Motor driving

[patroclus]

I read that Darlington transistors are good for controlling DC motors, or ULN2003, that it a darlington array, so you ground the terminal you want, and the motor runs.
I've also read that a normal transistor can be use for this task, BD140, for example (up to 1.5A Ic).

But this seems confuising for me. While Darlingtons in ULN2003 only support 500-600mA Ic, and have quite high hFE, BD140 have a much higher max Ic current, but much lower hFE (over 25 min).

What is the best choice?
Many thanks

 

[crust]

If you can live with the 500mA or so limit on the 2003, I would think that is a better choice, mainly because it is smaller (which is probably why it cant handle as much current) and has 4 darlingtons inside it. The discrete transistor will require more base current (but probably not so much more it would matter), but can dissipate more heat.

 

[Nigel Goodwin]

Darlingtons are simply two transistors connected together, there are a couple of ways of doing it. If you add a small transistor to feed your BD140, and the new transistor has a gain of 100, the resultant darlington will have a gain of 2500 - just multiply them together.

 

[patroclus]

I know that about darlingtons, but the circuit I was looking at is not in Darlington configuration.

**broken link removed**

when OPT1 is on, Q1 and Q2 are on, turning Q3 (BD140) off.
when OPT1 is off, Q1 and Q2 are off, turning Q3, and only Q3 on. there's no darlington, and hFE is very little (between 25 and 40), but base current at Vcc 6V+ won't reach 2,5mA... so colector current will be below 0.1A ?? This seems not posible...

 

[crust]

The 25-40 is the minimum hfe of that transistor. You should check the datasheet as their is (or should be) a curve that will show you the hfe given the collector current. I would guess that the hfe in operation would be closer to 150-200 or so, though the current gain decreases as the collector current increases.

 

[patroclus]

Well, there's not Hfe curve.. only a few values for some Ic currents.
Even hfe 150 is very optimist, and won't lead to more than 375mA, and that's very low value (for that, it is better to use ULN2003 for example), or any other transistor like BD547...

but wait a minute... If the transistor is saturated, and should be, shouldn't it? The hfe won't be usefull anymore as the colector current would be as high as posible (depending on the motor resistance), so it will achive 500mA with no problems.. am I wrong?

 

[Chippie]

So why not use a darlington tranny like the BD679?

 

[patroclus]

Yes, it's a posibility, but this design really works, and my question is (just to know, when going into more complex design) if a darlington is really needed, when this robot only requieres a BD140... yes, as I said, it supports quite high current, but this hfe....

 

[Nigel Goodwin]

Yes, it's a posibility, but this design really works, and my question is (just to know, when going into more complex design) if a darlington is really needed, when this robot only requieres a BD140... yes, as I said, it supports quite high current, but this hfe....

Have you actually seen it working?.

If so, check the gain of the transistor used (not what the book says, the actual gain of the particular sample used).

The site itself isn't in English, so I couldn't understand it, but what type of motor is it feeding?, and what kind of mechanical load is the motor having to move?.

You are quite right that the base current of the BD140 is supplied solely through the 2.2K resistor, and that from the gain figures you have quoted the collector current won't be very high. I've just checked my Towers book, and that gives the gain of a BD140 as 40-250 - so at the higher end it would be lot better. But it looks a very poor design, far too dependent on the specification of the particular transistor used - good design should take account of gain spreads, and you shouldn't need to select or change values to compensate (which this design would tend to need).

 

[patroclus]

Well, it works, and you can see a video of it
**broken link removed**

The motor is common 6V DC, but the author let's everyone decide which one, as soon as it is DC motor, and not over 6V. He says that must have some reduction, and they move a wheel each. also the robot is quite small and light.

One solution would be changing the 2.2k resistor to 1k, for example, but this really confuses me, because the robot has been made by many people and seem to work always. And this also confuses me, cause if a transistor like that works, why then many robotic web pages tell you that a darlington transistor has to be used??

By the way, isn't it supose to be saturated when on??

 

[Nigel Goodwin]

By the way, isn't it supose to be saturated when on??

It would certainly be advisable, is it's not saturated it will dissipate a fair bit of heat.

Buy a BD140, get a 2.2k resistor, and test it!.

 

[tavib]

The hFE of a bipolar transistor is minimum when the transistor is in saturation(when Vbc=Vb-Vc<=0V for PNP), so you must consider the minimum value for BD140 at 40-50. With this value you will have 100-150 mA with transistor saturated. When transistor go out from saturation(BC junction become revers polarized => Vbc>0 for PNP) the hFE will increse significantly over saturation value, up to 80-150 for BD140, but in this case the transistor disipation is increase over the saturation disipation.
Is recomended to decrese the 2K2 resistor from base of BD140 to 1K value for increase the BD base current, because the BC557 can source enough current through 1K resistor to rise voltage up to Vcc.

 

[patroclus]

Nigel, you're right, best thing is to get a BD140 and test the circuit.
anyway, how can you be sure, when you design a circuit like this, that the transistor will be saturated?? For that, base current should be more than

Ib > (Max colector current / minimun hfe) + some safety value

Isn't it?
So, if motor drives 300mA, this transistor won't be saturated unless Ib is greater than at least 7.5-10 mA...
If my analisis is wrong, please let me know.
thnks

 

[Nigel Goodwin]

Nigel, you're right, best thing is to get a BD140 and test the circuit.
anyway, how can you be sure, when you design a circuit like this, that the transistor will be saturated?? For that, base current should be more than

Ib > (Max colector current / minimun hfe) + some safety value

Isn't it?
So, if motor drives 300mA, this transistor won't be saturated unless Ib is greater than at least 7.5-10 mA...
If my analisis is wrong, please let me know.
thnks

Sounds good to me!. Personally I like to stuff plenty of current in the base, to make sure it's really turned on hard.

The circuit you gave the link to seemed very poorly designed, it uses three transistors, yet used a 2.2K to provide all the base current for the output.

 

[patroclus]

Could it be done, conecting only Q3 to Cyn70 output (the other 2 transistor just seem to invert twice the signal..), with a base resistor of 1k or less, and choosing another transistor with higher min hfe?? (or resistor of 330o, jeje, hard on)
Does it have to be darlington??

 

[patroclus]

Well, I supose I'm wrong about only one transistor, as Cyn70 may not always be saturated (depends on light intensity), so I think it may require 2 transistors? One for on-off and the motor driver?

 

[Nigel Goodwin]

Well, I supose I'm wrong about only one transistor, as Cyn70 may not always be saturated (depends on light intensity), so I think it may require 2 transistors? One for on-off and the motor driver?

I don't know what a cyn70 is, is it a slotted opto-switch, or a reflective opto-switch. By using the gain of the extra transistors it should help to switch the output more positively.

 

[patroclus]

It is a reflective optical sensor with transistor output.
I just was wondering on what transistor configuration to choose, better than these 3 kind of bad designed transistors.

 

[patroclus]

Look at this schematic
**broken link removed**

It also uses BD136 (kind of BD140) to operate a H bridge, with 4,7K base resistor! This takes us to a even lower colector current... I supose this is not coincidence..

 

[Nigel Goodwin]

I had a few spare minutes, so I did some tests. I used a BD132 (PNP), emitter to +6V, collector to a 6V 6W (1A) bulb, and bulb to -ve. I tried different value resistors from base to -ve, and measured the voltage drop between collector and emitter, these are the results:

1K - 2.50V
470 - 1.93V
220 - 1.35V
100 - 0.62V
47 - 0.22V
39 - 0.20V
33 - 0.19V
27 - 0.18V
22 - 0.18V
18 - 0.18V
10 - 0.18V

Certainly 1K looks a bit useless :lol: as low as 27 ohm is required for full saturation - I didn't even bother trying 2.2K!.