Alt-BEAM Archive

Message #00484



To: beam@corp.sgi.com
From: Steven Bolt sbolt@xs4all.nl
Date: Tue, 16 Feb 1999 17:18:08 +0100 (CET)
Subject: [alt-beam] Re: What is the connection between current and torque?


On Tue, 16 Feb 1999, jim cardona wrote:

> Increasing VOLTAGE to a DC motor will increase the motors speed. (only
> applies to DC)
>
> There is no way to change the internal resistance of the motor, Therefore,
> by changing current, you must mean that you want to add external resistors?
>
> What this will do is reduce the voltage at the motor terminals, which again,
> will reduce the speed.
>
> The only way that I know to increase torque is to add some sort of reduction
> mechanism to the shaft (reduction gears for example). An easy method that I
> used was to run a rubber band from the motor shaft to a large wheel with a
> groove in it. The motor was about 1000 RPM, so the wheel turned about 50
> RPM. With a much increased torque.

Actually, the starting torque of a reasonable motor does increase
with current, and the internal resistance varies with the rpm.
One has to take this into consideration when designing Solar Engine
driven `bots, as they have to make the most of the limited charge
in their storage caps. That's why the often used 2N3904 is not
really an appropriate driver in most cases.

A not untypical small recorder motor on my desk has a resistance of
less than 10 ohms when stopped, which means that ideally switching
2.5V to it will cause 250mA to pass through its windings. When it's
running free, it uses just 20mA, indicating a `resistance' of 125
ohms. A pager motor scores similar values and may actually use more
than 20mA (up to 50mA) when running free.

With a good storage cap in the usual range, a Solar Engine will
dump most of its energy in the first fraction of a second. For part
of that time, the motor isn't running. A high current is required
to get it going, which the transistor driving it must be able to
handle immediately and fully. The voltage over that transistor must
drop like this:

Vcap ----
|
|
| (VCEsat is what remains over the transistor
VCEsat |____ when it's fully conducting)
0 ----------------

And not like this:

Vcap ---
\
\
\
VCEsat \____
0 ----------------

Because in that case, some or most of the initial power `kick' is
used to warm up the transistor.

Now let's look at the characteristics of the 2N3904, as determined
using a little tool I posted some time ago:

Ic setting Ib Ic VCE

2N3904 1 10uA 100uA 0.1
2 50uA 1mA 0.1
3 500uA 10mA 0.1
4 5mA 100mA 0,24
5 10mA 240mA 1.2

Ib = base current, Ic = collector current, VCE = voltage between
collector and emitter, and the settings 1 - 5 switch resistors
between collector and Vcc from 47K downwards to 15 ohms. (The
diagram is at http://www.xs4all.nl/~sbolt/Other/hFEmeter.gif)

We can see that the transistor is performing reasonably well up to
setting 3, though base currents need to be rather high to get the
VCE where we would like it for our applications (below 0.1V). At
Ic setting 4, VCE rises no matter how high the base current, and at
setting 5 it is clearly overloaded.

But that small recorder motor wants a current higher than 100mA to
accelerate from stop. So let's have a look at the BC337 (npn like
the 2N3904; a good equivalent would be the 2N2219 or 2N2222):

Ic setting Ib Ic VCE

BC337-25 1 1uA 100uA < 0.1
2 10uA 1mA 0.1
3 100uA 10mA 0.1
4 5mA 100mA 0.1
5 10mA 300mA 0.3

This transistor clearly has much less trouble achieving a low VCE.
At Ic = 10mA it needs just 1/5 of the 2N3904's base current to do
so, and saturation is easy to achieve at 300mA. Thus it will do a
much better job accelerating a small motor from stop.

Like the BC549C, the 2N3904 is clearly intended for very small
collector currents. But it isn't as good as the 549C, as this
table shows:

Ic setting Ib Ic VCE

BC549C 1 1uA 100uA < 0.1
2 5uA 1mA < 0.1
3 50uA 10mA < 0.1
4 5mA 100mA 0,2
5 10mA 240mA 1.2

I also determined best hFE for all three transistors:

BC549C: 570
BC337-25: 250
2n3904: 200

The first two values are typical for these types. According to my
tables, the BC549C should score between 420 and 800, the BC337-25
between 160 and 400, and the 2N3904 between 100 and 300.
Finally, the approximate US$ prices may be interesting:

BC549C: $0.13
BC337-25: $0.25
2n3904: $0.15

Best,

Steve

----------------------------------------------------------------------
# sbolt@xs4all.nl # Steven Bolt # popular science monthly KIJK #
----------------------------------------------------------------------




------------------------------------------------------------------------
eGroup home: http://www.eGroups.com/list/alt-beam
Free Web-based e-mail groups by eGroups.com

Home