field test unit no 001 "model T"

[mickstv]

That's funny I just logged on to suggest the same thing. Might sound obvious but check that the descrim mosfet is not upside down.
Moodz

Hi Paul, I just checked again I have the source pin RHS going to coil + "7.5volts" and the drain going via a 100ohm resistor to L1 coil. I did notice that if I removed the 10uf cap from the gate, the waveform was also normal but couldn't be adjusted.


Mick

[moodz]

Hi Paul, I may have found the problem area I removed the descrim circuit including the IRF740 and the 100ohm resistor I then added a 470ohm resistor across L1 side of the Diff coil. The waveform now settles as you would expect and the descrim between ferrous and non ferrous as read on the cro works perfectly.

Now just have to sort out the problem with the descrim circuit.

All the checks are on the coil input no blocking fets fitted.



Mick

That's funny I just logged on to suggest the same thing. Might sound obvious but check that the descrim mosfet is not upside down.
Moodz

[mickstv]

Hi Paul, I may have found the problem area I removed the descrim circuit including the IRF740 and the 100ohm resistor I then added a 470ohm resistor across L1 side of the Diff coil. The waveform now settles as you would expect and the descrim between ferrous and non ferrous as read on the cro works perfectly.

Now just have to sort out the problem with the descrim circuit.

All the checks are on the coil input no blocking fets fitted.



Mick

[mickstv]

Mick,
your TX is mostly OK except that after the flyback the voltage is low ???? for a period ... this is definitely wrong ... see red area on your scopeshot. you can see a step in the voltage during the RX period this is not OK.



moodz

[ATTACH]17031[/ATTACH]

The only difference between the your second schematic and my design are the frontend diodes, mine are HER208 and yours are FS18G. I can't even find a match in google for FS18G what are they ?

I'll change them for BYW96e tomorrow and see what happens.


Apart from that I'm not sure what else to try.



Mick

[moodz]

hmmm ...

Mick,
your TX is mostly OK except that after the flyback the voltage is low ???? for a period ... this is definitely wrong ... see red area on your scopeshot. you can see a step in the voltage during the RX period this is not OK.



moodz

[mickstv]

Hi Paul, here are the extra shots I tool with out the blocking fets.

[mickstv]

Disconnect the blocking FETs completely and check the positive and negative flyback pulses on both sides of the coil ... should be about +/- 25O volts with that coil.

moodz.

No worries. I won't get a chance now but I'll check it tomorrow morning and post the results including pictures of both sides of the coil without the blocking fets attached. I'll also mark on the pictures the CRO gnd centre lines.


Mick

[moodz]

Hi Paul, yep the centre connection of the coil has 7.4volts on it. I should clear up that was with the CRO directly connected to the coil so it also has the diode in line with the fet.

Disconnect the blocking FETs completely and check the positive and negative flyback pulses on both sides of the coil ... should be about +/- 25O volts with that coil.

moodz.

[mickstv]

mick
where you have a picture with this caption ...
"The picture below is of fet sife of the coil and below is the tx pulse."

on the fet side of the coil your flyback pulse should be positive ... but it is going negative from the looks of it ... something is wrong .... make sure the centre connection of the coil is connected to the 7.4 volt rail.

moodz.

Hi Paul, yep the centre connection of the coil has 7.4volts on it. I should clear up that was with the CRO directly connected to the coil so it also has the diode in line with the fet.

[moodz]

mick
where you have a picture with this caption ...
"The picture below is of fet sife of the coil and below is the tx pulse."

on the fet side of the coil your flyback pulse should be positive ... but it is going negative from the looks of it ... something is wrong .... make sure the centre connection of the coil is connected to the 7.4 volt rail.

moodz.

[mickstv]

I just noticed that the descrim works when I have the TX set to 50us or less, but for some reason anywhere between 55 and 100us all targets get detected the same.



Mick

[mickstv]

Well here's what has happened. I've built another coil pictured below 24cm dia and 29 turns, wire from cat 5 cable. All Cro shots are 5v p-p at 50us per div.

With the new coil and once I looked in the right spot I can see the cro showing the difference between ferrous and nonferous objects.






The picture below is of fet sife of the coil and below is the tx pulse.




The picture below is FET side of diff coil and bottom picture is after the n-channel blocking fet.




The picture below is the other side of the diff coil top and after the p-channel blocking fet below.




This final picture is of the output top N-channel and bottom P-channel blocking fets.

[mickstv]

22 cm coil with 32 turns. I wound a coil using cat5 and it runs a little better than my usual coil ... I can sample at 5us. Descrim is good.

Watch out for the amplifier .. protection diodes at the input can upset the bias on the blocking fets and either blow them up or sink current which will stuff up the coil performance. This is the area I am fiddling with now ...previously I could barely run the amp at a gain of 10 without overload ... Now it is running at gain of 2000. ( the noise actually improves on these amps with higher gains as the bandwidth decreases )



I would like to see how you go with NP fet pack ... I am still moving things around on the front end circuit to get the demodulator channels completed ... the blocking fets are still in place so dont worry ... however I blow a set up every day on average ... the ones I am using bsp254 and bsn254 I have quite a few so I will try the fet pack when I stop probing.

moodz

Good to see the high gain on the THAT ic, it will be interesting to see the circuit mods once you've finished.

The circuit I build is exactly the same as your second schematic, only I haven't fitted the THAT ic yet, till I get this front end running correctly. What I'll do before I post up any pix tomorrow I'll wind another coil the same as the one you just made and try it. I'm still only on the first fet pack, who knows maybe they will be more robust.

I won't put the THAT ic in till you post up the possible changes, because as you would know it isn't cheap.



Mick

[moodz]

I know you have a thread on the diff coil but at the moment what diameter coil are you testing and how many turns.


Also did you end up ordering the PHC2300 fet pack IC's or do you want me to send you a couple for testing in your circuit.



Mick

22 cm coil with 32 turns. I wound a coil using cat5 and it runs a little better than my usual coil ... I can sample at 5us. Descrim is good.

Watch out for the amplifier .. protection diodes at the input can upset the bias on the blocking fets and either blow them up or sink current which will stuff up the coil performance. This is the area I am fiddling with now ...previously I could barely run the amp at a gain of 10 without overload ... Now it is running at gain of 2000. ( the noise actually improves on these amps with higher gains as the bandwidth decreases )



I would like to see how you go with NP fet pack ... I am still moving things around on the front end circuit to get the demodulator channels completed ... the blocking fets are still in place so dont worry ... however I blow a set up every day on average ... the ones I am using bsp254 and bsn254 I have quite a few so I will try the fet pack when I stop probing.

moodz

[mickstv]

That sounds fine ... must be the coil ...

I know you have a thread on the diff coil but at the moment what diameter coil are you testing and how many turns.


Also did you end up ordering the PHC2300 fet pack IC's or do you want me to send you a couple for testing in your circuit.



Mick