cool part

Okay, say I have 5V on the base, and 4.4V across the relay coil, and (say) 10mA is flowing out of the emitter. Steady state. Now I suddenly drop the base voltage to 0V. Current continues to flow out the emitter, but the emitter voltage drops to -0.6V (because of coil inductance). It starts out at 10mA or so, and drops towards zero from there.

Best regards, Spehro Pefhany

Yeah, mechanical relays are pretty darn good even for currents waaay below spec. I've grown fond of the Pansonic TN2 series. Which ones are you using?

robert

On Tue, 13 Nov 2007 22:20:07 -0500, Spehro Pefhany

....So long as the base driver supports positive current in that quadrant. Digital or overdriven linear parts are usually cut off, making the only source of base current the parasitic diodes to substrate.

......You might measure a larger negative voltage on the emitter than is anticipated here.

RL

A couple of other people make drop-in equivalents... Omron and NEC.

John

Any regular cmos part will go low pretty hard, whether sourcing or sinking current, it still looks resistive. Even a bipolar TTL will stop at ground in this situation.

John

Yes, that's the upside, but PCB layout could have been simpler.

Probably there are two dies inside as John L. suggested, so there wasn't much option.

Best regards, Spehro Pefhany

Have it your way. You didn't mention if you were dropping the base to common or just floating it? So, I'll assume after your description that you are pulling the base low to common. That being the case, with the return potential from the coil and expecting the emitter to handle that is risky. I've seen circuits like that many times where it simply destroyed the transistor gain but past the simple diode test.

and if your not pulling the base to common I wouldn't dare to think as to where that voltage is going.

I'll stick with my wheeling diode protection, thank you.

Why? It's still acting like a normal emitter follower.

What would kill the gain? Nothing's zenering here.

Speff's circuit is fine, if you don't mind losing 0.6 volts of coil drive. The relay is off if the port is hi-Z before being configured, it drives fine, it clamps the flyback fine.

John

Maybe.

I'll stick with experience and what's obvious to me. Maybe it's flawed how ever, it's been working for me so far.

look at these articles. They have Emitter relay driver examples. You'll notice they insist on diodes. I wonder why.

A diode won't hurt; it's just unnecessary. You could use seven diodes, too.

It's obvious; the guy doesn't understand things.

John

Because they're idiots. Figs 4 and 5 are dangerous nonsense, for a number of reasons.

John

UofG a.k.a. "Moo U"

Best regards, Spehro Pefhany