Transistor theories

[Nigel Goodwin]

Actually i did read through the websites. Maybe my question was abit unclear. So what i was actually asking was about transistor Q1 and Q2. If Q5 and Q6 are OFF, wouldnt Q1 and Q2 base be connected to nothing(basically an open circuit)? Therfore, is it alright to leave it like that?

As I said above, it's advisable to have resistors across the bases and emitters of Q1 and Q2 - overwise any slight leakage in the NPN transistors could bias the PNP ones ON - which could well be disasterous!.

It 'may' work OK without them, but it's not guaranteed, and under some circumstances it could all go badly wrong - it's not good practice to leave them out!.

 

[Roff]

Base-emitter resistors will also allow the PNPs to turn off faster. When a BJT saturates, excess charge is stored in the base region, and it has to be cleared out before the transistor will turn off. Resistors will speed that process. Without the resistors, the charge has to drain off through the base-emitter and base-collector junctions.

 

[John Sorensen]

its not the base current that controls the behaviour of the transistor. its the base-emitter voltage Vbe that determines the behaviour of the transistor

This is fundamentally wrong. BJTs are current control devices.

j.

 

[Roff]

its not the base current that controls the behaviour of the transistor. its the base-emitter voltage Vbe that determines the behaviour of the transistor

This is fundamentally wrong. BJTs are current control devices.

j.

You can look at it either way, since voltage and current in the base-emitter junction are tied together in the **broken link removed**, but I agree that, unless you are using a transistor in some sort of log converter, current control is a lot easier to understand intuitively. I would like to meet the guy who designs a saturating switch with the diode equation in mind. :roll:

 

[fabbie]

dont mind me asking this, but what are darlington transistors used for? some h bridge drivers seem to be using it

 

[samcheetah]

This is fundamentally wrong. BJTs are current control devices.

okay lets talk about fundamentals.

why is a transistor called a transistor???

well, that because a transistor "transfers resistance"!!!!!

okay so what is resistance??

from Ohm's Law we have R = V/I which means that when you change the voltage across a resistor the current changes proportionally. so a change in voltage results in a change in current.

okay now what is transfer of resistance??

resistance is transfered when a change in voltage between two points results in a change in the current at some third point. so its basically just like resistance but the change in voltage and current is happening at different places.

thats exactly what happens in a transistor. a voltage across two terminals controls the current through the third terminal.

so does that mean that the equation Ic = BETA * Ib is wrong??? no, it isnt. it isnt wrong but it is just an approximation. it works in explaining alot of applications of transistors and thats why it is used so much. but that doesnt mean that a transistor is a current amplifier. to understand differential amplifiers, logarithmic converters, temperature compensation and other important applications you must think of a transistor as a transconductance device - collector current is determined by the base emitter voltage. and according to the Eber Molls model of the transistor

Ic = Is [ exp( Vbe/VT ) - 1 ]

where Is is the saturation current and VT is the thermal voltage.

the quantity Ib is just a by-product of the whole transistor-action. its something like who came first? the chicken or the egg? :lol:

and if you remember the voltage divider bias amplifier (VDB) you would remember that Ib and BETA are eliminated from the calculations. when designing a good amplifier Ib must be zero. so now i ask you that lets say you have a control panel consisting of Ib, Is, BETA, VT etc and lets name that control panel a "transistor" and the manual of this control panel says that keep Ib = zero for an optimum performance then how would you say that Ib is a controlling factor. either it is the controlling factor or it is an un-necessary factor in the whole process. thats up to you to decide.

now i know that it Ib can never be zero in a BJT. but thats the defficiency of the transistor. if it had been ideal there werent be any Ib. so if you see that Ib is always present then that doesnt mean that Ib is the controlling factor.

one thing more, do you know why Ib should be kept zero. because if Ib is zero then all the electrons entering from the emitter will go to the collector and become the collector current (in an NPN transistor) but if some of the electrons combine with the holes of the base they wont become the collector current. thats why Ib should be zero.

for further reference read the book "The Art of Electronics" from page 79 and onward.

and read the article http://amasci.com/amateur/transis.html

i hope that helps

 

[fabbie]

Im thinking of using this Hbridge cct. Actually this belongs to PIKE. And i hope you dont letting me make of it as a reference.

As Nigel mentioned previously, a resistor is required in between the PNP base-emitter junction. Any value would do right? Just as long as it keeps the transistor OFF.
As shown below in this diagram, there are 2 10K resistors connected at the NPN emitters, why are they required ?

 

[Roff]

Im thinking of using this Hbridge cct. Actually this belongs to PIKE. And i hope you dont letting me make of it as a reference.

As Nigel mentioned previously, a resistor is required in between the PNP base-emitter junction. Any value would do right? Just as long as it keeps the transistor OFF.
As shown below in this diagram, there are 2 10K resistors connected at the NPN emitters, why are they required ?

1. Lower value base-emitter resistors (Rbe) will allow the PNPs to turn off a little sooner, at the expense of requiring more current from the driving transistors. The current through those resistors when the PNPs are on is Vbe/Rbe. Even with 1k, that current is less than a milliamp. Since you are driving about 10ma into the base nodes, that is pretty insignificant. On the other hand, since you have to build in generous dead time at switching time, turn-off speed is probably not critical. For the circuit you have drawn, my personal choice would be 1k base-emitter resistors.
2. Notice that the resistors at the NPN emitters also connect from base to emitter of the lower H-bridge transistors. Without the resistors, those bases would be floating when they are off, just like the PNPs.

 

[fabbie]

i have forgotten to ask about this. Is it alright to place a motor in parallel with another motor in the H bridge?
Im facing this problem where I dont have enough PWM channels to drive the hbridge. What im planning to use is the PIC16F876 which has 2 PWM channels. I'll also be using 2 motors as my forward and reverse(2 motors required due to heavy load).
If i build 2 separate hbridge for each motor, i would require 4 PWM channels to control the forward and reverse. Therefore, isit possible to place the motors paralel with each other?

THX for any suggestions!!!!!

 

[samcheetah]

why not use a single motor capable of driving the load you have???

 

[Nigel Goodwin]

As Nigel mentioned previously, a resistor is required in between the PNP base-emitter junction. Any value would do right? Just as long as it keeps the transistor OFF.

10K would be fine!, it's not at all critical.

As shown below in this diagram, there are 2 10K resistors connected at the NPN emitters, why are they required ?

They perform the same as the top 10K's, as Ron H says.

 

[fabbie]

why not use a single motor capable of driving the load you have???

This is mainly because my battery is 12V. I dont think there is any point using a 24V motor when my supply can only provide up to 12V unless im mistaken
So can motors be placed parallel to each other?

 

[Nigel Goodwin]

why not use a single motor capable of driving the load you have???

This is mainly because my battery is 12V. I dont think there is any point using a 24V motor when my supply can only provide up to 12V unless im mistaken
So can motors be placed parallel to each other?

Yes they can, but it seems really pointless designing something that way?.

You don't need a 24V motor, just a larger 12V one!.

 

[fabbie]

I cant seem to find any large 12V motors. Most of them are in standard size. meaning its rather small. and the battery weight is quite heavy for a single 12V motor to support.

Nigel, when u said it was pointless, what do u mean by that? wouldnt the torque increase by doing that

 

[Nigel Goodwin]

Nigel, when u said it was pointless, what do u mean by that? wouldnt the torque increase by doing that

As a modification, after design, there might be some point to it - but from the initial design stage you should be looking to design it properly - not design it from scratch with fairly dodgy twin motors.

Also, two small motors will probably cost more than one decent one, it's not a good design idea!.

There are a great many sizes of small electric motors available, how big are we talking about?.

Oh - one quick thought! - assuming this is a four wheeled vehicle?, using a seperate motor feeding each wheel would be a reasonable design choice.

 

[samcheetah]

You don't need a 24V motor, just a larger 12V one!.

Also, two small motors will probably cost more than one decent one, it's not a good design idea!.

exactly, thats what i suggested. and connecting two motors might give you more mechanical losses. you will have to couple two motors on a single shaft. that would be a problem!!!!

ive seen alot of BIG 12V motors with gearboxes and they have alot of torque! so i would say that you should find a big 12V motor.

 

[fabbie]

i will be making a four wheel vehicle. one 6V motor(approximately as it was taken from another r/c) to control the left right motion for the 2 front wheels, 1 motor for the left rear wheel and 1 motor for the right rear wheel. I have actually already bought the 2 12VDC motors since they were selling quite cheap.
I dont plan on using 2 motors on 1 shaft. 1 motor for 1 wheel. Of course other than the mechanical losses issue, i am also abit worried on the motor alignment. This would mean 1 motor could be slightly faster than the other one. But both motors i bought are the same type, so i suppose it shouldnt be a problem.
Oh forgot to mention the motor size, i cant exactly give the precise size but its slightly larger than the size of a thumb.
So should i go ahead with my plan of connecting it paralelly or buy a large 12V motor. But buying a large 12V motor or motor with gears at my place could go up to around 100 bucks.

 

[samcheetah]

i have no experience in building robots and R/C cars but i think that if you use 4 small motors for each wheel then the steering will be more accurate. the type of steering you are trying to use is used in cars. but when all four wheels have a motor you will have to use a different type of steering mechanism simillar to a tank. and by using 4 motors you wouldnt need to have big motors at all. but ofcourse that depends on the load that the motors have to drag.

just my 2 cents !!!!!!!!!!!

 

[fabbie]

yup i was thiking about that. But the Hbridge is still the problem. placing motors parallel to each other isnt really effiecient. but i guess i dont have much choice

 

[samcheetah]

so, the problem you are now facing is that PICs dont have more than 3 PWM modules. while you want 4 PWM modules in one PIC. well then how about using two hardware PWMs and two software PWMs. i think that will solve the problem