Radiated Electromagnetic test fails

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

Hi all ,

The circuit should sense 2A and 4A current , so a 20m shunt is used

When EMI enters (20MHz-1GHz) , instead of 4A detection , 2A is detected

Load is car lamp

It seems the shunt is acting like an antenna , it gets induced by EM field and effects the ADC readings

I added a 100n capacitor on shunt and the board had better reaction

I removed the D6 diode and it seems there is no problem now

How can I protect the shunt from EM field ?

 

Thanks in advanced

 

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

Ferrite beads?

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

I'd be trying some ferrite beads. Like thai love beads, it pays to place them carefully.

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

Have you tried decoupling pin 1 to ground?

#1 This forum helps those that help themselves

#2 All grounds are not created equal

#3 How have you proved that your chip is running at xxMHz?

#4 "If you think you need floating point to solve the problem then you don't understand the problem. If you really do need floating point then you have a problem you do not understand." - Heater's ex-boss

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

Headlight circuits often have a ground connection near the sockets. Use that for the ground, to avoid the large loop area, resulting from a distant ground at other chassis point. The decoupling cap also essential. Incandescent bulbs generate noise.

It all starts with a mental vision.

Last Edited: Mon. Apr 27, 2015 - 01:54 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Decoupling pin1 to ground doesn't solve the problem

The PCB GND may solve it and I'll make a new PCB , But till then I'm looking for a fast solution

 

ferrite bead :

Would you please explain more , I can't get it

Where to place it , which part number , any application note ?

 

Thabks

 

 

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

Lots of info on the web. Here for example.

 

An excerpt:

 

Greg

Last Edited: Tue. Apr 28, 2015 - 12:02 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

How about a cap across the zener?

 

Imagecraft compiler user

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

Wurth Elektronik have a number of app notes.

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

bobgardner wrote:

How about a cap across the zener?

I didn't test it

 

Thanks for your guidances

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

LOSTISLAND wrote:
It seems the shunt is acting like an antenna

Seems rather unlikely - that's the extremely low impedance part of the system!

 

Far more likely that interference is entering in the high(er) impedance parts of the circuit - around Vout - or in the microcontroller parts (not shown)...

 

Top Tips:

  1. How to properly post source code - see: https://www.avrfreaks.net/comment... - also how to properly include images/pictures
  2. "Garbage" characters on a serial terminal are (almost?) invariably due to wrong baud rate - see: https://learn.sparkfun.com/tutorials/serial-communication
  3. Wrong baud rate is usually due to not running at the speed you thought; check by blinking a LED to see if you get the speed you expected
  4. Difference between a crystal, and a crystal oscillatorhttps://www.avrfreaks.net/comment...
  5. When your question is resolved, mark the solution: https://www.avrfreaks.net/comment...
  6. Beginner's "Getting Started" tips: https://www.avrfreaks.net/comment...
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Trying to solve EMC issues via a forum is fun!

 

Theres jack sh!t filtration on that schematic and it resembles a text book circuit rather than a practical one, automotive just means even more margin of safety and extra crap coupled with equipment and materials you can and can't use

 

Ferrite beads are OK for bringing levels down that extra few db but for me personally it always feels like a bodge, common mode chokes on the board are far superior, more inductance can be inserted, its more predictable and it looks better its better in every way, add a common mode choke a couple of Y caps to the chassis and an X cap across the supply and you have yourself an EMC filter, after this you still need all your standard decoupling caps and for measuring current a differential filter could be a good idea

 

That goes a hell of a long way to reducing emissions but that circuit isn't a full system and there is so much more to consider, immunity is another beast alltogether which measuring 2A instead of 4A sounds like a radiated immunity fail to me but I dunno what standard you are working to?

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

Is your SHUNT a standard wirewound resistor? If so, it's an induction coil, and picks up changes in the ambient magnetic field caused by nearby wiring. You could wind your own, folding the wire in half so that induced current is coming and going along the same path and cancels, or put it straight on the PCB in a zig zag pattern to get the same effect, with a ground plane on the other side of the board. 20 milliohms is not a lot - 2 inches of 50 thou track in 1 ounce copper. If you went for a somewhat higher resistance shunt, the induced interference would be proportionately less.

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

Bignoob, being a vehicle i'd suggest a common mode choke might ve difficult to implement. The chassis is effectively the ground plane. Agreed re solving emc via a forum.

As noted by others:
The exact test is not specified
The shunt might be causing problems
There is zero filtering
No details if there is any relation to frequency - if there was a peak or trough suggests resonance so this can be isolated. If it is across the sweep, then its probably rectification in a silicon junction.
Exact circuit layout- the schematic only tells half the story.
You need to do tests to isolate where the signal gets coupled in.
My guess is the the power in or power out.
How are you measuring the sense voltage? Is this not affected by the test?

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

Ok friends ,

Yes , so many issues are there that must be taken into consideration

I think the simplest solution is utilizing a hall effect current sensor like ACS715 and modifying the PCB

The current is high and ZXCT1008 which needs a metal shunt in this case is a bad choice

The PCB is very bad designed , ADC route is dirty , The GND plane is weak , ...

 

Thanks for your help

 

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

Is the circuit in a metal box?  If not, put it in a metal box.

 

Yours Faradaycageingly,

 

David 

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

The hall sensor is a sensitive analog device. How do you think this will work better? Especially considering you haven't identified the root cause in the first place.
Besides, using hall effect devices in an auto environment presents different challenges.

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

Hi.

if you change your current measure method perhaps help but your problem will not be the shunt.

20mR at you do not make a antenna with that and not problem in the input of your opamp.

But your zener is a pure capacitor and send alt noise directly to your output.

bypass the output and try again. 

if high frequency use 2 or 3 cap in parallel 47p NPO, 10nF +100nF or more with different ceramic to get best cut off.

 

Thierry

Thierry Pottier

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

Metal Box is too expensive

 

The hall sensor mentioned is made for high currents , there is no shunt and I think the frequencies will not have that much effect like ZXCT1008 IC

ZXCT1008 IC is too sensitive to voltage change too but the hall effect IC is sensitive to current change and it's too for away from having a current change up to 1A with the EM fields , is't this true ?

But for a little change on the voltage we have a great misreadings in the output

 

 

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

You've got at least one big antenna loop. This couples energy into the small signal circuits. So don't think the hall effect sensor is immune. So short of using a bimetal strip and gauge linkage, you need to address the root cause.
The hall effect sensor also needs magnetic shielding. Anyrhing nagnetic close by will be measured.

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

8< Metal Box is too expensive

 

How much would a metal box cost?  If you have a working circuit but it is susceptible to EMI then you need to shield it.  This could be a simple metal enclosure or a shield covering the part of the circuit that is susceptible to or generating RF.

 

Of course you can spend a whole bunch of time changing this and that and try to fix the problem but if you have a naked board being irradiated with RF -  "EMI enters (20MHz-1GHz)" - it will be very difficult to remove susceptibility without a simple shield.

 

If your board is in any form of enclosure you could put a conductive coating on the inside of the box.

 

David 

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

Sounds like its for an automotive application. Probably for a taxi. Still no mention of the exact test.

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

As I understand it, ABS plastic can be plated.

 

Ross McKenzie ValuSoft Melbourne Australia

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

Looking at the circuit I can't see that it is acting as a transmitter and the EMI is entering on the power lines and not the short bit of conductor that is Rsense. Why would Rsense pick up EMI that the big ass antenna that is the Vin line is not picking up? You removed the zener and the problem went away, I suspect it offers no protection to your circuit. I also suspect a cap of some sort across Rsense may mitigate your problem further.

 

If I made a mistake, ignore it. it's been a while since I did automotive electronics.

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

Perhaps you just need to put inductors in the power supply lines to filter out the RF?

The largest known prime number: 282589933-1

It's easy to stop breaking the 10th commandment! Break the 8th instead. 

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

Torby wrote:

Perhaps you just need to put inductors in the power supply lines to filter out the RF?

Would you please explain more , which inductor ? is there any guide ?

 

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

Kartman wrote:

Sounds like its for an automotive application. Probably for a taxi. Still no mention of the exact test.

Yes automotive , The test is accomplished with GTEM cell and the board is placed inside the cell

We tested it in the car which was placed in a room too (EM fields present) and the result was the same

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

As to where to place ferrite bead, inductors and capacitors, it helps to have a background in electronics otherwise you're just guessing and wasting expensive test time. I asked was the interference related to frequency - no response. If you can identify if there were peaks and troughs with the unwanted effect, then you can specifically target those frequencies. If not, the problem is the unwanted interference getting rectified in a silicon junction. Thus you need to shunt the high frequencies away. A simple low pass filter in the input snd output should have the desired effect. What value of L and C? That involves some measurement and experimentation. The smaller the L value required the more likely it will be physically smaller and cheaper. Form an experiment and do some tests.

Last Edited: Wed. May 6, 2015 - 01:11 PM
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I asked was the interference related to frequency :

 

I just know that the readings of ZXCT1008 IC decreases dramatically

This is read by ADC

 

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

Best be doing some more research then. Have you tried to replicate the problem outside of the test lab?