PNP switching method

[wingerr]

I saw a suggestion on using a PNP for switching a load on the low side with the collector grounded, emitter following the input (first diagram) as opposed to the high side. I normally would switch the high side (second diagram), but I guess the first method should also work, although it would dissipate more power in the transistor and not have a full swing on the output. If the output swing isn't all that critical and the transistor has enough reserve, would that be a good way to go, being simpler?

**broken link removed**

 

[Nigel Goodwin]

You should still have a resistor in the base in either case - and if you want to switch the low side, then just use an NPN transistor.

 

[ikalogic]

NPN are for high side switching
PNP is for low side

**broken link removed**

Do the opposite and the transistor will never saturate.

the math behind that is quite simple,

read this tutorial for more infos about switching transistors:
**broken link removed**

 

[jpanhalt]

Your drawings are correct, but the terminology is confusing when compared to mosfets. In other words, low-side/high-side with mosfets refers to the location of the mosfet, not the load. Thus, one usually uses a PNP mosfet for high-side switching. John

 

[crutschow]

Your drawings are correct, but the terminology is confusing when compared to mosfets. In other words, low-side/high-side with mosfets refers to the location of the mosfet, not the load. Thus, one usually uses a PNP mosfet for high-side switching. John

To nit pick, it's a P-Channel mosfet, not PNP (which of course refers to a bipolar transistors).

 

[jpanhalt]

I knew that. My fingers get tired of typing. I spent 3 hrs. trying to fill out an order for one of the common suppliers. John

 

[wingerr]

Normally I'd say to use the right part and switch with the method that allows full turn on of the transistor, but this happened to be a case where someone was just looking for a bump in the drive current in an existing setup, with a low going signal being used to switch a relay, but apparently was either halfway defective or just didn't have enough drive.
The advantage of the simple setup was that it didn't invert the signal and wouldn't need rewiring of the relay connection.
I guess adding base resistance would probably be safer by limiting the initial surge of base current when the transistor is switched; with a tradeoff of lower drive further by effectively adding more negative feedback.

 

[Roff]

Normally I'd say to use the right part and switch with the method that allows full turn on of the transistor, but this happened to be a case where someone was just looking for a bump in the drive current in an existing setup, with a low going signal being used to switch a relay, but apparently was either halfway defective or just didn't have enough drive.
The advantage of the simple setup was that it didn't invert the signal and wouldn't need rewiring of the relay connection.
I guess adding base resistance would probably be safer by limiting the initial surge of base current when the transistor is switched; guess it'd have to be a tradeoff of lower drive further by effectively adding more negative feedback.

You don't generally need a base resistor for an emitter follower. If it were truly driven by a switch, I would be inclined to add a resistor from base to vcc, just because I don't like floating bases. I don't think you are actually driving this transistor with a switch. Are you?? If not, what are you driving it with? If it is another circuit, the voltage needs to swing from ~zero to Vcc.

 

[wingerr]

No, I just drew in a switch as input for simplicity; it's supposed to be a relay drive output from an alarm module. Most likely open collector, so the pullup would probably be a good idea.
I was just looking at the general idea of it's use as a simple workaround; not something for a new design. For driving a relay, it wouldn't be too fussy, so it looked like it would work.

 

[ikalogic]

No, I just drew in a switch as input for simplicity; it's supposed to be a relay drive output from an alarm module. Most likely open collector, so the pullup would probably be a good idea.
I was just looking at the general idea of it's use as a simple workaround; not something for a new design. For driving a relay, it wouldn't be too fussy, so it looked like it would work.

the pull up will always create a voltage drop, the bjt will never saturate, as i told you in my previous post

 

[Roff]

...the bjt will never saturate, ...

But for some loads, e.g., relays and LEDs, 0.7V to 1V across the transistor may not matter. If your driving signal can't supply much current, it might even be an advantage, as the emitter follower only draws as much base current as it requires, whereas a common emitter stage needs to be driven with Ib=Ic/10 (less if you're adventuresome).