two relays

Den fredag den 24. marts 2017 kl. 00.10.56 UTC+1 skrev John Larkin:

A) uses cheap resistors and lowside control

so I'd say it has the other beat on everything but current and if that such is a big issue get some 12V relays

But it uses twice as much supply current as is necesary. The switches are isolated DIPswitch sections. Ckt B is out because it needs SPDT switches.

I have the 5V relays in stock.

Circuit C looks most efficient. I'd prefer to not use a switcher.

The relays are DPDT, so it is possible to use the extra contacts to do the more-pullin-than-holding-current trick. Apply 12 volts until the relay pulls in, then switch to a low holding current. That's circuit A, with the extra NC contact shorting the resistor.

Den fredag den 24. marts 2017 kl. 00.46.42 UTC+1 skrev John Larkin:

yeh, if holding voltage is less than 2.5V that could get A on par with C so it is down to 4 resistors vs. 2 zeners

Resistors are a waste of power, Zeners are relatively expensive, and choosing to use 5V relays when one only has 12V power is... not the best design decision. Besides, Zeners sometimes fail short.

One could also use an LM317 as current source, putting the two relay coils in series, with the switches shunting each coil... GAAA now you've got ME doing this silly stuff!

I don't get B, but C looks good, (as long as the power works). George H.

Agreed. A silly problem that wastes everyone's time.

What drives the SPST and SPDT switches in (A) (B) and (C)? Are they 12V relays? If not, do they require additional resistors and zeners?

What happens to two zeners in series with an overvoltage on the 12V supply?

Why not simply use a 12V version of the 5V relays?

Or use a 5V supply?

The problem is not scalable. What if you have to drive three relays?

Actually, I do have to drive three relays.

Easy Peasy. Just use the SPST or SPDT contacts in your (A)(B) and (C) switches.

Problem solved. Next?

Noise an issue? Then I don't suppose you could just PWM the drives?

Then you get a WDT for free. ;-)

???

Rather not. This is a picoseconds-jitter thing, and switchers are jitter generators.

That's great! What's a WDT?

I thought engineering was sometimes described as "making what you need out of what you have."

Sometimes a 12V supply and 5V relays is what you have.

Anyone can make a "design decision" to spend more money

The best way to do three or more relays is probably to take advantage of the pullin/hold current differential, using the second set of DPDT contacts on each relay. That only needs one resistor per relay, and saves the most power when they are on. Or off.

I guess that's the best way, period.

Thanks for that insight.

Watchdog timer I guessed.

NT

We try to avoid adding new parts to stock. We have 6640 stock numbers already.

But ... the contacts are most certainly break-before-make, so timing could be a problem. Depending upon inertia of armature, flyback diode current, ... No experience, just a thought.

My idea for a "fault tolerant" version:

Haven't tested. Or Spiced. May not work...

Are you one of those unlimited budget people?

NT