Speed measurement Test setup

Thread Starter

phuzionz

Joined Dec 5, 2008
47
Hello,

We want to make a whole test setup that drive a pulley with a preset value. Thus I'am looking for a DC motor, 6000rpm, 1n/m.I think that a DC motor is much easier to control. I hope that a encoder is included in the motor itself. Most of the time is that encoder signal, a DC voltage, is that right?
Have sombody already experience with that. I somebody have some good references, please let me know.

Jef
 

leftyretro

Joined Nov 25, 2008
395
On the large industrial motors and compressor used in my refinery they just used a passive magnetic pickup (variable reluctance) looking down at a 60 or 30 tooth gear to get a frequency output pulses that were then counted in a simple digital controller. The probes were installed in threaded holes in the machines outer case and set the probes to about .005 -.020 inch clearance from the outer circumference gear face. There were typically 2 or 3 probes installed and the measurement was used for speed control feedback as well as a safety over-speed trip protection system. Very simple, accurate and reliable.

http://www.daytronic.com/products/trans/t-magpickup.htm

http://www.woodward.com/pdf/ic/82510.pdf



Lefty
 

jpanhalt

Joined Jan 18, 2008
11,087
Controlling a dc motor @ 6000 rpm is no an easy game. You need acceleration ramp and a very sofisticated control due to the high frequency preduced by the encoder turning at that speed. (Assume an encoder with only 360 clocks per turn at 6000 rpm gives a clockout = 2.1 Mhz)

It is not clear if you need rotational movement or linear translation nor the kind of accuracy you need in positioning.

Alberto
Can you review that calculation? 6000 rpm is only 100 revolutions per second or 36,000 ticks per second if your encoder gives 360 ticks per revolution.

In my experience with RC model DC motors, controlling a 6000-rpm motor is no harder than controlling a 2000-rpm motor.

John
 

jpanhalt

Joined Jan 18, 2008
11,087
@Alberto, I think you missed my point. 6000 rpm is not particularly fast for a DC motor, either brushed or brushless. It is quite easily controlled using common controllers. The geared BLDC in one of my smaller models goes well over 6000 rpm and it draws well over 400 W. The controller is about 3 X 2 cm. Of course, any design has to consider the masses and power needs, but that is just part of engineering.

John
 

Thread Starter

phuzionz

Joined Dec 5, 2008
47
Thank you for all the answers.

Which motor type i gonna use depend on several items.
-speed(max 6000rpm) and a torque of 1n/m.
-low cost driver to control the speed. I mean, if i use a DC motor, i can control his speed with varying the DC power supply, i think very simple and effective. But not sure if it is possible.
- low cost feedback device to measure the speed. When it's possible, i want to use something without a digital card, because it is most of the time not cheap.

So every motor type is possible for me, but i thought that a DC was much easier.

If a universal motor is acceptable, you might salvage one from an angle grinder...
How can i feedback then the speed?
 

russ_hensel

Joined Jan 11, 2009
825
I think an encoder is overkill for this. It really allows you to control position. All you really need is a tachometer ( sp ) and some sort of feedback loop. How accurate does your speed control have to be?

Another approach is to use an AC motor, probably poly phase and use an ac drive, typically I think this is called a brushless motor. Like the motor in a hard drive. Like the motor in a hard drive it may be hard to drive, there are chips for this. The RC guys who build their own motors and controllers probably have a lot of info on this.
 

Thread Starter

phuzionz

Joined Dec 5, 2008
47
I have no idea on how accurate i can go. Normally,the accuracy i need is 6000r/m +2 and -2. I have also to consider the price of the whole installation.

Now i was reading in a book, and there i have to choose the maximum power.
Pmax = Mmax * n max = 0.55 * 6000rpm = 3312W
Is that realistic?
 

beenthere

Joined Apr 20, 2004
15,819
That is very tight control - 6000 RPM (how we express revolutions per minute) +/- 2 is .033%. You will probably need multiple pulses/revolution to sense RPM accurately enough.

Where did you get your power figure? My physics book suggests that power is related to torque and angular displacement, so I get about 328 watts for the motor running at 6000 RPM. Is there a load on the system, or do you only need to hold the speed steady?
 

Thread Starter

phuzionz

Joined Dec 5, 2008
47
Ok, so it is very expensive and difficult to reach that accuracy.
I don't know, but a "standard" tacho, how accurate is that?
For good understanding, what you mean with a tacho, is that a meter that converts directly the "frequency" into a value displayed on a digital screen(7 segment display).

Well, the torque that i gave in the first post was the maximum torque.
That means that he need that to start the pulley in the beginning, afterwards he need probably less torque.

On the pulley is no load connected.

Well, i think that i did something wrong with the calculation.
The "n" in the formula must be in Rad/s.
P = M *n = 0,55 * 628 = 345W

When i see this power value and look parallel in a catalogue, I can't find a DC motor for so high power rating. I think i have to find alternatives.
 

beenthere

Joined Apr 20, 2004
15,819
I think we can stop being concerned about power - http://www.rapidonline.com/producti...0+rpm+Miniature+motor&moduleno=61542#techspec.

The Google search term was "8000 rpm dc motor". I used 8000 RPM because I wanted to find a motor with some ability to exceed 6000 RPM. It should control better at a point under the absolute maximum performance.

If that motor can run 8000 RPM with 230 ma input current, even at 6 volts the power is only 1,38 watts.

Tachometry can be done in several ways, but a disk turning with the motor shaft is easy to arrange. Either slots or holes in the disk will let an optical device count a circuit with each slot or hole passing.

With a good time base, you can track the interval between holes to get speed, or you can simply count pulses for a accurate duration of time to get a value that directly relates to motor speed.

An electrical (analog) output would be less accurate without going to great expense.

A microprocessor can count pulses and control DC motors by means of pulse width modulation (PWM) techniques. That is least expensive for the parts, but more expensive for you unless you already can do programming.
 

Thread Starter

phuzionz

Joined Dec 5, 2008
47
If that motor can run 8000 RPM with 230 ma input current, even at 6 volts the power is only 1,38 watts.
Well yes, But you have also to consider the maximum torque?
If he can't handlethe torque during starting, it will burn your motor.


Tachometry can be done in several ways, but a disk turning with the motor shaft is easy to arrange. Either slots or holes in the disk will let an optical device count a circuit with each slot or hole passing.

With a good time base, you can track the interval between holes to get speed, or you can simply count pulses for a accurate duration of time to get a value that directly relates to motor speed.

An electrical (analog) output would be less accurate without going to great expense.

A microprocessor can count pulses and control DC motors by means of pulse width modulation (PWM) techniques. That is least expensive for the parts, but more expensive for you unless you already can do programming.
Ok, thank you for the tip, but is it really needed do control the motor with a PWM signal?
 

Thread Starter

phuzionz

Joined Dec 5, 2008
47
I was just looking in a book to DC motors.
For instance the EC45 from Maxon Motors has a starting current of 85A, that means that i have to buy also a very expensive powert supply?
Can i handle this problem in another way?
 

Thread Starter

phuzionz

Joined Dec 5, 2008
47
I have a lot of choices after looking on the internet and contacting suppliers.

- DC brushless motor with control and tacho with 500 pulses.
I have to read the analog signal from the encoder into a microcontroller to convert it to rpm.
- A servo motor with controller where i can control the speed with a potentiometer.
- brushed DC motor(3000rpm) with reduction to reach 6000rpm + tachodevice + controller.

Have someone from maybe other suggestions. It is my first experience with motors, thus a second opinion is welcome.
 

Thread Starter

phuzionz

Joined Dec 5, 2008
47
Has nobody an oppinion? Servomotor is the most expensive in this case, i don't know why but it is what the suppliers sayed to me.
 

Bernard

Joined Aug 7, 2008
5,784
What is the size and weight of the pulley?
Under steady state conditions it seems the only power needed is to ofset losses,bearings, air resistance,IR, so forth. Need a motor substantial enough to support pulley or added complication of adding bearing system for pulley.
One example of" tacho " on a motor generator; brushless generator, AC output: approx. 2V per1000 rpm. Can be used as a V or pulses per revolution.
To start, use a constant current power supply set for max. operating motor current.No time limit was stated to bring up speed, so no hurry. As speed builds up, no load current drops off signaling time to switch to voltage speed control.
 
Top