Intermittent reset of 16C6440 - possible noise issue

[Christopher4187]

I'm at a loss here. I have a 16C6440 that resets sporadically on *most* of the equipment it is installed on. It's a noisy environment but seems like the filter is doing somewhat of a good job keeping the noise at a lower level. I've attached two pics - one of the schematic for the filter and one of the measurements taken between VSS and VDD. The peak-to-peak ripple is approximately 180 mV and the frequency is approximately 233 kHz. When using a DMM, the voltage is steady at 5.223 VDC.

I don't think the ripple voltage is a problem but what about the frequency?

[Christopher4187]

Still struggling with this one. I had the PIC reset today (by the way, the schematic says 16C6440 but this doesn't seem to exist) and the voltage never changed. I'm still leaning towards a noise issue but I can't really prove it. Also, pins 15 and 16 (RC0 and RC1) were left floating and I wonder if that is an issue. I also noticed there there isn't a capacitor after the DC/DC converter but again, not sure if this is a huge problem or not.

I think I may have an issue with common mode noise so I'm going to take different measurements tomorrow without the ground clip on. My oscilloscope is battery powered but I wasn't sure if this would be an issue or not.

Can anyone help me out here?

[Christopher4187]

One more question. Does a 40 pin DIP PIC need a .1uF cap across both sets of VDD and VSS pins?

[shahidali55]

A 0.1uF capacitor accross VDD and VSS is always recommended. Just make sure it is located real close to the PIC.
Also, you can add a 5.1V zener diode in between VDD and VSS like below:

[Ramius]

Hi Chris!
The short answer is yes, .1mfd capacitors are very necessary. The techniques for dealing with noise vary based upon the environment and signals in use. If you wish to "dump" your 233Khz then a capacitor of a value based upon frequency in hertz time the capacitance in farads divided by .159 aka Xc= .159/FC. The best choice of capacitors are the tantilum type as the act very much like a diode. Other considerations are things like how your grounds are handled. Typically a "single point" grounding system is best to avoid ground loops which will induce noise and stray voltages. Another tatic is to use a double sided PC board where all unused areas on the board (both sides) become a "ground plane". Lastly is to completely shield the PIC board in a metal box and use what are called "feed thru" capacitors. This of course is for extreme conditions in high emi and rfi environments. It is a good practice to make all ground traces at least twice the size of signal and power traces on a PC board. A "sprinkling" .1 mfd, 10 mfd, 100 mfd, wherever possible will also help. Not knowing what type of 5 volt regulator you are using I would check to ensure you are not using the "switching" type regulator and stick with the known "standard" or analog regulators such as the LM7805. Hope this at least provides some ideas and directions. Best, Ed