Please consider the following for switching an relay:

Change the PNP to NPN must be care about duty cycle to match what he want.

 

Change the PNP to NPN must be care about duty cycle to match what he want.

Of course. They are worried about lack of voltage for the relay, but there would be none since the transistor is SINKING current straight to ground.

 

Yes well I must have missed something then. Operating a 5 volt relay off 3V3? Well the whole schema seems wrong then.

Omron G5V-1 relays are available with coils rated for 3V, 5V, 6V, 9V, 12V and 24V. I think we can assume he's using the right one...

Also, consider what your NPN solution does to his relay ON and OFF times: it reverses them. He wants a short, fixed ON duration and a long, variable OFF duration.

Of course. They are worried about lack of voltage for the relay, but there would be none since the transistor is SINKING current straight to ground.

Huh???? Any transistor in series with the relay, whether it's an NPN or a PNP, is going to have a Vce(sat), and it's going to subtract from the net voltage available to the relay coil. That loss may or may not be significant, and in this case it might not be judging from the relay pull-in specs, but it's still there. What makes you think that just because the transistor is sinking current rather than sourcing it, there would be no Vce(sat) loss at all????

 

Huh???? Any transistor in series with the relay, whether it's an NPN or a PNP, is going to have a Vce(sat), and it's going to subtract from the net voltage available to the relay coil. That loss may or may not be significant, and in this case it might not be judging from the relay pull-in specs, but it's still there. What makes you think that just because the transistor is sinking current rather than sourcing it, there would be no Vce(sat) loss at all????

Yeah I guess you will get your obvious drop of ~0.65 volts between base and emitter. But this is ridiculous to think that such and entity cannot be done with just an 2-cent common garden variety BJT.

 

Why have you switched from an N-channel MOSFET in the original circuit to a P-channel MOSFET?
The results will be reversed.

Because I believe I had my logic reversed from what I intended in the first place, and I think this reversal fixes it. The goal is to energize the relay coil for short, equal length bursts of time and to leave it off for longer, variable lengths of time (set by pot.) I copied some formulas wrong when I first worked out 555 timer calcs and had duty cycle reversed, among other issues. I think (hope) the current p-channel arrangement provides the duty cycle I want...

 

Are you going to breadboard this gadget first?

I always do, simply because it brings good luck: almost every time I've tested a design before committing to PC board or sending it to a client, the thing turns out to work exactly as expected; and the few times I've skipped that step (mostly because I thought the design was too simple to bother breadboarding it), most often something doesn't work right and I end up red-faced.

Call me superstitious if you will, but that's the way it's been for me.

It is not being superstitious. It is practicing due diligence.

You guys are both absolutely right! I must confess I was considering not breadboarding this project, because money's tight, I'm developing it for free, and I don't specifically need my own version of this circuit for anything yet...

But, thanks to your nudges in the right direction, I think I will just go ahead and get the necessary parts to test it anyway. It won't cost much and it'll be good practice for me. Plus, it'd be really rotten to send a flawed circuit plan to someone thousands of miles away who has less diagnostic/troubleshooting ability than I do. Whether I keep it as a finished circuit or just keep the parts around, I'm sure I'll find some use for it all in the future.

 

Yeah I guess you will get your obvious drop of ~0.65 volts between base and emitter. But this is ridiculous to think that such and entity cannot be done with just an 2-cent common garden variety BJT.

Wouldn't it be the collector emitter saturation voltage that would be the issue here? That's what will be missing from the relay coil, not anything to do with the base.

Regardless, this project will be a one-off, maybe at most 2 or 3 units, so a little extra expense to ensure success is probably a worthwhile investment.

 

Wouldn't it be the collector emitter saturation voltage that would be the issue here? That's what will be missing from the relay coil, not anything to do with the base.

A transistor is like two diodes back to back. So yes drop between emitter and collector as well.

 

Wouldn't it be the collector emitter saturation voltage that would be the issue here? That's what will be missing from the relay coil, not anything to do with the base.

Exactly. Vbe(sat) is irrelevant here.

Regardless, this project will be a one-off, maybe at most 2 or 3 units, so a little extra expense to ensure success is probably a worthwhile investment.

A wise choice. I think the design you presented in post #47 is fine just the way it is. Go for it.

 

Plus, it'd be really rotten to send a flawed circuit plan to someone thousands of miles away who has less diagnostic/troubleshooting ability than I do.

Know what's worse? Sending a flawed design to someone who has more diagnostic/troubleshooting ability. Been there, done that. Really embarrassing...

 

Really embarrassing...

Only if you have tickets on yourself it would be. An expert makes an slip up?

 

Yeah I guess you will get your obvious drop of ~0.65 volts between base and emitter. But this is ridiculous to think that such and entity cannot be done with just an 2-cent common garden variety BJT.

Please do not used your "personal emotional word" to insult other members and just focus on your professional viewpoint.

 

Many thanks to everyone who helped me with this (especially OBW0549 - you did a lot of the heavy lifting towards the end!)

Everything is working beautifully so far, although I've only got a 5V power supply to test it with at the moment. I'll have to dig out some batteries and/or some other wall wart supplies to make sure it handles higher and lower voltages without any surprises. I'll update again once I've done that. For now, it just feels good to get something made and have it work right!

 

Excellent. Good work!

 

Just tested my worst case low voltage scenario: <3VDC supply, which is actually lower than I would CLAIM the circuit should be reliable at, but it was originally the lower end of what we HOPED to be able to use (based on part specs, I think it should be very reliable down to 3.2V, but anything below that depends on some measure of luck.)

I just ran the circuit off of two AA batteries in a tv remote control. 2.97V into the circuit, 2.78V out of the LDO while the relay is energized, and steady action on the relay.

Taking a 3V supply and losing voltage drop through a regulator and a tiny bit more through a transistor, I had doubts about getting the 3V relay to work. That's why I tried hard to spec the least drop possible everywhere I could, and it appears to have paid off!

Guess the only thing left to test now is giving it 9V or maybe even 12V supply and see if the LDO starts to warm up.

 

At 12 v that regulator will be dissipating around half a watt when the relay is on, but if it's on only a low percentage of the time I don't think you'll notice much heating.

So far, so good!

 

Yeah, that's about what I came up with too. Is there any general rule for how much heat a TO-92 package can dissipate? Or does that depend too much on the internal workings of the component? Do you have to calculate it based on the thermal resistance specs? So much to learn!

The data sheet just says power dissipation is internally limited. Like you said though, at the highest I'll be around 40% duty cycle, and there's practically no current when the relay is off, so we're probably well under 250mW average dissipation in the worst case 12V, maximum speed scenario.

 

Is there any general rule for how much heat a TO-92 package can dissipate?

I don't think so. For TO-92 packages, I'll just do a quick back-of-the-envelope calculation of maximum power dissipation and if it's less than 250 mW I don't worry about it any further. If it's more I'll take a more detailed look, considering thermal resistance, maximum junction temperature, and so on.

Or does that depend too much on the internal workings of the component?

I'd say yes. I reckon there's usually three distinct temperature limits: first, there's the Absolute Maximum junction temperature, above which a device can suffer catastrophic failure during a fault condition (or by careless design). Below that, there is usually an operational maximum junction temperature, above which the device might not meet all (or any) of its published specs, or even cease operating altogether until the temperature drops again. And finally, below that, there is what I would call a "performance" maximum temperature that I have to stay below to get the device performance I'm looking for. One example might be a voltage reference IC that provides the voltage reference for an extremely high performance analog-to-digital converter. I want that voltage reference to be as stable as I can make it, so I will take care not to draw any more current from it than is absolutely necessary. That way, the part will experience minimum temperature rise.

The regulator you've chosen appears to be pretty much bulletproof with regard to overheating (unless you stick it in an oven, of course), because of its internal overtemperature shutdown feature. If it decides things are getting too hot, it will simply turn off. Once things cool down again, it will turn back on. Note that this can sometimes result in some pretty bizarre behavior, because the heating/cooling cycles can be as short as a few dozen milliseconds due to the short thermal time constant of the silicon die. If you see your regulator output rapidly switching on and off at a 10-50 Hz rate, that's a clue that it may be cycling in and out of thermal limiting.

Do you have to calculate it based on the thermal resistance specs?

Sometimes that's necessary, sometimes not (see above).

So much to learn!

Yup. That's electronics for ya!

That's one of the great things about boards like this: no matter how much you know, and how much experience you've had (I started tinkering with hobby electronics 57 years ago, when I was 9), there's always someone around who knows something you don't know. ALWAYS. Ergo, there's always something new to learn.