VHF interference

Go To Last Post
31 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,
I have a circuit which has a ATMEGA64 and about 17 LEDs. The micro is reading an analogue voltage from a pot and displaying that voltage using LEDs on board by turning various LEDs on or off. I am turning the LEDs on and off using transistors so the micros output current limits aren't exceeded.

The design works perfectly fine doing exactly what it should. No problems.

THE ISSUE IS WITH VHF INTERFERENCE. I had a handheld VHF radio transmitter. As soon as I press the transmit button it basically screws up the whole circuit. The voltages get affected all over the place and nothing seems to work.

I am thinking I need to filter the noise induced by the radio. But HOW? I have tried screened cables for connections. They DON'T work.

Help, Ideas??

Thanks

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Don't use it close to the transmitter?

Leon Heller G1HSM

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Put the PCB inside a metal box.
Filter the incoming power leads.
If the LEDs are mounted on the PCB and show through holes in the metal box'es lid you may be OK, but if there are long wires going to them then each of the leads will need a filter, too.

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Leon Thanks. I was thinking more along the lines of an actual Engineering solution.

JC that sounds good. I thought of that too. It does have long wires going to it. Which will pick up interference. So I do need a filter. But filter as in just a simple capacitor to filter the noise down to negative or...?

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Are the LEDs individually controlled, and either full on or full off, or are they either scanned or PWM'd?

Either way you need a filter to let the DC, or ___ kHz PWM signal through in one direction, while blocking the 148 MHz / 220 MHz (?) from passing in the reverse direction.

Not something I could draw up off the top of my head... Do you have access to an ARRL Handbook?

JC

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

they are individually controlled full on. Yes i figured i need a filter to let the DC through and block the induced noise.

ARRL handbook. NO. Not yet. But if there is something in it that is going to set me straight then I can definatly find one. So you could tell me what am I looking for and I will start digging.
Thanks

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ferrite beads on power lines, any avr pin configured to be an input, and all the rest of the pins in that order.
Hope you have a lot of ground plane area.

Rick

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Leon Thanks. I was thinking more along the lines of an actual Engineering solution.

Avoid or mitigate. Leon went down the avoid route, seems you want to mitigate!

Basically as the others have suggested - ferrite beads, capacitors on the inputs (try 100pF), metal enclosure. Anything with a length of wire is an antenna - you need to add filtering so that the wire looks like a short circuit to RF.

Mobile phones tend to be worse - they pulse the rf as well.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Use feed through capacitors ( I think if You search for TUSONICS You will find a range of components suitable for the task to hand)

here... http://www.google.com.au/url?sa=...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You should protect first of all input and Vcc circuitry. This can be achieved by thorough shielding and correct ground routing - and one will not work without another. Thus get prepared for serious redesigning.
Happened to read John R. Barnes book on Robust electronic design (translation) - a must read for such a case.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you desire more than guesses, show us a schematic, and high resolution photos of the board.

Immunity is not difficult when done right :-) Watch out for 'unimportant' details.

/Kasper

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Your problem is your led flasher is radiating spurious radio frequencies. The squareness of the squarewaves is the problem. If you can 'slow down' the rise and fall times, there is less radiation. One solution is turn on the clkdiv8 fuse and run the avr slower. Adding some Rs and Cs to the outputs will round out the corners on the squarewaves and solve a lot of the problem.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Other way around Bob. His VHF transmitter is interfering with his LED flasher ... or are you just being your inventive self? :lol: Not enough work on the bench perhaps; nothing on TV; don't have a facebook or twitter account?

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You have been given some sensible ideas which should work if implemented properly. The problem is that there is no single magic bullet and the "mitigation" really has to be designed in at the start rather than tacked on.
Good EMC comparability comes is an electronics art in itself.
Some of the really old ARRL handbooks (pre 1950 when semiconductors did not really exist) cover the material as well as the latter editions.
You need to consider what happens at HF & UHF as well as VHF! You probably need radio in the blood to be really effective at the game.

Charles Darwin, Lord Kelvin & Murphy are always lurking about!
Lee -.-
Riddle me this...How did the serpent move around before the fall?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Somehow I read the problem the wrong way around. Oh well, my advice was good for reducing emanated radiation I think.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Many of the tricks to reducing radiated RF also help to reduce received RF. Probably the easiest to implement is a metal box, and a capacitor to ground on all wires entering the box. If that's not practical, a low pass filter of a 100 ohm series resistor and a 0.001uF ceramic capacitor on your power and input wires might help.
In any case, you're going to need a good ground to connect those capacitors to. A ground plane is best, but short, fat ground traces on a circuit board might help.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've got it. The rf is coming in the analog part. I bet adding a .1 from the pot wiper to ground will help most, and also across the pot hi side to ground, and aref to ground.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

"Feedthrough caps" are excellent in conjunction with a metal box. Here is an example just to show you what it looks like - this one is quite expensive.

http://search.digikey.com/script...

Do NOT use a cheap sheet-metal box. Use a good die-cast one. Why? It turns out the the sheet metal ones can be quite RF-leaky. Any place with a joint results in currents that flow along the joint. That happens to couple currents inside, very nicely. The die cast boxes have a single joint around the lid and it can be very RF-tight.

Jim

Jim Wagner Oregon Research Electronics, Consulting Div. Tangent, OR, USA http://www.orelectronics.net

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi all,
All the suggestions are great. The simple and most effective seems to be caps on the wires connected to the board. I have tried that and that has made it 'almost' perfect. Not much of an affect from the radio any more. I tried .1u and 1u. They seem to be working similarly well.

I have a question though. You all seem to be saying, caps to 'GROUND'. What I have is caps from say the slider, or the Pot high side down to 'NEGATIVE' and not 'GROUND'.

I believe what this is doing is filtering most of the noise down to 'negative'. By GROUND do you mean GROUND or are you using this as a term for 'NEGATIVE' from the power supply or battery.

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ground is usually ment to indicate signal ground.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My lab supply has 3 banana jacks... + - and GND. If you have opamps somewhere and really need +5V, -5V and gnd, then gnd and negative are two differnt pins. If there only is +5V and gnd, then you would probably tie - to gnd, and I can see that in this case, one might consider that supply negative and gnd were at the same potential.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ignoramus wrote:
Ground is usually ment to indicate signal ground.

sorry for sounding dumb but by that you mean NEGATIVE ya? not the GROUND from Live, neutral GROUND??

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
My lab supply has 3 banana jacks... + - and GND. If you have opamps somewhere and really need +5V, -5V and gnd, then gnd and negative are two differnt pins. If there only is +5V and gnd, then you would probably tie - to gnd, and I can see that in this case, one might consider that supply negative and gnd were at the same potential.

GOT IT. Thanks

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

The wall socket green wire ground does go to the ground rod out by the meter. BUT after the power supply transformer, you have +5V VCC and GND and there is no connection from the power supply ground and the wall socket ground.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
The wall socket green wire ground does go to the ground rod out by the meter. BUT after the power supply transformer, you have +5V VCC and GND and there is no connection from the power supply ground and the wall socket ground.

I understand. But the thing is this is technically WRONG terminology though isn't it? When you call something Ground, it is possible for someone to assume that it is at ground potential. But these DC negatives that are often called GND can be on 'above earth' systems which means they will be sitting above the solid Ground.

So I think if it's the negative, it should be called that. And Ground should only be used for Ground potential.

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I guess ground is like dB. You need to specify the reference. The whole thing could be floating at some DC potential with respect to earth ground, and in that case, the power supply negative wouldn't be earth ground, but if you put the black probe on the voltmeter there, you could pretend it was ground. Just keep your other hand in your pocket.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Ground is usually taken to be a point of reference. Since current flows in a loop, it goes back to the point of reference. Negative is also nebulous. Negative on a battery vs negative on a bipolar power supply. So it depends on context. How about grounding in an airplane?

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Kartman wrote:
How about grounding in an airplane?

Exactly, or a boat. On boats AC ground is usually connected to the hull of the ship/boat. I guess similar with aeroplanes.

When talking about reference points for measurements the term 'ground' can be considered just a point. But when looking at a system as whole, the DC negative shouldn't be called that I think. e.g Take a DC Neg point which is floating 130 V above 'Ground'. If that is then labelled GND and a fresh sparky comes along to work on the thing, reads the label GND and touches it with full confidence, he is going to see stars isn't he? or SHE!

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hence my comment ... SIGNAL ground ( with some amplification on signal)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

ignoramus wrote:
Hence my comment ... SIGNAL ground ( with some amplification on signal)

Ya, I can live with that.

If it is meant to happen... it WILL happen!!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My spin on these for beginners (whom I still teach in beginners classes)
Earth or ground .... actually connected to earth via an earth stake.

Common .... a reference point to take voltage measurements from and for connecting one side of an input, output or power supply.

Common earth/ground ... common which is connected to earth or ground.

Shield/screen ... electro/magnetic barrier. For the electric barrier part to work it must be connected to common somehow(like a sreen grid in a tetrode).

The terms are then used interchangeably as if there was no difference.

It is as clearly understood as voltage, voltage drop, potential difference & EMF (not).

Charles Darwin, Lord Kelvin & Murphy are always lurking about!
Lee -.-
Riddle me this...How did the serpent move around before the fall?