity. I hope am not confusing anybody.........

Regards,
Michael

[Max]

Hi,

I have another problem for discussion.
If freq of PI is not synchronized with main AC then it causes a lot of periodic noise in the circuit. One has to manually adjust every time PI is used to get optimum results, as i think this is the main sourse of noise when using detector in room. Is there any way through which we can adjust automatically.

I tried one method but wasnt fool proof. At power ON calib read peroidic devitations thourgh PIC micro and keep freq band changing & at the end put PI to best freq which has lowest noise, but this option sometimes worked and sometimes didnt. Might be due to bad program code.

But do we have any easier way out ? has anybody worked on it ? I think with better sync one can exploit more sensitivity. I hope am not confusing anybody.........

Regards,
Michael

Hi Michael,
PIC could be a problem source if you exactly don't knows what's going on (e.g. watchdog on etc). Simple method to avoid much ac noise from power lines is a simple frequency adjust pot...with some hundred hertz range over original frequency.
You can find positions with pot where harmonics are greatly reduced.
Take a look at goldscan IV schematic and you will see something like this on main oscillator circuit (555).

Best regards,
Max

[revmaster]

Hi Michael,
PIC could be a problem source if you exactly don't knows what's going on (e.g. watchdog on etc). Simple method to avoid much ac noise from power lines is a simple frequency adjust pot...with some hundred hertz range over original frequency.
You can find positions with pot where harmonics are greatly reduced.
Take a look at goldscan IV schematic and you will see something like this on main oscillator circuit (555).

Best regards,
Max

Thanks Max, but am already using sync pot to adjust freq. Now am looking for a solution with automatic adjustment for freq to deal with ac noise. ???

[Max]

Thanks Max, but am already using sync pot to adjust freq. Now am looking for a solution with automatic adjustment for freq to deal with ac noise. ???

Hi revmaster,
yes, sorry. OK, I don't know for sure but maybe a pic could do the work too.
Problem is identifing beats due to noise and main clock frequency. But this requires multi-frequency searching by something like a freq-voltage converter then connected to an adc that go to the pic inputs...complex.
Maybe another device with integrated analog-adc inputs can reduce complexity. But think that program code must be smart enough...and this is a problem if one have to use pics...something with few memory and speed and limited instruction set too.

Seems not so easy to do in a cheapy way, but I think could be done.
In effects...it's boring to tune manually every time...

Best regards,
Max

[Carl-NC]

If freq of PI is not synchronized with main AC then it causes a lot of periodic noise in the circuit.

You're talking about house mains, 50Hz/60Hz? A good way to deal with that is to use a Figure-8 coil.

- Carl

[revmaster]

You're talking about house mains, 50Hz/60Hz? A good way to deal with that is to use a Figure-8 coil.

- Carl

....but i think figure 8 coil will get me lower sensitivity ? there should be a gain loss of about 50%.

[revmaster]

is there any way to handle this sync automatically ???????

[Carl-NC]

Well, if your concern is the lab bench, then the Figure 8 coil is easiest. Sensitivity will be roughly equal to a round coil that's the same size as half the 8.

But let's say you want to deal with interference in the field. You can only account for one frequency, so let's assume 50Hz (60Hz in the US) interference. I would probably set up a separate coil in the control box with a resonance at 50Hz, and make that part of a PLL loop. Then the output of the PLL can be whatever multiple you want to run the pulse at, say 750Hz. If the signal level in the 50Hz tank is below some threshold, then an internal 50Hz oscillator takes over.

However, I suspect that running at a multiple of 50Hz will still give some sort of time-varying integration signal. Ideally, you would want to run the pulse rate at the 50Hz interferer.

- Carl