I believe I have gotten an LED to light up in both configurations, but my understanding is that the configuration in which the emitter is on the ground side of the circuit is the correct way. Is this right? Are both ways right? Thanks.
d_vee:
I'm helping a kid make a motor driver using 2n2222's and ran into a bit of confusion. My understanding of transistors is somewhat limited. I encountered two Instructables that show the collector going to ground,
Unfortunately the authors of those instructables have limited understanding of transistors too, only one had an actual schematic and the wiring pictures from the others are hard to follow. All of the designs have problems.
The all npn arrangements of those h-bridges are not going to work well. They might light an LED but the upper transistors will never be turned on properly. They will get hot and burn up with any real load. They also have no diodes for protection.
They should be using PNP's for the top transistors, the emmiters of the npn's will be connected to ground. The emmiters of the pnp's to +v. An additional couple npn's are needed to drive the pnp's if the supply voltage is greater than 5v.
Search around for npn-pnp designs for simple low current and voltage. Ideally mosfets are a better choice than bjt's. H-bridge IC's make the job much simpler.. even a l293d can handle more current than a 2n2222
Yes, the technical term is saturation, when a transistor is fully switched on and the collector is only 0.1 to 0.3V
above the emitter. For that the base current has to be a sizable fraction of the collector current, normally 5 to 10%, and you have to use either common-emitter circuit, or a bootstrapped supply to drive the high-side
bases. (This is commonly done for n-MOSFETs, less commonly for BJTs due to the massive base currents
needed)
A lot of people assume that a transistor with a gain of 250 (say) will switch 1A with 4mA of base current -
whereas it will only bring the collector-emitter voltage down to 1.5V or so (still in the linear region with
the base-collector junction reverse-biased).
To saturate the base-collector junction has to be forward-biased and the normal transistor action stops
completely, all that is left is gain due to carrier diffusion from the massively doped emitter to the
very lightly doped collector - this is driven by concentration gradient, and gains of 5 to 20 are about it
for this mechanism, whatever the small signal gain is.
Thanks guys. If I have some PNPs I'll use those, but if not I'll just make sure I have my circuit arranged so that the NPN emitters are "pointing towards" ground. MarkT, I definitely have to read up a lot more on transistors. Past using them as switches, I know very little about them.
I found the circuit below online. As a simple circuit to drive a small motor, should this setup, plus four diodes, be sufficient? (Ignoring the specific transistors and just focusing on NPN vs PNP).
Maybe, maybe not. If there is any delay while A side PNP turns off while NPN turns on, then "shootthru" can occur and cause one or both transistors to burn up when 5V goes unchecked thru both parts to Gnd. Check the specs on the two transistors, see what their turn on & turn off times are like. You may find you need 4 separate control signals for PNPs and NPNs so you can control their on/off times, with slight delays between PNP off and NPN on.
Here is a simple 4 transistor bridge that works. It doesn't have the 5 volt motor limitation and the shoot-through problem. You do have to take care in your program that both inputs are never high at the same time.
d_vee:
Thanks guys. If I have some PNPs I'll use those, but if not I'll just make sure I have my circuit arranged so that the NPN emitters are "pointing towards" ground. MarkT, I definitely have to read up a lot more on transistors. Past using them as switches, I know very little about them.
I found the circuit below online. As a simple circuit to drive a small motor, should this setup, plus four diodes, be sufficient? (Ignoring the specific transistors and just focusing on NPN vs PNP).
That circuit has shoot-through - both top and bottom transistors conduct for a while when switching
over because transistors take a lot longer to switch off than on. It also lacks free-wheel diodes
so will rapidly fail to voltage spikes.
It also shows a motor powered from the Arduino 5V rail - never do this.
