Using the Rlp/Tlp with microcontroller and 12V

Wireless & IoT
1

I guess, technically, my question is about the transmitter.

I’d like to increase the range as much as possible (ok, as much as the FCC will allow with

out a license… :D). I think I’ve got the antenna figured out (433 mhz 1/4 wave).

I also know that you get better range with a lower baud rate so I’ll experiment

to see what it buys me and I’ll use some sort of encoding scheme,

probably manchester.

I also want to run the transmitter at 12 volts. However, the datasheet

says that input high should be not less than Vcc - .5V. So, I’ll need to

boost the output of my PIC (might be an AVR, doesn’t really matter) to

about 12V. I’m thinking of a small signal NPN transistor as a classic

emitter-follower like so:

http://www.geocities.com/pcbs4less/rfboost.gif

I’d probably feed the transmitter’s Vcc through another NPN to a) match

the drop of the input and b) be able to turn the transmitter off when not in

use. I probably don’t need R1. I’m not sure of the R2 value but was thinking in the 10K range.

Does this make sense? Any ideas that would work better?

Phil

2

You’re not going to get 12V out of that. With 5V at the left of R1, you have something less at the base and 0.7V less than that at the emitter. Another way to think of it is if you do have 12V at the emitter and 5V at the base, you’re reverse biased.

Try a PNP, 3906 or whatever, collector to gnd, emitter to a 4.7K, other end of the 4.7K to 12V, output between emitter and the 4.7K. Mostly the same circuit, only R2 and Q1 exchanged…sort of.

Pete

3

Thanks. I expected to get a drop but my breadboarding shows something more like 5V which doesn’t do it.

The PNP circuit doesn’t seems to work for me. I still get a big drop but more importantly, I’m seeing 7V between base and gnd (not good for a 5V uC pin). Maybe the PNP is blown… I did try the NPN low, resistor high, output from the collector/resistor junction and get an inverted output but at least it gets nearly full Vcc (within a quarter V). I can live with inversion since I’m bit banging. Still not sure if this right way to do it. By the way, I’m using a 10V supply for testing (lm317, resistors on hand…) rather than 12V.

Bipolar Transistors still mystify me. Gotta spend more time in Horowitz and Hill and play with SPICE.

Thanks again.

Phil

4

An emitter-follower will follow the voltage of the base (hence the name) so you’ll just get 5v out again, minus the Vbe drop (0.7v). This is what I’d try:

http://www.hhhh.org/wiml/tmp/ckt/pnp1.png

R1/Q1’s job is to convert the microcontroller’s output of “+5v or 0v” into “connected to ground or floating”. R2/Q2 does the actual power switching. You can almost, but not really, get rid of R1/Q1 by programming the uC’s output port to be high-Z, because the port will still have input protection diodes keeping the pin’s voltage between the uC’s supply rails. Most microcontroller data books have a schematic of what their pin’s circuitry looks like for each kind of I/O pin, which is good for that kind of detail.

AFAIK there’s no reason not to use a FET for this if you’re more comfortable with them than with bipolars.

5

thanks for all the help. I was considering a fet but think I would have

to go to logic level devices (or use a more complex triggering scheme).

I put together the various circuits in SPICE and they make a tad more sense

to me now. I like the 1 transistor inverter:

http://www.geocities.com/pcbs4less/5to12.gif

SPICE and breadboard says it works. Current through R2 (4.7K) is a little

above 4mA so it seems a little high on draw. Given the cost and

compactness, this makes a lot of sense to me. I’ll just need to invert

the bits when I output them Any reasons why not to use this one?

Phil

6

That’s good as long as the 12v side doesn’t draw much current. (For some reason I thought you were using this to switch the module as a whole on and off. For a logic input, this circuit is fine. At least, it’s worked when I’ve used it. :slight_smile: )

The larger the current draw, the smaller R2 has to be to keep the ‘on’ voltage near +12v, and then R1 has to be smaller (and the transistor beefier) to pull the ‘off’ voltage down near 0v, and you end up just burning a lot of power in the resistors.

7

I thought I posted this yesterday but must have muffed something…

I can see how you thought that, I should have been clearer.

On the topic of switching power to the unit, I would like to avoid the voltage

drop from the bipolar transistor as the input signal high can not exceed Vcc,

per the datasheet. So, I’m looking at a P Channel MOSFET to do this. The

circuit is cobbled from Horowitz and Hill and is similar to Wiml’s with the

exception of using a P channel MOSFET in place of Q2.

http://www.geocities.com/pcbs4less/p-mos.gif

I SPICED this and it appears to do what I want. R1 can be fairly high - even

at 100K, the circuit simulated just fine. R2 simply needs to pull the voltage

up to 12V when the Q1 is off so it can be fairly high. I’ve simulated 1K to

10K with no apparent problems. The good news is that Load sees a full 12V.

Its not super fast - turn on takes about 8 uS and off takes around 60 uS

with a 4.7K R2. Higher values of R2 slows it way down but draws less

current, engineering trade-off to be made here.

Of course, the load must not cause Q2 to exceed it’s current spec.

Any thoughts on this? Should I have a resistor between Q1 collector and Q2 gate?