Noise Induced on Ajacent lines

Status
Not open for further replies.
J

jnnewton

Member
I have a project where i have 6 24Vdc inputs coming in from a connector to a max6818 switch debouncer that have been causing headaches due to the fact that when one comes on/goes off , it induces a signal in the ajacent lines that is longer than the debounce time. Will lowering the pulldowns from 2.2k help? How far should I go? Any other Ideas?
 
Last edited:

1) lower pull downs until DC dissipation is at limit.
2) Put filter on the incoming offening signals. Place them right at the connector. If you slow their rise/fall time down, they will couple less.
3) Can you put some shield inbetween the sensitive signals and the offending signals? If so, it should help.
4) Re-route the sensitive signals to be far away from offending signals.

There are some ideas to try. Some may be more practical than others for your situation that you only briefly described.
 
5) A resistor, about 1/0 the value of the pullup, in series with each switch. Locate it near the switch.
 
Ok, I am redesigning this circuit, and I wanted to add a few things, including a debounce ic, and an indicator LED and possibly reverse polarity / overvoltage protection.

The data sheet on the debouncer says its input current is 1mA at 15V. I calculated 12mA through the filter (see diagram for calculations. How does this current 12mA relate to the datasheet info (1mA at 15V)? Does anyone see any other problems with the values of the resistances?

How would I (and i don't know if it is worth it) protect the input against someone trying to put 120VAC into it?
 

Attachments

  • Input Circuit.jpg
    377.1 KB · Views: 157

If the path from J203 to U212 is a physically long one (lots of area to couple noise), then place your filters right at the input of U212.

Also, use a ground plane. I dont see any issues with your resistances.

If you want overvoltage protection, put suitably rated TVS diodes on all of your lines and place a fast acting fuse in series with your lines at the connector. The TVS diodes (with a good layout) can limit the voltage to a low/safe level within nano-seconds. If there is a sustained higher current from said overvoltage, the fuse will open removing the condition. The fuse should be rated for the standoff voltage. If the 120VAC you ask about is mains, I would rate the fuse for 250V ( a standard rating)
 
Should i use this same method to protect the power input to the whole board ( also 24V) or is there some reason why people use a simple schottky diode in line or I have even seen a backwards transistor ( hooked up reversed)? I have had two of the boards i built fried b/c someone hooked them to the wall instead of to a 24 v supply!
 
Thanks, I will probably go with 10uF to get 88 Hz Fc (1/(2*R*pi*C). Any suggestions on the power protection?
 
If you use a ceramic capacitor make sure it is X7R or X5R. Don't use Y5V or other types. Electrolytics would probably be fine.
 
For protection you can use a TVS (transient voltage suppresor) diode and a fast-blow fuse. Put a 30V bidirectional TVS zener diode (fast response, they are on digikey) in parallel with the 24V switch lines, and then a fuse between that point and the main power connector. If someone hooks up anything > ~30V the TVS clamps, fuse blows.

**broken link removed**

This will handle the current just fine until the fuse blows.
 
Last edited:
Thanks. I really appreciate the help. I will use X5R's they are avail on digikey in 1206 packages. Just out of curiosity, why X7R or X5R?
 
Voltage coefficient of capacitance. At their full voltage rating they will still show about 80% of their rated value of capacitance. Y5V types fall all the way down to as much as 20% when operated at rated voltage. They're fine in some apps but I always try to avoid them.

**broken link removed**

Once you get above 0.1uF, you can't really get NP0 ceramics and X7R/X5R becomes the preferred type in most applications. You could stick a Y5V in there, just be sure to make it a larger value than you normally would have used.
 
Last edited:

Finally got around to trying this, and the results are in! I built the following circuit, and it protects against 24VDC reverse voltage, but if i hook 120VDC up to it (just rectified mains to try dc before trying 120VAC) it fails. The TVS is no good anymore, fuse blows even at 24VDC correct polarity. Do you know what might be causing the problem?
 

Attachments

  • Power Protection.gif
    10.1 KB · Views: 141
Status
Not open for further replies.

Similar threads

N
  • Question
Controlling 12 watering lines with Arduino
Replies
18
Views
3K
For The Popcorn
Z
Timer on/off delay
Replies
14
Views
2K
Zool
Z
  • Question
5 second 'slow on' circuit
2
Replies
25
Views
3K
shokjok
S
N
  • Question
Light pulse then on
2
Replies
22
Views
4K
danadak
S
  • Question
WAH on Telecaster
Replies
19
Views
5K
ZipZapOuch
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…