DC Motor

Thread Starter

mlkcampion

Joined Nov 16, 2006
59
hEY fORUM
Iam trying to use a Power mosfet (IRL520) to control the
speed of a DC motor, i have setup 2.2v on the base of the
transistor. The source is connected to ground. I then put the 24v
to the positive of the motor and the return from the motor is attached
to the drain of the transistor.
When i power up the circuit, the motor starts and then continues to
speed up until the full 24v is dropped across the motor? Whay is this
and how can i stop it. The circuit is pretty similar to the circuits
you would see interfacind a motor to a PWM output.

Thanks
Michael
 

beenthere

Joined Apr 20, 2004
15,819
Hi,

From your description, it shouldn't be happening. The IRL520 should not go into saturation until the gate voltage is higher. Are you able to verify the polarity of the 24 volts? There is an inherent diode in the internal structure of the hexfet that will conduct with reverse voltage across it.

The part about the motor starting and speeding up is wierd. If this takes some time, I wonder if there is some path to the gate that will allow the voltage there to increase over time? A post of your circuit would be very helpful
 

Thread Starter

mlkcampion

Joined Nov 16, 2006
59
Hey
Yea it takes about a minute to speed up, it would seem
that the reluctance of the motor to start causes a
change in the mosfets drain-source characterisics, this in
turn causes the motor voltage to change therefore speed up,
the motor attempts to speed up and causes further change
in the mosfets characteristics and so on. Basically with the 2.2v at
the gate i should see approx 8v across the motor drawing appro 150mA.
Iam unsure how to upload the picture. The circuit is similar in many ways to
circuits used with PWM. Iam certain of the polarity of the 24V. I simulated
this circuit and got the results i expected even though i replaced the motor
with a 50ohm resistor, i was unsure what the effect the motor would have and it appears it has quite a significant effect. Maybe it's an impossible circuit with the gate voltage been so low!!!!

Thanks
Michael
 

beenthere

Joined Apr 20, 2004
15,819
Hi,

Dunno. I've only used hexfets as switches, never as linear devices. The published curves says they're good for it.

Do some experimenting with the gate circuit - the effect/problem about has to be in it.
 

richbrune

Joined Oct 28, 2005
126
Is this a brushless motor? I don't know why, I've never seen a DC motor controlled by an analog signal to the gate to drive motors, they "latch up". I've always seen Pulse Width Modulation, or pairs of MOSFETS, or other means to control the motors. I hope you post the answer when you get it---Thanks, Rich
 

n9352527

Joined Oct 14, 2005
1,198
If you look on the datasheet, the typical output characteristics figure, at Vgs = 2.2V, the Id ranges from ~0.1A to ~0.2A depending on Vds value. In other words, the Vds will depend on the Id value for a constant Vgs.

This is what you actually observed, when your motor first started, it drew more than the stated current (150mA @ 8V?), and probably would hit the ceiling of Id(sat) ~ 0.2A. When this happened, the current is limited to ~0.2A and the motor started to spin slowly because it didn't have enough power.

When the motor was turning faster, the current drawn decreased, and the Vds also decreased. When the motor is not loaded, it might only draw a bit less than 150mA. Looking on the same figure, you can see that the Vds would be a little bit less than 0.2V. So that it appeared that all the supply (24V) dropped across the motor.

It is just not straigthforward to control the output voltage through Vgs when you are crossing between the linear and saturation regions. Also, you might not have a linear Vmotor to Imotor relationship, which would not give a straight load line, and would make it more difficult to control the Vds against Ids with Vgs.

Bottom line is, both Ids and Vds depend on Vgs. Ids and Vds for a constant Vgs have two regions, linear and saturation and the relationship between Ids and Vds at constant Vgs depends on the region.
 
Top