0
\$\begingroup\$

I have a 9V circuit to operate latching relays. Because there are instances where two independent DC power supplies meet briefly, I decided to implement schottky diodes in the circuit to prevent them from affecting one another. Initially, I used IN4004 blocking diodes (https://docs-apac.rs-online.com/webdocs/14f5/0900766b814f5b51.pdf), but took them out because I realised they could only carry 1A of current. I subsituted them with SBYV27-200-E3 switching diodes (https://docs-apac.rs-online.com/webdocs/14ae/0900766b814ae442.pdf). I understand that switching diodes are used in instances where there are rapid and frequent switches in the power supply in both directions i.e. AC, however I have used them since they were already available and could carry up to 2A of current (vs 1A of IN4004). When I used my relay to disconnect a part of the circuit (9v measured before disconnection), I found it unusual when I measured around 2V in potential difference (with reference to GND) upstream to the diode at the disconnected part of my circuit. I was puzzled by why there was 2V when that part should have been disconnected by the relay, and any reverse current blocked by the diode. I then took those switching diodes out and switched back to the IN4004 blocking diode. With these, I measured 0V when the relay perform the same disconnection, showing that the circuit was effectively cut off by the relay and no reverse current was allowed through the blocking diode.

Thus, my question is why are my IN4004 diodes able to effectively block all current, while my SBYV27 diodes seem to allow some reverse current through? I want to clarify it this is indeed the case and whether a switching diode is similar (or not) to a blocking diode in blocking reverse current.

schematic

simulate this circuit – Schematic created using CircuitLab

I added a schematic. This schematic only has one DC source, but I tested the voltage measurement on this as well. Circuit maker does not have the exact relay model I am using so I combined two of them here. When one the coil is activated, one circuit closes and the other opens, and vice versa when the reset coil is activated. My circuit might appear unusual but I've designed it to work such that the coils are operated by the very same current that runs through the switching circuit it controls.

\$\endgroup\$
  • 1
    \$\begingroup\$ 2V of current?????? \$\endgroup\$ – Andy aka Feb 1 '18 at 12:34
  • 2
    \$\begingroup\$ Show your schematic. \$\endgroup\$ – Andy aka Feb 1 '18 at 12:40
  • 1
    \$\begingroup\$ I "do" electronics for more than 30 years, heck, I even have a Master's in electronics. Yet I still am left guessing what you mean by "blocking diode". A diode is a diode and usually it is in forward mode or it isn't. You really need to show a schematic so that it is clear how you're using the diodes. There's a schematic editor available when you edit your question. \$\endgroup\$ – Bimpelrekkie Feb 1 '18 at 12:50
  • 1
    \$\begingroup\$ OK, there's a circuit now yet I fail to see what it is supposed to do. What is your goal with this circuit? The many wires crossing does not help to see what's going on. Don't hesitate to place diodes "upside down" if that enables a more straight (short) connection (example: D3, D5). The voltage source is 1 V, is that correct? I have yet to see a relay able to work with 1 V. D3 and D5: flyback diodes, OK but D1 and D2, why are these needed? I smell overcomplicating things. \$\endgroup\$ – Bimpelrekkie Feb 1 '18 at 13:22
  • 2
    \$\begingroup\$ ... and use the GND symbols to eliminate wires and make it very easy to see all the ground-referenced parts on the schematic (although it won't make much difference in this case). \$\endgroup\$ – Transistor Feb 1 '18 at 13:23
1
\$\begingroup\$

Schottky diodes tend to have lower forward voltage at the same current than similar conventional diodes, but also tend to have considerably higher reverse leakage. One follows from the other if you look at the Shockley diode equation.

At high junction temperature you can have VERY significant reverse current flowing in a large low-voltage Schottky diode (tens or even hundreds of mA).

Usually a 1N400x or even a 1N4148 (200mA rated diode) is fine for the flyback diode on a small relay coil. It only sees the coil current and only briefly.

\$\endgroup\$
  • \$\begingroup\$ I read that it is not advisable for two independent DC power supplies to be connected together in the same circuit as they can compete and destroy each other. To overcome this, diodes can be used so that the current coming out from each DC source do not contact the other source. As an amateur hobbyist, I was wondering if the reverse leakge current from the Schottky diode (from the other DC source) could be significant in any way in causing harm to an upstream DC power supply? Or would that not matter since the overall potential difference is overwhelmingly in the forward direction? \$\endgroup\$ – Craver2000 Feb 1 '18 at 13:55
  • 1
    \$\begingroup\$ Diode OR logic using Schottky is fine, or Silicon and now dual FET IC's for low drop voltage in batteries. For line voltage Silicon tends to be cheaper and higher PIV \$\endgroup\$ – Sunnyskyguy EE75 Feb 1 '18 at 14:11
  • \$\begingroup\$ Generally a bit of reverse current won't cause problems but it's really case by case and I can think of many (perhaps pathological) exceptions where it could violate specs or cause problems. \$\endgroup\$ – Spehro Pefhany Feb 1 '18 at 15:51
1
\$\begingroup\$

The SBYV27 diode has a reverse leakage current of 0.4 uA (at 10% of its maximum reverse voltage and at 25 degC). See figure 4 in its data sheet.

So it could be leaking maybe 0.4 uA through D3 (then via the coil) and through D1 to your open circuit measurement node. If you are using a multimeter with 1 Mohm input impedance you could measure 0.4 volts due to 1 uA flowing.

If you look at figure 5 in the 1N400x data sheet from Vishay you will see that at about 5 volts (reverse) the leakage current is about 20 nA at 25 degC. That current into a multimeter of 1 Mohm input impedance would produce a voltage of 20 mV.

\$\endgroup\$
0
\$\begingroup\$

What's all this stuff about reverse diode leakage?

Note above diode reverse voltage leakage is 0.5 uA typ at 25'C and curve is fairly flat (i.e. its a really unstable constant current source).

In general, reverse leakage current is fairly flat (until zener voltage) but temperature dependent (rise with T) and is higher for diode rating and higher with faster switching silicon diodes and much higher in lower voltage Schottky diodes.

When relay coils release, the voltage reverses due to reverse EMF and the diode clamp current is initially the same as the coil as it decays down with somewhat a linear L/R=t time constant.

Since the question is about latching relays, we know they take much more power initially to switch from a magnetic resting position than a non-latching but can save energy over time.

We know everything has thermal mass and thermal resistance and a time constant to reach critical temperatures and we need to keep safe margins and not choose parts operated at their Absolute Max limits for reliability considerations.

Check your assumptions

  • 1N400x is rated for 30 A for 1 cycle 60 Hz and not 1 AMP.

The old workhorse 1N400x is rated for 1 A continuous and 30 A surge for 1 cycle.

(The x suffix indicates peak inverse voltage (PIV) in units of 100 hundred volts.

Like all semiconductor switches can handle much higher peak currents than the sustained current. This is often described in several ways;

  • I vs t with an inverse characteristics with graphs (trending towards deleting these curves in datasheets for common legacy parts
  • Maximum non-repetitive current for 1 cycle (based on either 10 ms sine or 16.7 ms (60 Hz) sine)
    enter image description here

The other part of this question is about a fast recovery diode which has more leakage (0.5 uA) than a 10 Mohm DMM so it reads a pullup voltage when reverse biased (no big issue with power drain here, I expect).

schematic

simulate this circuit – Schematic created using CircuitLab

enter image description here

Conclusion

1N400x is rated for 30A max 1 cycle 60 Hz and is likely to work perfectly for your miniature latching relay.

\$\endgroup\$

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.