500 nano second sampling

[Tinkerer]

Tinkering with coils, specifically trying to build a large fast coil.

This TX coil has a diameter of 1 meter. Balanced RX coil.


The picture shows the scope set at 500ns

The red trace is the TX switch OFF
The blue-white trace shows no target
The yellow trace shows a foil target.

I am quite happy with the coil arrangement, however, I also have a problem that I need to resolve:

We see on the red trace some switching noise oscillations. These oscillations have propagated to the battery ground and of course get amplified in the pre-amp.

The system is soldered on perforated board, so now I want to hunt down the exact sources of the noise so that I can eliminate them when designing the PCB's.

Any suggestions are welcome.

[6666]

Eric used to use a 2ohm resistor in the negative lead to decouple the battery from the TX

[Chet]

Hi Tinkerer

Attached is a paper that has some good information on power supply bypassing. Paragraph 7 has some good points on circuit layout. I am curious where you are measuring the -3v red trace?

Regards,
Chet

[Tinkerer]

Hi Tinkerer

Attached is a paper that has some good information on power supply bypassing. Paragraph 7 has some good points on circuit layout. I am curious where you are measuring the -3v red trace?

Regards,
Chet

Thank you for the link. This is the kind of information I am looking for.

The red trace is measured at the mosfet gate driver input.

All the best

Tinkerer

[Tinkerer]

Eric used to use a 2ohm resistor in the negative lead to decouple the battery from the TX

Thank you for the information.

All the best

Tinkerer

[DOOLEY]

hello tinkerer,

been looking real close at your scope shot,

things I have noticed, an its quite subtle .....

your red trace when it returns to rest after the on pulse is still recovering ,
well more accurately your regulator is recovering,
assuming your using a regulator that is........
battery's would do the same thing,
but wouldn't show the harsh spike.

take a close look at the red line from start to finish,
bet there's about 10 to 20 mv recovery there,
only spotted it straight away as I had a similar problem
recently on a project.

the noise you have on the red line after the pulse is still evident
during the on pulse also.

for what it's worth my 2 peneth's worth is a regulator fault / problem.

recently I had a similar problem on a Pi i'm building,
regulator droop and recovery due to high currant load followed by off time,
and also regulator oscillation resulting in line noise.

ah hell, I could be wrong, only a thought.

[Tinkerer]

hello tinkerer,

been looking real close at your scope shot,

things I have noticed, an its quite subtle .....

your red trace when it returns to rest after the on pulse is still recovering ,
well more accurately your regulator is recovering,
assuming your using a regulator that is........
battery's would do the same thing,
but wouldn't show the harsh spike.

take a close look at the red line from start to finish,
bet there's about 10 to 20 mv recovery there,
only spotted it straight away as I had a similar problem
recently on a project.

the noise you have on the red line after the pulse is still evident
during the on pulse also.

for what it's worth my 2 peneth's worth is a regulator fault / problem.

recently I had a similar problem on a Pi i'm building,
regulator droop and recovery due to high currant load followed by off time,
and also regulator oscillation resulting in line noise.

ah hell, I could be wrong, only a thought.

Dooley,
thank you for taking the time to analyse the problem. I will try to improve the power supply for better response to a current step.

All the best

Tinkerer

[tony_av]

Tinkerer,

Use 10x1 probe to verify these measurements

Tony.

[Tinkerer]

Tinkerer,

Use 10x1 probe to verify these measurements

Tony.

These measurements were made with the probe 10x1

[green]

Tinkering with coils, specifically trying to build a large fast coil.

This TX coil has a diameter of 1 meter. Balanced RX coil.


The picture shows the scope set at 500ns

The red trace is the TX switch OFF
The blue-white trace shows no target
The yellow trace shows a foil target.

I am quite happy with the coil arrangement, however, I also have a problem that I need to resolve:

We see on the red trace some switching noise oscillations. These oscillations have propagated to the battery ground and of course get amplified in the pre-amp.

The system is soldered on perforated board, so now I want to hunt down the exact sources of the noise so that I can eliminate them when designing the PCB's.

Any suggestions are welcome.

Hi Tinkerer, Just noticed your new thread. I had a similar problem. Unregulated DC power supply with a 2000uf cap on the circuit board. Adding a 10uf ceramic in parallel on the board, helped a lot. Looking at your scope traces your are doing something I'm not familiar with. Balanced coil, DD,OO? If the coil is balanced why the signal when Tx on? Are you measuring the Rx coil or amplifier out? Why no large signal when the Tx decays. I'm impressed that you could sample so soon. Thanks

[6666]

Was it realy a 10uf ceramic, thats a very big ceramic cap ?

[green]

Was it realy a 10uf ceramic, thats a very big ceramic cap ?

It's a X7R, 16 volt. 3.2x1.6mm chip capacitor. Probably should have ordered a higher voltage. I know I should have for some of the lower capacitance values. Maybe I could see the higher voltage caps.

[Tinkerer]

Hi Tinkerer, Just noticed your new thread. I had a similar problem. Unregulated DC power supply with a 2000uf cap on the circuit board. Adding a 10uf ceramic in parallel on the board, helped a lot. Looking at your scope traces your are doing something I'm not familiar with. Balanced coil, DD,OO? If the coil is balanced why the signal when Tx on? Are you measuring the Rx coil or amplifier out? Why no large signal when the Tx decays. I'm impressed that you could sample so soon. Thanks

Hi Green,

a picture is worth a thousand words.
This picture was to show my problem with the oscillations on the battery minus lead. The feedback has been helpful and little by little I am getting the noise down. Like most of the time, there is no simple solution, because the noise is made up of several causes or sources.
So, lets go step by step and find all the solutions.
Attached are 2 pages from the pdf file that CHET kindly sent. They show us the oscillations caused by the step function and the recipe to solve it. The way I understand it, the combination of the 2000uf cap and the 10uf, X7R cap fall into this category. To get deeper into it, I guess we need to look at the specific characteristics of the actual capacitors used. We should do that.

Now to your other questions:

None of this setup is of a traditional PI.
The coil balancing is done in a totally different way. With the traditionally used metal detector coil balancing, once the coil is cast, there is little we can do to re-balance if the balance is upset, by extreme ground condition or physical shifting.
This coil, of 1 meter diameter is being balanced in the lab, for testing target response, and re-balanced on site for the conditions encountered in the field. Initially we were doing this with a 25 turn pot, then with a digital pot controlled by software, now I am working on a more robust solution, as the digital pot has limited voltage parameters. Suggestions for solutions are welcome.

Why is there a signal with TX ON, with a balanced coil?
When this picture was taken, the Flyback was about 500V and the peak coil current about 10A. The balanced coils reduced the Flyback to near zero V, but alas, nothing is perfect in this world, so there are all sorts of parasitics and noises, that make up some 200mV of noise on the RX signal trace that should be a perfectly straight line.

If I remember right, this picture was taken at the preamp input.

Why no large TX coil decay? This was my best shot at building a large coil. It had a self resonant frequency of about 1300kHz. Of course I learned something more by building this coil, so the next one can be a bit better.

Don't pay too much attention to the sampling so soon. As mentioned, this is not a traditional PI. It is a machine meant to sample and recognize targets with TC's from 500 ns to 10us.

I am always open to suggestions and welcome any joint effort to solve problems and improve out joint knowledge.

All the best

Tinkerer

[green]

Hi Tinkerer, Thanks for the reply. I'm curious why the 1 meter coil for targets with TC's less than 10us. Maybe the challenge. I apologize for the questions, like you said it's about the oscillations on the power supply. I can't get mine clean either, but the oscillation does repeat and the integrator does seem to average it out. Good thread, maybe can come up with ways to solve the problem.

[Tinkerer]

Hi Tinkerer, Thanks for the reply. I'm curious why the 1 meter coil for targets with TC's less than 10us. Maybe the challenge. I apologize for the questions, like you said it's about the oscillations on the power supply. I can't get mine clean either, but the oscillation does repeat and the integrator does seem to average it out. Good thread, maybe can come up with ways to solve the problem.

1 meter coil is for deep targets, like 1 meter deep for small (relative) targets and 2 meter deep for large (relative) targets. The targets could be voids in the ground or different conductivity of the ground/soil itself.
This machine is to measure the characteristics of the soil, not metallic targets. However, in the lab I often use metallic targets of different TC's to test the system.