@Danko, you were once kind enough to help me model a split capacitor induction motor. Now I need to model a permanent magnet, brushed DC motor. The motor is of the 90V type exactly like this one:

​

Its rotor has 8 coils total, and uses ordinary carbon brushes. How do I go around modeling this motor in LTSpice? Do I just measure its inductance and resistance and represent it as a simple inductor with the measured values?

Also, what's the best way to filter an H-bridge to drive this motor? Would something like this do the job?

​

My intention is to PWM the high-side mosfets, that is why those are the only ones with RC filters installed. That is, C5 and R6 for M1, and C6 and R7 for M4. With R6 and R7 = 51 ohms @ 1W, and C5 and C6 = 2.2nF

They seem to be working well in the simulation, but I'd like to know if they're being used correctly.

 

Now I need to model a permanent magnet, universal type motor. The motor is of the 90V type exactly like this one:

Are you sure ? Permanent magnet - Universal motor?
Generally a Universal motor is a wound field with a brushed rotor connected in series, and used on AC as well as DC.
It appears to be a PM DC motor.
Max.

 

Are you sure ? Permanent magnet - Universal motor?
Generally a Universal motor is a wound field with a brushed rotor connected in series, and used on AC as well as DC.
It appears to be a PM DC motor.
Max.

My bad, Max. I misquoted. It is indeed a permanent magnet brushed DC motor as you've just said.

EDIT: I've just edited and corrected my first post to make sure it doesn't confuse future participants.

 

@Danko, you were once kind enough to help me model a split capacitor induction motor. Now I need to model a permanent magnet, brushed DC motor

R1 - depends on mechanical load, determines current of motor under load.
R2 - measured resistance of motor [on motor terminals].
L1 - measured inductance of motor [on motor terminals].
C1 - depends on inertia of rotor+inertia of mechanical load, imitates inrush current, imitates generation mode.

 

So for a 1/10 HP motor (74.6 W) all I have to do to calculate R1 would be to apply ohm's law, right?

But what about C1? Assuming that a normal motor's inrush current lasts for 1/4 second, and it's in the order of about four times its rated current, do I just adjust its value empirically until it more or less imitates that condition?

 

I may have posted this before, but there is a fairly decent representation of a Bridge circuit http://tahmidmc.blogspot.com/2013/01/using-high-low-side-driver-ir2110-with.html
To avoid the high currents associated with inrush, when using a PWM controlled H-bridge, the acceleration (mean current) can be controlled in order limit the high current associated with the application of immediate full voltage.
Max.

 

I may have posted this before, but there is a fairly decent representation of a Bridge circuit http://tahmidmc.blogspot.com/2013/01/using-high-low-side-driver-ir2110-with.html
To avoid the high currents associated with inrush, when using a PWM controlled H-bridge, the acceleration (mean current) can be controlled in order limit the high current associated with the application of immediate full voltage.
Max.

Thanks, Max. I'm already familiar with Tahmid's blog. It is indeed a very useful reference. Thanks for sharing.

 

BTW, I did an empirical test of current of a unloaded motor of around that size using a variable DC voltage supply in order to start at 0v and gradually increase the rpm up to the rated rpm value.
The measured current was a identical low value from start throughout the range up to maximum.
Max.

 

BTW, I did an empirical test of current of a unloaded motor of around that size using a variable DC voltage supply in order to start at 0v and gradually increase the rpm up to the rated rpm value.
The measured current was a identical low value from start throughout the range up to maximum.
Max.

I assume you increased the applied voltage very carefully. A motor's coil behaves almost like a short circuit when first powered up. It's the back EMF that works as a limiter of sorts that prevents it from overheating while unloaded, am I correct?

 

Correct.
The initial stationary current is dependent on armature resistance and actual voltage applied.
The rpm is decided by the BEMF coinciding with the applied voltage value (almost), when equal, acceleration ceases, any load will lower rpm and will increase this voltage difference and cause higher current to flow, (load dependent).
Max.

 

Correct.
The initial stationary current is dependent on armature resistance and actual voltage applied.
The rpm is decided by the BEMF coinciding with the applied voltage value (almost), when equal acceleration ceases, any load will lower rpm and will increase this voltage difference and cause higher current to flow, (load dependent).
Max.

That was a lot of useful info in very few words, thank you.

I assume that for the same reasons a vacuum cleaner goes berserk when its hose is plugged? because the normal load (the power required to suck air in) has been removed, and therefore its BEMF (and normal bearing and rotor air friction) becomes its sole rpm limiter?

 

Basically yes, The Vacuum (Universal motor) is a little of a different species, it is a series connected motor which tends to operate in a run-away condition due to field weakening, this style of motor tend to have very high starting torque, hence their use as Automotive starters.
The high armature load current also flows through the field.
If the load is reduced from it normal loaded condition it can run away, in some cases to destruction, this is why even a shunt wound field motor often has a field loss detector in order to prevent a possible runaway condition if this occurs.
Of course in a P.M. field this cannot occur.
In the case of the vacuum style radial suction pump it has, the load is very much decreased when the intake is blocked.
Max.

 

So for a 1/10 HP motor (74.6 W) all I have to do to calculate R1 would be to apply ohm's law, right?

You are right. But better way is measure current with real mechanical load and calculate R1.

But what about C1? Assuming that a normal motor's inrush current lasts for 1/4 second, and it's in the order of about four times its rated current, do I just adjust its value empirically until it more or less imitates that condition?

You are right. In DC simulation it should becomes 1/4 sec (if you belief this value) by adjusting C1 value. In practice it is better to measure this time by oscilloscope, with connected real mechanical load. Load may contain, for example flywheel, then time may be some seconds.
By the way, any mechanical load has its own inertia, and it will increase inrush current time.

 

​

R1 - depends on mechanical load, determines current of motor under load.
R2 - measured resistance of motor [on motor terminals].

It is wrong. You should measure resistance between terminals of motor, using some ohmmeter.

I thought that R2 is the value that should be measured directly from the motor? ... and I meant to say that R1 would be calculated as in:

74.6W at 90VDC (rms) -> I = P/V -> I = 0.83A at rated motor power​

Therefore:

R1 = V/I -> R1 = 108.43 Ohms, which would be the equivalent rated continuous load.​

Am I right, or am I wrong?

 

Am I right, or am I wrong?

I changed my previous post, because I mix up R1 with R2 and my last post was about R2.
Sorry.

 

I don't have the motor with me yet. But I will have it by monday. I'll be back then with the measured values and then see how it goes.

Thanks!

 

Here is more accurate model of permanent magnet motor:

R1 - depends on mechanical load, determines current of motor under load.
R2 - measured resistance of motor [on motor terminals].
L1 - measured inductance of motor [on motor terminals].
C1 - depends on inertia of rotor+inertia of mechanical load, imitates inrush current, imitates generation mode.

 

The motors are here!

​

Measurements are:
R2 = between 12.5 and 15 ohms
L1 = between 25.5 and 26.0 mH

Measurements were done at 1 KHz. The variations are there because of the carbon brushes (and also possibly because of the presence of the permanent magnets at the stator), and were manifest when I slightly rotated the motor's shaft, so as to test it in different positions. What I'm going to do, is feed the mean values into the sim and see how it goes. That is 13.75 ohms for R2, and 25.75 mH for L1.

As for R1, following the math I used in post #14, but this time using the data shown in the motor's plate: R1 = 90V/1.1A = 81.82 ohms.

​

Now I have to determine a proper value for C1. But as I've already mentioned, I'm going to do that empirically assuming a 1/4 second motor startup time. I'll later see if I can measure that condition in real life.

 

Question, shouldn't the value for L1's series resistance in LTSpice be the same as R2? That is, shouldn't we take R2 out of the circuit and instead feed the 13.75 ohms value into L1 instead?

 

Here's my first sim, using the previously defined values, and 10,000 uF for C1:



It's interesting to see how spikes of up to 200V begin to appear after the first 1/4 sec.