[SOLVED] Shielded cable with no metal case

[brain-dead]

Hi Everyone,
I have looked through many threads and researched on the net and have not found a clear answer to the following question.

How to ground a shielded cable when the PCB is enclosed in a plastic case.

My design is a high gain microphone amplifier, the board has a solid ground plane underneath and signal traces on the top. The circuit basically contains two quad op amps, a filter then multiple gain stages. The circuit works great until I put a long length of cable for the microphone > 1m. For testing I soldered an electret mic to the board with a short length of twisted pair, all was fine. I need to keepp the signal free from EMI/RF as the gain is so high

Alot of answers say ground the cable to the case at the point of entry, but I have a plastic case.

I am wondering which option to use:

1. Leave the cable unshielded
2. Connect the shield of the cable to 0V on the pcb side and not the microphone
3. Connect the shield to the microphone 0v and not the PCB side
4. Connect the shield to the PCB 0v, however place an inductor in series with it
5. Have a nervous breakdown and write a letter to someone requesting that RF transmissions be reduced

The problem is RF pickup, mostly intermittent, I performed two experiments:

1. Stuck a spec analyser round the board with the short twisted pair mic < 6' length
No RF noise anywhere

2. Stuck a long cabled mic > 1.5m and then tested with the analyser
Intermittent RF of all things weird and wonderful. Noise from 100MHZ upto about 900. I have one of them weather stations outside that transmits weather data to a console in doors, it is on 868MHz, that got picked up very well

Hopefully someone can shed some much needed light on this cold, dark and damp situation for me.

Thanks in advance

Brain-Dead

 

[pstuckey]

Have you tried a clamp-on ferrite? Something like a Laird type 28 would work well over your frequency range.

Digikey has kits of different sizes, see for instance K-403 or K-404.

 

[sky_123]

I sometimes find this link useful
(click on the check-marks in the yellow, cyan and purple table boxes).

Edit: updated with a better URL.

 

[pstuckey]

View attachment Cybershield - Metallization of plastics.pdf

Another thing to consider, at least before trying your fifth option is to metallize the plastic. The attached has several options for that.

 

[godfreyl]

To start with, use shielded cable and connect the shield to the PCB ground. The ferrite, metalized plastic etc are optional extras if there's still a problem.

 

[brain-dead]

Thanks for the help everyone. I will try and screen the cable to ground, then if that doesn't work I will add the ferrtite. I have some long pcb traces, is it a good idea to place a decoupling capacitor at either end of the trace?

 

[brain-dead]

Hi everyone, thanks for all your help. I have finished my PCB and everythings is much better with the cable screen connected as close as possible to the GND on my PCB, in this case where the battery is connected to the PCB. I have also clamped on a ferrite and all is well. Thanks once again