3
\$\begingroup\$

I need to derive a low-voltage galvanically isolated 50 Hz signal from 230 V mains. I'll use an opto-isolator for this. I'd like to use IC sockets for all ICs in this personal one-off project.

Opto-isolators are typically suitable for mains voltage, being rated for thousands of volts of isolation between the diode and the transistor. But are DIP IC sockets suitable for mains voltages?

I know IC sockets are made of plastic, but I preferred to ask rather than possibly compromise isolation. I don't think there are any datasheets I could check either.

\$\endgroup\$
  • \$\begingroup\$ "I don't think there are any datasheets I could check either." So then find a vendor that provides one. \$\endgroup\$ – Ignacio Vazquez-Abrams Dec 11 '14 at 16:33
4
\$\begingroup\$

I just grabbed one socket at random on Farnell UK, and got this drawing datasheet. In it is this little bit of text:

ELECTRICAL:

  • Contact Resistance: 10 milliohms Max
  • Contact Rating: 3 Amps
  • Capacitance: 1.0pF per MIL-STD-202, Method 305
  • Insulation Resistance: 5000 Ohms Min @ 500 VDC Per MIL-STD-1344, Method 3003.1
  • Dielectric Withstanding Voltage: 1000 Volts (RMS) Per MIL-STD-1344, Method 3001.1

So I would conclude from that two things:

  1. Datasheets with the relevant information in them are available if you look, and
  2. Yes, that one would be OK for mains at low currents.

Besides, you'll have a resistor there too, right, which would mean you'd end up with maybe 2V across the LED anyway?

By the way, MIL-STD-1344A Method 3001.1 states:

  1. PURPOSE. The purpose of this test is to prove that a given electrical connector or coaxial contacts can operate safely at its rated voltage and withstand momentary overpotentials due to switching, surges, and other similar phenomena. The dielectric withstanding voltage shall be established as 75 percent of the minimum breakdown voltage of the connector or coaxial contacts. It is suggested that the operating voltage of the connector or coaxial contacts be established as one-third of the dielectric withstanding voltage.

From that I can tell that the maximum normal operating voltage of that socket would be 333V RMS.

For clarification on the insulation resistance, Method 3003.1 states:

  1. PURPOSE. The purpose of this test is to establish the methods and procedures to be followed in determining the resistance offered by the insulation materials and the various seals of a connector or coaxial contacts to a direct current potential tending to produce a leakage of current through or on the surface of these members.

So it's intentionally making the insulation "break down" and testing the resistance during that catastrophic event.


Addendum: Don't sue me if I'm wrong.

\$\endgroup\$
  • \$\begingroup\$ Ah, I should have checked another shop before posting a question... The local electronics shop doesn't display the manufacturer or link to a datasheet for IC sockets, and Google wasn't very helpful as it thought I was searching for mains sockets. Thanks! \$\endgroup\$ – ntoskrnl Dec 11 '14 at 16:44
  • 1
    \$\begingroup\$ Is that saying that there might be 100mA of leakage between pins at 500V, or am I misreading something? Something that behaves like a 5K resistor at 500V would be dissipating 50W. \$\endgroup\$ – supercat Dec 11 '14 at 16:46
  • \$\begingroup\$ At 500V DC, yes, but you're not working at 500V DC. \$\endgroup\$ – Majenko Dec 11 '14 at 16:48
  • 1
    \$\begingroup\$ @supercat that's probably a typo. I just checked the data sheet for a similar device and that's spec'd as 5000MΩ minimum, not 5000Ω. See assmann.us/specs/Axx-LC-TT.pdf \$\endgroup\$ – TheNoonMoose Dec 11 '14 at 17:29
2
\$\begingroup\$

No.

DIP pitch is 10th inch correct? But taking into account the pads the actual clearance between pins is painfully low.

Check out the IPC regs for creep distances.

\$\endgroup\$
  • \$\begingroup\$ The optos I looked at are standard DIP6 packages, yes. I don't think the PCB pads are different for IC sockets and directly soldered ICs. \$\endgroup\$ – ntoskrnl Dec 11 '14 at 17:46
  • 1
    \$\begingroup\$ The high voltage shouldn't be between pins that are 0.1" apart. The relevant distance is between the LED side and the transistor side, edge to edge of tracks/pads. The problem is that the socket manufacturers usually only give one voltage/insulation rating, which we have to assume is for the 0.1" pitch. \$\endgroup\$ – Martin Dec 11 '14 at 19:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.