Filtering?

[Anonymous]

Hi all,
I was playing tonight with my bredboard and thought I'd tell you what I did and see if I can do 2 things. 1)learn something 2)spark some action in the class room. Hey, Summer vacation is over, time to go back to school right?
Out of the front end op amp it appears to be pretty noisy so I thought about what I had seen in some of the patents I looked at. There seems to be a low pass filter incorporated in the "receiver circuit" but of course all I've seen has been block diagrams, not schematics in the patents. I wired up a Sallen-Key low pass filter after the front end and then looked at the output on the scope. It does seem to clean up the signal but...
First let me say that the collapse of the magnetic field induced from the transmit is finished in about 10uS so watching the scope connected at the output of the front end amp, I pass a gold wedding band and watch the results. I can see it affect the curve before it flatens out. If I look at the output of the filter it appears that there is no movement in that area of the curve on the scope. This would seem to indicate to me that it would reduce the depth detection of less conductive metals (thin gold wedding bands).
Has anyone else tried filtering this way? Did you find it to improve things or not? Any thoughts on advantages or disadvantages to filtering would be appreciated.
Thanks,
Charles

 

[Anonymous]

Hallo, I'm also thinking on filtering. It should give better sensetivity! But as far as time constant of some finds could be even less then 10us, the bandwith of the LPF should exceed 100kHz. So I would trie with about 200kHz. What bandwith did you use? Eric, do you have any recomendations?
thanks
Walter

 

[Anonymous]

Hi Charles
Yes I have thougt a lot on filtering. I can not see the reason why picking up the whole frequence range 0 Hz to several MHz. I would suggest to start filtering from around 200Hz low to a few hundrede kilohertz. There is a lot of 50/60 hz in the ground from main power lines and a lot of noise from radio communication like celluar phones at higher frequences. So defenately we must focus on the area our signals is covering to boost sensitivity. So from my point of view we need a band-pass filter of some kind.
Mark

 

[Anonymous]

To add to Mark's BPF suggestion, I would hope that that a Bessel is used as it will not distort pulse's shape (ie. square wave in , square wave out). A Bessel has the best group delay and transient response of all topologies, which is why it's used most in signal processing.
Randy Seden

 

[Anonymous]

Hi all,
I did use a bessel because of the characteristics of it, I figured it would be the best choice in this application. However, I did have a much lower cuttoff frequency of only a few Khz so maybe I'll try opening it up and see what affect it has. As just a shot in the dark last night I tried it and then posted the message in hopes of hearing other experimenter's thoughts and it seems there are a lot of good ones out there. I will try to widen the passband and see what difference it makes and then add another filter to see how a band pass filter affects it.
Thanks,
Charles

 

[Anonymous]

Hi Walter and everybody,
Now there

 

[Anonymous]

Hi Mark,
One of the problems with frequencies that are outside of the front end amplifiers bandwidth is that the protection diodes often demodulate the carrier and it is is resultant audio that causes the interference. I have used r.f. filters before the front end with some success. Ferrite beads or sleeves on the Rx input leads help in this respect.
Eric.

 

[Anonymous]

Hi Eric
Thanks for the good answer, I will try to do some filtering with ferrite beads and sleeves. Talking about protection diodes, have you noticed that the NE5534 you often suggest as a frontend amplifier actually have build in reversed protection diodes.
Would there be any idea using those instead of external ones ?.
Mark