voltage spike of AC power line

[[email protected]]

I was having false triggering on my LM393 input when I switch on the
lamp (tungsten or florescent) in my room. After experiment a bit, 2
ceramic capacitors(2200pF) at both LM393 input seems to solve the
problem. Why is there such a "switch-on" voltage spike & what is the
voltage frequency?

Thanks in advance
ck

 

[Tim Shoppa]

I was having false triggering on my LM393 input when I switch on the
lamp (tungsten or florescent) in my room. After experiment a bit, 2
ceramic capacitors(2200pF) at both LM393 input seems to solve the
problem. Why is there such a "switch-on" voltage spike & what is the
voltage frequency?

Thanks in advance
ck

Without the bypasses, the LM393 will respond to sub-microsecond
pulses.

With a 2200pF C on the inputs with circa 10k-ohm impedances, the time
constant is now 20 microseconds.

If the R is a megohm instead of 10K, the time constant with the 2200pF
bypasses is 2 milliseconds.

Does that answer your question about spikes? Or where the "false" in
"false triggering" lies?

Tim.

 

[Marra]

I was having false triggering on my LM393 input when I switch on the
lamp (tungsten or florescent) in my room. After experiment a bit, 2
ceramic capacitors(2200pF) at both LM393 input seems to solve the
problem. Why is there such a "switch-on" voltage spike & what is the
voltage frequency?

Thanks in advance
ck

I think the problem is down to your power supply.
It is important to filter the incoming mains properly.
The transformer will pass through much higher frequencies than just
mains frequency.
Any high frequency noise will pass right through unless you have
sensible capacitors across the power supply.

High value electrolytics are useless as they are quite unductive and
allow the HF through.
You need 100nf or similar across the electrolytics.

 

[John Fields]

I was having false triggering on my LM393 input when I switch on the
lamp (tungsten or florescent) in my room. After experiment a bit, 2
ceramic capacitors(2200pF) at both LM393 input seems to solve the
problem. Why is there such a "switch-on" voltage spike

---
Switch bounce and:

di
E = L ----
dt

---

 

[[email protected]]

---
Switch bounce and:

            di
     E = L ----
            dt

Sorry if I understand it incorrectly.What I would like to know is the
value of capacitance just enough to surppress such spike induce by
switching ON/OFF the AC socket.

ck