Using 12v at base of transistor

Question

Hi all,

I'm thinking of using a transistor in a car circuit, to switch a light depending on the state of another light.

So it will be 12v as the signal at the base of the transistor.

I have a bunch of pn2222a transistors, and I can't figure out from the datasheet what the maximum means for that. It seems to be a reverse maximum (ie, don't let the output be more than 5v over the base voltage).

So the question is - is it possible to switch a transistor like that using 12v at the base?

Thanks!

-Nico

ps, I know I can easily do this using relays, but those take up more space, and I'd prefer to avoid the mechanical bits.

michaelkellett · Accepted Answer

Short answer - no. The base of a transistor is current driven - think of the base emitter junction as being a diode (the arrow on the emitter shows the direction of conventional current when forward biased). When the transistor is operating the base emitter voltage of a silicon transistor like the pn2222A will be about 0.6V (actually it will vary according to the base current). R2 controls the base current so for 500mA in the load you'll need about 10mA in the base so make R2 = 1k2. But actually, don't do this at all - use a relay ! The pn22222 is rated 500mA max so it will struggle to turn on a 2W light bulb (might be enough - you don't say what the load is - but remember that normal light bulbs draw perhaps as much as 10x the normal working current at switch on). The pn2222 is rated for 40V max so it should be protected form car power supply transients. If the load is shorted the transistor will fail. Relays can cope with all the above - but don't forget a diode across the coil (cathode to +ve supply) to stop it generating huge voltage spikes when it is turned off. MK

D_Hersey · Answer

Maybe there is a POTS soln: How To Install A Trailer Light Taillight Converter In Your Towing Vehicle I am not quite sure how the brake light signal is derived from turn signals. If you can come up with the FSM I am sure this can be configured with a coupla logic chips

D_Hersey · Answer

Picture perfect! If you get a big, Zener protected MOSFET, and drive it with your dinky bipolar, you will have a smaller and quieter soln. than a relay, don't forget a heatsink and a shut-off resistor. For an example: http://www.st.com/content/ccc/resource/technical/document/datasheet/5e/25/44/34/36/e7/49/93/CD00045633.pdf/files/CD00045…

michaelkellett · Answer

Hello, I think I can do the right thing with one transistor and one relay but it needs to be double pole changeover.

When both turn lines are high the relay will energize because its coil is connected between LT and a transistor switched on by RT. When the relay is on it routes LT and RT to the brake lights, when it is off LT and RT go to the turn indicators.

I'll scribble a circuit for you later today if I get a moment (or too bored with the real work !!).

MK

dougw · Answer

I'm not sure this state table is what you have, but once the states are populated, the logic to derive each state can be determined. It the table your vehicle controls (in blue) generate control signals (in green) and you want the output signals in yellow and red. The output signals must be derived from the green control signals. there are only 2 control signals, but they each have five states (off, on dim, on full, flashing off-on and flashing dim-on) The darker colours indicate flashing on-dim. If I have it right, there are 14 input states and 11 output states. This doesn't need a microcontroller to sort out as your yellow turn signals only have 2 states and your tail lights follow the green control signals, except when the yellow signals are flashing. The dim condition doesn't matter if you use the green signals as power rails for your tail lights. etc. Some states are not possible because, for instance, the left and right turn controls cannot indicate both at the same time. In order to figure out the state it is necessary to detect flashing. The two types of flashing don't matter, but it is harder to detect flashing because it isn't always the same. I-STATE 0 1 2 4 3 5 8 9 10 12 11 13 16 17 16 16 17 17 Controls Lights OFF ON OFF OFF ON ON OFF ON OFF OFF ON ON OFF ON OFF OFF ON ON Controls Left Signal OFF OFF ON OFF ON OFF OFF OFF ON OFF ON OFF OFF OFF ON OFF ON OFF Controls Right Signal OFF OFF OFF ON OFF ON OFF OFF OFF ON OFF ON OFF OFF OFF ON OFF ON Controls Brakes OFF OFF OFF OFF OFF OFF ON ON ON ON ON ON FLASHING FLASHING FLASHING FLASHING FLASHING FLASHING Inputs Left Wire OFF DIM FLASHING OFF FLASHING OFF ON ON FLASHING OFF FLASHING ON FLASHING FLASHING FLASHING FLASHING FLASHING FLASHING Inputs Right Wire OFF DIM OFF FLASHING OFF FLASHING ON ON ON FLASHING ON FLASHING FLASHING FLASHING FLASHING FLASHING FLASHING FLASHING Outputs Left Turn OFF OFF FLASHING OFF FLASHING OFF OFF OFF FLASHING OFF FLASHING OFF OFF OFF FLASHING OFF FLASHING OFF Outputs Right Turn OFF OFF OFF FLASHING OFF FLASHING OFF OFF OFF FLASHING OFF FLASHING OFF OFF OFF FLASHING OFF FLASHING Outputs Left Tail OFF DIM OFF OFF DIM DIM ON ON ON ON ON ON FLASHING FLASHING FLASHING FLASHING FLASHING FLASHING Outputs Right Tail OFF DIM OFF OFF DIM DIM ON ON ON ON ON ON FLASHING FLASHING FLASHING FLASHING FLASHING FLASHING O-STATE 0 1 2 3 4 5 6 6 7 8 7 8 9 10 9 9 10 10 Note the brakes can be flashing if you pump the brakes - you don't want this to cause turn signals to come on unless the turn signals are also on. I am not sure there is enough information in the control signals to know what is going on if you pump the brakes at the same rate as the turn signal flashes - while they are flashing. If you use a flashing detector, it will delay your turn signal activation. I hope this doesn't look too confusing and complicated - it is just trying to set up a way to document what you and Michael have been saying in words, and in translating to a circuit, it should simplify to something similar to what you and Michael have been designing.

michaelkellett · Answer

No much time today but just a couple of quick thoughts. 1) I would go for the AVR processor but code in C - I just don't trust the Arduino code environment for precise timing, start up etc. The Atmel software tools are free and nice. 2) @Doug - I'm used to classic state diagrams which show how events cause transitions in the machine's state - I don't see quite how your table can show how the reaction to new input conditions is affected by the state the system is already in. MK

michaelkellett · Answer

Hello Nico, Here's my scheme: The relay can be any 12V coil double pole change over (often called DPDT) you fancy. The coil resistance should be at least 100R if you want to use up your pn2222a transistors. The relay will only be energised when both turns inputs are high and the brake lights get the power. If only one turn input is high then the turn indicators get power. With R1 and R2 as shown the turn on threshold for Q1 on the LT input is only about 0.8V, you could reduce R2 to 100R which would set the threshold at 6V or so. The transistor only switches the relay coil current and is protected against supply spikes and relay coil inductive spikes by Z1. A small capacitor (100nF ceramic) from Q1 base to emitter wouldn't hurt and might reduce sensitivity to electrical noise. MK

michaelkellett · Answer

Hello Nico - you are correct - I hadn't registered that problem correctly. I think I can tweak it so that it's fixed. The software I use for pcbs is EasyPC - by far the best cheap PCB CAD that I'm aware of - it doesn't do everything but what it does it does very well. I'll post mk2 in a bit. MK

michaelkellett · Answer

Hello Nico, Here's a better design - two extra diodes but the relay only needs to be DPST so will be cheaper and easier to find. I think this nearly does what you want - if only one input is high the appropriate turn light comes on, if both inputs are high the relay energizes and the brake lights come on (but so do the turn lights), the brake lights stay on until both inputs go low because the c terminals on the relay hold it on so long as there is power through either diode. I don't really like the turn lights coming on at all when braking but this gets much harder since you need quite a lot of 'statefullness' in the controller - are you up for a micro ? MK

michaelkellett · Answer

I've only ever seen one state machine project of any complexity implemented in PROM and it was a total and complete disaster. The idea was to avoid software and its approvals for an application of the highest integrity but the result was to end up implementing software in delays in Rs and Cs - Nico will have all the same kind of problems. A nice little AVR will do him fine - can be reprogrammed as often as he likes, is supported by free software and probably costs a lot less than a PROM and supporting logic MK

ntewinkel · Answer

I was just thinking about this some more... With your circuit locking the brake lights on, I think I can now make the flashers work properly. The logic is: if the brakes are not on, then light the signal light that is ON. When the brakes ARE on, light the signal light that is OFF. That would save getting into more complex microchip or PROM or whatnot. Cheers, -Nico (ps, camping this weekend, so I'll likely be MIA most of the time)