field test unit no 001 "model T"

[moodz]

Hi Mick,

I think what your seeing there is the Miller effect.

http://www.richieburnett.co.uk/mosfail.html

It shouldn't normally be a problem and it indicates that your not driving the gate of your coil drive mosfet low hard enough.
Which raises the question how are you driving the gate of your main coil driving mosfet?
I noticed the mosfet driver dropped off the schematic a while back.. I'm not sure why. I hope you didn't make the not so unreasonable assumption that meant you could drive the mosfet directly from the PIC.
If you are using a suitable driver then as a next step I'd recommend scoping the mosfets gate and make sure it is working as it should driving the gate high and low nice and smoothly.

I hope this helps you get it working properly, I'm keen to see someone else replicate moodz's excellent work.

Hi guys ... that is something to check out ( Thanks Midas ).... I am using an IRF740 and it drives OK off the PIC but I guess this issue can happen.

I have got myself some of those blocking FETs you are using Mick so I will look into this on the weekend ... Just a thought ... your "top" diode is not the wrong way round is it ? ( not FET end diode ).

moodz.

[Midas]

Hi Mick,

I think what your seeing there is the Miller effect.

http://www.richieburnett.co.uk/mosfail.html

It shouldn't normally be a problem and it indicates that your not driving the gate of your coil drive mosfet low hard enough.
Which raises the question how are you driving the gate of your main coil driving mosfet?
I noticed the mosfet driver dropped off the schematic a while back.. I'm not sure why. I hope you didn't make the not so unreasonable assumption that meant you could drive the mosfet directly from the PIC.
If you are using a suitable driver then as a next step I'd recommend scoping the mosfets gate and make sure it is working as it should driving the gate high and low nice and smoothly.

I hope this helps you get it working properly, I'm keen to see someone else replicate moodz's excellent work.

[mickstv]

This seems to be looking worse every time I do something to it. no diodes on output of blocking fets

[mickstv]

Mick there seems to be some current flowing ... if you have diodes connected to the ouputs of the blocking fets disconnect them then measure.

moodz

[ATTACH]17143[/ATTACH]

Ok shall do will post up result in about 1/2hour.




Mick

[moodz]

Mick there seems to be some current flowing ... if you have diodes connected to the ouputs of the blocking fets disconnect them then measure.

moodz

[mickstv]

Just took this shot the cro is set to DC input 2 volts p-p taken on one of the blocking fet output's. If this picture isnt correct I don't know whats going on I must mention that I may have done previous measurements on AC input.

[moodz]

Mick ... I think you still have a problem ... the output of the blocking fets should look like below .... if the coil is sitting at 7.4 volts then that is what the voltage should decay to ...

I have wired up a board for field tests ... just finalising the amplifier ....



moodz

[mickstv]

Some more pictures with the blocking fet package attached. No amp attached. Descrim running at 710.

For some reason it's taking approx 20us to 0 volts. I can get it faster but the damping code has to be set to 10 on the display.

[moodz]

Hi Paul, the lead was basically flat with a slight wave in it. I did have one lead sinker 3cm diameter that detected the same way as iron also but all other smaller sinkers detected the same way as copper.

Have you done tests with different TX widths did you find the disc changing.




Mick

I have not had time to sit down and test all combos of pulse timings vs targets ... I imagine there could be some change ...

I generally run Tx at 100 us.

moodz.

[mickstv]

That is good news ... I have not tested a large piece of lead ... was it rolled up ? that could affect the descrim.

You can fit the blocking fets and diodes as per the last schematic but no amplifier yet and check that blocking should now be working.
The input of the THAT amplifier has diodes fitted and I not sure that these are interacting somehow with the front end ... looking into it.

Otherwise the channel mods are coming along ... but I have been busy at work so not as fast as I would like.

moodz.

Hi Paul, the lead was basically flat with a slight wave in it. I did have one lead sinker 3cm diameter that detected the same way as iron also but all other smaller sinkers detected the same way as copper.

Have you done tests with different TX widths did you find the disc changing.




Mick

[moodz]

I've found the problem with the descrim front end, the 78L05 that I had fitted was drawing too much current. Once replaced with a new one the input voltage went upto 15-16 volts and the Descrim section started damping correctly.

Without the blocking fets I've been mucking around with various targets. Copper and lead sinkers go one way and Iron targets go the opposite way when looking at the CRO.

The only exception was a sheet of plumbers lead it's amplitude went the same way as iron, not sure why that would be.

Next will be fitting the blocking fets, Paul you were walking about mods to the frontend just wondering how that's going and whether the blocking fet setup is different to the second circuit you posted. If it is I'll wait till you post up the next circuit.



Mick

That is good news ... I have not tested a large piece of lead ... was it rolled up ? that could affect the descrim.

You can fit the blocking fets and diodes as per the last schematic but no amplifier yet and check that blocking should now be working.
The input of the THAT amplifier has diodes fitted and I not sure that these are interacting somehow with the front end ... looking into it.

Otherwise the channel mods are coming along ... but I have been busy at work so not as fast as I would like.

moodz.

[mickstv]

Without the blocking fets I've been mucking around with various targets. Copper and lead sinkers go one way and Iron targets go the opposite way when looking at the CRO.

The only exception was a sheet of plumbers lead it's amplitude went the same way as iron, not sure why that would be.


Mick

Further to the the story about the A4 sized lead sheet, I cut off a piece that was 2cm wide and 15cm long this smaller piece tracked on the cro the same way as the lead sinkers and copper.

Also the TX pulse width seems to effect the result I have found the higher you go the more likely the targets are going to be detected the same way on the CRO.


Mick

[mickstv]

I've found the problem with the descrim front end, the 78L05 that I had fitted was drawing too much current. Once replaced with a new one the input voltage went upto 15-16 volts and the Descrim section started damping correctly.

Without the blocking fets I've been mucking around with various targets. Copper and lead sinkers go one way and Iron targets go the opposite way when looking at the CRO.

The only exception was a sheet of plumbers lead it's amplitude went the same way as iron, not sure why that would be.

Next will be fitting the blocking fets, Paul you were walking about mods to the frontend just wondering how that's going and whether the blocking fet setup is different to the second circuit you posted. If it is I'll wait till you post up the next circuit.



Mick

[mickstv]

Measure the gate voltage as you adjust damping ... also measure the voltage on the cap connected L2 via diode ... should be something like 14 volts or so. The FET might be punched through ...though I supposed you changed it already.
moodz

From memory the cap on L2 had about 12-13 volts. I haven't replaced the fet yet I'll do what in the morning and see what happens.


Mick

[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

Measure the gate voltage as you adjust damping ... also measure the voltage on the cap connected L2 via diode ... should be something like 14 volts or so. The FET might be punched through ...though I supposed you changed it already.
moodz