Is this relay datasheet incorrect? (Omron G6SK-2)

Question

I'm using Omron G6 series relays, both the non-latching (G6S-2) and and dual coil latching (G6SK-2) for a few PCB designs.

What I didn't know was that some models of relays give their connection diagrams from the "Top" view (as expected) while some give them from the "Bottom" view (annoyingly, often both of these will be present in the same datasheet).

As such, my PCBS were produced with the polarity of the relay coils reversed. Ouch! And of course, I only discovered this after soldering everything together.

Despite this, the latching relays work perfectly. I'm using transistors to create the ground connection for each coil, using a microcontroller signal to drive the Set and Reset pins high and low to toggle the relays. No problem at all.

Of course, the non-latching relays don't work (as expected).

My suspicion is therefore that the datasheet for the latching relays incorrectly specified "Bottom" view when it should be the "Top" view, making my PCB inadvertently correct. Any value to this suspicion, or is there another reason my latching relays are working despite screwing up the PCB connections?

(Both latching and non-latching coils are driven by the exact same circuits - same transistors, same microcontroller, same resistors, etc. In addition, cutting out the microcontroller and providing a 5V signal to the transistor pins does the same - the latching relay works, the non-latching relays don't.)

Datasheet here.

Answer 1

Omron relays that have polarized coils use internal magnets to help activation: -

Taken from here.

I suspect that a latching relay does not have a permanent magnet because it would "help" activation in one direction but be a problem when un-latching.

Hence, a two coil latching relay can probably be driven in reverse quite naturally. If you also look at the activation currents for the standard 5 volt coil version versus latching you will see that the standard requires 28 mA whilst the latching requires typically 40 mA. This kind of substantiates what I'm suggesting.

Answer 2

I don't seen anything inconsistent in the datasheet. Pin 1 is the + for set on all the relays.

They just consistently show the through-hole parts from the bottom, as one did in the old days (when single-sided PCBs were the norm), whereas the SMT parts are consistently shown from the top.

As to whether the dual coil relay will work at all, let alone work well "enough", with the polarity reversed, I'm not sure how we can guess that. I suppose you can try testing it and take some apart to see exactly how it works, or get out the X-acto knife and soldering iron and make it right (assuming it is actually incorrect according to the datasheet, of course).