Temperature change the depth (target)

[Gilmar Ribeiro de Oliveira]

I finsh my PI detector and I have proble with change of tempereture, The depth decreise with the increse of the temperature.

I'm using the OPA132 a verry good opamp with temperature and I use microcomtroller for generate the pulse and delay.

I see that the diode that limite the sign from the coil is cause the tempereture problem, I'm using the 1N4148.

Have someone had this problem ?
Do you know a good diode to use ?

Regards,
Gilmar

[Brian King]

Re: Temperature change the depth (target)

Hi Gilmar,
Silicon diodes have a forward voltage drop of 2.2mv / deg C. (1.25mV / deg F) these voltage drops can vary widely from .4 to .7 volts from device to device +/- the temperature coeficient
(2.2mv/d C).
As a matter of interest, what detector circuit have you built?

Cheers
Brian K

[CsZS]

Re: Temperature change the depth (target) just a question

What would happen if the equipment and coil will put into liquid nitrogen? That is good for noise reduction?

[Gilmar Ribeiro de Oliveira]

Re: Temperature change the depth (target)

Hi Brian, thanks for reply.
I have built a PI detector. Do you had this proble.

Regard,
Gilmat

[Gilmar Ribeiro de Oliveira]

Re: Temperature change the depth (target)

The circuit that I built is from carl with some change for use microcontroller.

regards,
Gilmar

[Reg]

Re: Temperature change the depth (target)

Hi Gilmar,

Your problem is an interesting one. Now, do you have a scope and if so, just what do you see changing?

I have a question about your problem. If you increase the sample delay a little, does the problem still exist? I have seen slight changes as the result of the FET characteristics changing with temperature. What I have seen is the decay curve alters a little so the signal at the output of the preamp changes along with it.

Now, if you are sampling right at the edge, meaning that you are sampling on the edge of the decay curve, then a slight change could cause the diff integrator to go into saturation. This would give the appearance of a sensitivity loss. This type of problem I have seen.

If this isn't what is happening, then I suspect you may have a component failing with temperature. If this is the case, then a can of freeze could be used to cool the components until the culprit is found.

Reg

[Brian King]

Re: Temperature change the depth (target) just a question

If you mean freezer spray used to cool components as a service tech might do and as Reg has suggested then yes,you could better locate a heat related component problem this way.
You could simply have an out of spec part. What are you using as a time base for the PIC chip , an RC network or crystal? Is your oscillator stable?

I would do the following;

1) check supply voltages using a CRO. You should see clean voltages with no (or very little) digitally generated noise riding on top or below the supply voltage. If the supply is not clean (filtered and decoupled)or does not have enough current capacity you will see a lot of peaks and troughs on the supply lines as the transmit pulse switches the FET(s) on and off.

2) Look at FET gate drive circuitry. You should see clean sharp rise and fall times at the gate of the FET eg transmit pulse. FETs tend to get warm when operating in their linear region and this can happen with slow rise and fall times on their gates. Not sure what you are driving the gates with?

3) Look at the gain stage(s) with a CRO .Are you operating within the dynamic range of the OPAMP? Knowing your pulse freqency and width is a good thing. Some Opamps have low noise and high gain but trade this off against bandwidth and input offset current. Look at the step recovery data of your particular amp and compare this against other opamps used in PI circuits.

4) What do you see with a CRO on the coil signal line? Diode clamp circuits should not impact on a signal other than to clamp at forward voltage drop above or below a reference. -v / +v or GND.

I am probably preaching to the converted here but sometimes it pays to go back to basics!


Cheers

Brian K

[Brian King]

Re: Temperature change the depth (target) just a question

On the subject of liquid nitrogen.
It would fix some noise problems but also crack a few components!

Cheers

Brian K

[CsZS]

Re: Temperature change the depth (target) just a question

Hello Brian,

Thank You for the advices. I have a desktop panel only, it works fine, only I would like to know there is any reason to think about it, to do efforts and spend some money in the future for it.
Here is my test panel:
http://www.moricz-bp.sulinet.hu/~tklub/kapcs1.jpg

[Brian King]

Re: Temperature change the depth (target) just a question

Nice job on your test panel. Well done.

Sorry I messed up my replies to both yourself and Gilmar....late nights...

Cheers

Brian K

[Gilmar Ribeiro de Oliveira]

Re: Temperature change the depth (target)

Hi Reg Thanks for your help.

I'm working in this problem for 1 month, without success, I change amp op, delay, pulse with, resistor 1%, capacitor NP0 and I discovery that the diode IN4148 is the problem, because I separate with a wire the diode and if I warm or cold the diode and through the scope I see output change more the 1V.

If you want I can send to you the schematic.

Thank you.
Gilmar

[Gilmar Ribeiro de Oliveira]

Re: Temperature change the depth (target) just a question

Hi Brian Thanks for your help.

I use the PIC with crystal.
The drive of gat is good.
I'm working in this problem for 1 month, without success, I change amp op, delay, pulse with, resistor 1%, capacitor NP0 and I discovery that the diode IN4148 is the problem, because I separate with a wire the diode and if I warm or cold the diode and through the scope I see output change more the 1V.

If you want I can send to you the schematic.

Thank you.
Gilmar

[Reg]

Re: Temperature change the depth (target)

Hi Gilmar,

I would appreciate a copy of the schematic. You can send it to:
[email protected]

Thanks,

Reg

[Brian King]

Re: Temperature change the depth (target) just a question

Hi Gilmar,
Yes it might be a good idea if you could send schematic, then I can help you further.

Cheers

Brian K

[Brian King]

Re: Temperature change the depth (target) just a question

Gilmar,
Had a quick look at your circuit.
You could try a small signal schottky diode such as BAT49 or similar in place of the 1N4148 types.
Schottky diodes tend to have a much lower forward voltage drop,tighter temperature coef and faster reverse recovery or switching time typically around 4 or 5 nano secs.

Also try and limit the current into the Opamp.

Let me know how you go.

Cheers

Brian K