Menu
What's new
Find's Treasure Forums

Welcome to Find's Treasure Forums, Guests!

You are viewing this forums as a guest which limits you to read only status.

Only registered members may post stories, questions, classifieds, reply to other posts, contact other members using built in messaging and use many other features found on these forums.

Why not register and join us today? It's free! (We don't share your email addresses with anyone.) We keep email addresses of our users to protect them and others from bad people posting things they shouldn't.

Click here to register!

Need Support Help?

Cannot log in?, click here to have new password emailed to you

Changed email? Forgot to update your account with new email address? Need assistance with something else?, click here to go to Find's Support Form and fill out the form.

MD-Sensitivity

A

Anonymous

Guest
Good morning Bill and Erick, I've been "lissening" carefully in your classroom, and I find it great, so thank you for your time and for sharing your experiance with another starting detektorist. Beeng an electronics engineer and hoping to be able to build a well sofisticated detector, I have some hot questions:
Coil design - as you mentioned "there is only a limited amount of shielding that can be put round a coil" - can you precise please, how should a good shielding look like?
Methods of noise reduction - as far as I understand it, the cirquit interpolates (analog) the voltage sampled at 100s of Pulse-Responses, gaining this way an S/N-Ratio. Did somebody ever try a (fast) digitalisation of every Pulse-Response and process digitally the signal for better noise/disturbance reduction? Can you give me some advice in this matter? What is, what you call "novel filter systems"?
Sample delay - it is among the factors effecting PI detection range. If it would be possible to colaps the field quicker without reducing the Pulse Power, lets say 10us instead of 20us, how much increase of sensitivity (on smal objects) should one expect? The question concerns first, that the speed of field colaps is doubled, ergo emc doubled, and second, sampling is performed halve the time earlier, meaning bigger signal level. (please dont take discrimination problems into account)
Switching element - what kind of switch is common used - high voltage MOSFET, an IGBT or a high voltage Bipolar transistor? Can you give a short comment for this choice?
regards Walter
 
Hello Walter,
I did some investigations on direct sampling the amplified pulse response because I expected some differences in the reponse between ferrous and nonferrous materials and intended to use these differences for target identification.
Until now my effort led only to poor results. The reason for this is mainly my low experience with the design of low noise, high speed amplifiers and poor knowledge about proper coil shielding. I think I have good experience in the field of digital signal processing. My last system offered a sampling time of 1
 
Hello Stefan,
my opinion about MOSFET on the transmitter side is that it has too much parasitic capacitance on the drain (which appears parallel to the coil), therefore, I think, a BIP or IGBT should perform better. Did you verify that there is no ringing / What R value have you in parallel to the coil? I also believe, that a 2-coil PD should have better performance. For the pre-amplifier I pointed out the SSM-2017 (Analog Devices), which has the lowest voltage noise density I know. Eric, can you tell us a good choice for the cutof frequency for noise shaping (3dB corner of the gain of the pre-amp)? Stefan, can you provide me digital records of the sampled signal? Besides the sampling time, what repetition rate do you use? How long after Pulse do you start sampling? regards walter
 
Hello Walter, in my last prototype I used a 470 Ohm damping resistor in parallel to the coil. I started a try out with higher values and reduced it until the point of critical damping was reached and no ringing occured. Next I'll try out a series diode between coil and mosfet and an additional resistor like Eric mentioned some time ago in his forum.
I used a sampling time of 1
 
Hi Stefan and Walter,
I will try posting a front end circuit in the next day or so. I drew one out yesterday but the quality when I scanned it was not too good. Will redraw today in blacker ink. I'm still very interested in pursuing this with you; just been very busy of late producing detectors with little time for development.
Eric.
 
Hi Stefan
It seems that your having problems with these op-amps. I know from experience that these current feedback amplifiers can be extremely nasty and do strange things. A good look at the application notes by Analog components gives some useful advice. They are very sensitive to any reactive comnponent in the feedback path. This can include capacitive coupling caused by track layout. It may be better to choose a voltage feedback amplifier. These are easier to apply since stablity is less of an issue with them. The gain bandwidth product can be sorted out with the required feedback components (R//C) to give the required frequency response shutting out a bit of the noise.
kind regards
Steve
 
Hello, adding a diode is a good idea. It should be a fast one (ev. shottky?). Differential input is also preferable. Band limitation to about 100kHz (Eric?) should also reduce HF-noise. Can you tell me, what distance between samples do you use? also 1.0us? That would mean, the coil needs 10+20=30us or 10+30=40us to come within the range of the ADC! With separated coils it should be possible to get it much quicker. I think, the earlier you can measure, the better the sensitivity. But it is very interesting how much. Take care also for the 50Hz Power line disturbance - you can simply take the middle value of two corresponding samples with 10ms distance (at 100Hz repetition rate very easy) regards Walter
 
You must log in or register to reply here.
Top