Jump to content
Sign in to follow this  
Shree

Mind Boggling Emi Problem With Pic

Recommended Posts

Hello everyone,

I came across this problem when I was making a small high frequency inverter of upto 50W (extremely low cost; system cost approx less then $10) for rural areas in india, where there is no electricity. The battery for this inverter would charge from a solar panel and provide energy to the load (CFL: Compact Flouroscent Lamps) in the night. I used PIC16F684 and deployed its PWM module. everything worked fine in the initial stages. But when I created a PCB, all the problems started. Now the problem is that whenever I switch On the supply, the controller keeps on reseting until I dont place my hand on the negetive/negetive track of the battery. If I place my hand on the negetive, it starts working fine. I tried everything from using filter caps to tranzorbs across 7805. But nothing worked. Lastly I tried taking a wire from earthing and directly shorted it to the negetive of the battery. Voila!!! Now there were no problems! But now my problem is that I am not sure I will have the luxury of 'Earthing' for my product as the electricity condition in rural india is more then pathetic. Can anybody suggest me what can be done to resolve this issue?

For Info:

The inverter voltage is created by applying 12V DC to the central tap of a ferrite core transformer, of which the ends of winding are connected to the ground through MOSFETs. So the MOSFETs act as switches (PUSH-PULL Config) and hence create a stepped up high freq square wave (40Khz) of 200V AC approximately at secondary of the transformer. This square wave is rectified by fullwave rectifier and used to power load (CFLs, which have a ballast in built, so they work on DC as it is). So we get a output of 250V DC . The switching of MOSFETs is done by PIC micro and the voltage conversion(5V controller signals to 12V MOSFET gate voltage) is done by using transistors and a TOTEM POLE configured transistor stage is further added to ensure proper switching. The pic micro is 16F684 and works on 8Mhz internal clock. PWM module used in PUSH-PULL mode is used to generate gate voltages and a dead band delay of 1.5uSec is programmed in between the switching of 2 gates. MCLR, Power-Up Timer, Brown-Out detect all are turned off.

I am utterly confused as what should I do now to eliminate this noise issue. I tried re-designing the PCB thrice. I hope somebody can guide me through this.

 

Thanks and Regards

Shree

Share this post


Link to post
Share on other sites

Shree,

I came across this problem when I was making a small high frequency inverter of upto 50W (extremely low cost; system cost approx less then $10) for rural areas in india, where there is no electricity. The battery for this inverter would charge from a solar panel and provide energy to the load (CFL: Compact Flouroscent Lamps) in the night. I used PIC16F684 and deployed its PWM module. everything worked fine in the initial stages. But when I created a PCB, all the problems started. Now the problem is that whenever I switch On the supply, the controller keeps on reseting until I dont place my hand on the negetive/negetive track of the battery. If I place my hand on the negetive, it starts working fine. I tried everything from using filter caps to tranzorbs across 7805. But nothing worked. Lastly I tried taking a wire from earthing and directly shorted it to the negetive of the battery. Voila!!! Now there were no problems! But now my problem is that I am not sure I will have the luxury of 'Earthing' for my product as the electricity condition in rural india is more then pathetic. Can anybody suggest me what can be done to resolve this issue?

Sound like an interesting project.

I'm not an expert in such matters, but consider the following:

1) Make sure than all inputs that are unused are tied to 0V or 5V through a pullup resistor, ie nothing is left floating.

2) Make sure the PIC has appropriate decoupling capacitors fitted close to the device, and same around you 5V regulator.

3) Consider putting a metal screen cover over PIC and underside PCB to protect it from the radiation, bonding this to 0V,

 

Regards

Dave

Share this post


Link to post
Share on other sites

Hello Dave,

Thanks for that piece of advice. I have done most of the things you mentioned. Would have to try to sheild the controller though and yes its a real nice project, although there are been similar products existing here in the market, but they either have lot of failure rate or cost above $20. So if I am been able to do that, it would help ppl a lot. The cost of solar panel is also a cause of concern. Thanks alot again for your advice.

Regards

Shree

Share this post


Link to post
Share on other sites

Hi,

Not directly related according to your description, but I had a similar problem with a PIC project. It had a 16F877a in a TQFP fairly close (~5cms) to a couple of relays switching 230VAC. Mostly it worked fine, but occasionally it was clear the code had gone berserk!

I eventually discovered that an earthed aluminium screening can cured the problem, presumable by protecting the PIC from EMI produced by arcing at the relay contacts.

You may consider doing the same, as you may have HF EMI emanating from any inductors or transformers?

Share this post


Link to post
Share on other sites

Your content will need to be approved by a moderator

Guest
You are commenting as a guest. If you have an account, please sign in.
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoticons maximum are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

×