Simple Overvoltage Protection

I was asked lately to design the simplest possible circuit for mains overvoltage protection (mains supply 220V, 50Hz).
I attached its Kicad schematic, A34S_P380V_11_simple.pdf.

Notes:
R1= 220Rh (R means Ohm, and h means half-watt type)
C1= 220n (rated not less than 600V)
CP1= 470u (rated not less than 35V)
K1= small relay (24V-1200R), it turns on when the mains voltage exceeds the high limit.
R2 and RV1 to set the high voltage limit (decreasing their sum resistance increases the high limit)

It worked well in protecting an SMPS (when Vac > 250V, it could be lowered). But, unlike the previous linear protection (link below), it just turns off the SMPS completely (till the mains voltage is turned off then on again).

AC Overvoltage Protection for SMPS | Electronics Forum (Circuits, Projects and Microcontrollers) (electro-tech-online.com)

Have a nice day,
Kerim

So do you have a question, or are you just posting what you have done?

crutschow said:

So do you have a question, or are you just posting what you have done?

Click to expand...

I am sorry if projects are not supposed to be posted here without a question. I simply expect hearing possible comments, because nothing is perfect

Kerim said:

I am sorry if projects are not supposed to be posted here without a question. I simply expect hearing possible comments, because nothing is perfect

Click to expand...

Posting that here is fine.
It just wasn't clear, the reason for your post

You have some pretty substantial Pdiss in some components :




Regards, Dana

A Varistor can be useful for the same, MOV

The MOV good for short term transient protection, but long term sustained
overvoltage not appropriate (depending on power levels of course).


Regards, Dana.

danadak said:

You have some pretty substantial Pdiss in some components :

Regards, Dana

Click to expand...

I wished you were more specific.

For example, let us calculate Pdiss of R1, 220 Ohm 1/2W, in series with C1, 220 nF 1000V:
Zc = 1/wC = 1/2/pi/f/C = 1/2/3.14/50/220e-9 = 14476 Ohm
Ic_max = Vrms_max / Zc = 400 / 14476 = 0.027632 A
Pdiss_max = Ic_max^2 * R1 = 0.027632^2 * 220 = 0.168 W
Actually, it is a bit lower due to the DC opposing voltage, on CP1 470 uF, in series with R1 and C1.

Did I miss something?

Regards,
Kerim

My screwup in sim, used 1 Khz, not 50 Hz. my apologies. Here is 50 Hz :





I ran sim out to a second, seems like your relay current causing a lot of latency,
eg. time V exceeds trip to time load disconnected......In our neighborhood its also
aggravated by pole breaker cycling when coming back on, in many instances several
cycles over a minute or so. Note I looked at several datasheets, not exactly sure what
your relay pullin current is....? Also I estimated Relay L as 1 mH ???? Although I think that
unimportant at 50 Hz.




Regards, Dana.

danadak said:

My screwup in sim, used 1 Khz, not 50 Hz. my apologies.

Click to expand...

Such situation may happen in best families. And it happens to me once a while

danadak said:

I ran sim out to a second, seems like your relay current causing a lot of latency,
eg. time V exceeds trip to time load disconnected......

Click to expand...

Good Remark. The worst case is when the initial voltage of the mains is 246V (assuming the set limit is 245V). The lag time before disconnecting the load would be about 2s! Even for 380V, it is 0.5s. This is the time that 470u capacitor, CP1, takes to reach about 16V (the lowest DC voltage of my small 24V relay).
Now, I wonder if the circuit works without CP1. As you know, the relay coil has an inductance which also smoothes the rectified current. But this needs a test in real because the coil resistance only could be known/measured while here its inductance determines the result(s).
The main drawback of this simple solution is that it doesn't have hysteresis which needs suitable positive feedback. So, I was fortunate that the SMPS (the load) turns off normally (without oscillation) even at the limit voltage (when the minimum relay pulling current is reached).

For instance, CP1 is usually charged already near the lowest voltage of the relay (when the mains voltage is below limit, as 220V). Anyway, the real test of this circuit is running now at some customer's houses

Regards,
Kerim

====================================

IMPORTANT: (in general):
I am afraid that testing circuit on mains is not for newcomers to electronics. The AC high supply on the test table should be isolated from the mains lines by 1:1 transformer (primary and secondary coils are electrically isolated). When one of the two floating terminals (secondary coil) is touched, it becomes at earth level. It is safe as long the two secondary terminals are not touched simultaneously. Adding a fast fuse (or equivalent) at the primary side protects the wires in case of a short circuit during a test. Also adding an autotransformer helps increasing the AC voltage gradually when working on the first test(s) of a new circuit.

Thanks a lot. I can see the schematic only. If you have made any PCB, you can share that at

Here are two PDF files, one for PCB and the other for assembling: