Is there a way to power an LVDT such that it could be used as a linear motor, or a linear actuator?
I need something relatively fast like a solenoid, that has a 3-3/4" stroke and can be accurately positioned to act as a microwave waveguide tuner.
Where I would probably put a diode in the waveguide near the dummy load, and when reflected radiation is detected, move the tuner to scan for the position that minimizes reflected power.
I would use a solenoid, but I don't know if they can be accurately positioned. Maybe a solenoid in combination with an LVDT?
I am building a microwave plasma jet and plan to inject glassy ceramic material into the flame to be melted and deposited. I suspect that varying rates of material deposition will change the load impedance and require continuous tunning.
In theory, yes. In practice, no. LVDTs, like all measurement devices, are optimized to have the least possible effect on the systems they measure. A typical tachometer is a poor motor, in part because the winding resistance is too high, and because the magnets would be weakened by high currents.
There are other devices you can use. Google for "linear actuator", and think about making use of rotary actuator. (Voice coils are used to position the head in some disk drives. They can't be all that bad.)
In theory it's possible, however the LVDT windings are too small to generate any appreciable force. BEI makes some nice little voice coil actuators which can be run from a linear amplifier such as Trust Automation, or Quanser. Some form of feedback will be required, LVDT, potentiometer.
Actually, if you are in scrounge mode the voice coil actuator from a disk drive wouldn't be a bad thing to use. Voice coil actuators are just the bee's knees when you need precision, speed, and when you're not pushing against a big load.
I have a few old 5" full-height 5-megabyte hard disks with stepper motors and ban-around-a drum drives to position the head. 050 is welcome to all or part of one, including the drive transistors.
"relatively fast" meaning what? What the inertial load? What's your bandwidth requirement and required position accuracy?
You can do some simple calculation assuming sinusoidal movements to determine your power requirement. This will help to select an approproate actuator. Check out the following:
formatting link
3.75 inches is a large stroke. You'll have a hard time getting a bandwidth larger than 100 Hz. BEI has some state of the art commercial linear actuators and lots of good application notes. Their model LA19-40-000A as a +/- 0.9 inch motion (1.8inch total stroke).
formatting link
or maybe a linear motor is better for your application. You can searh on "fast steering mirrors" (FSM) for info on fast actuators to get other ideas, but nothing with a 3.75 inch stroke.
My junk pile turned up a linear actuator with a 5" stroke. It consists of a six-lead rotary stepper motor with hollow shaft that is threaded
1/4-16 acme. (Remarkably little backlash. Loaded ball nut?) At 24 steps/turn -- a wild guess -- the resolution is about .0026", although the backlash may exceed that. Half stepping will halve that.
There is a driver chip in the package (and a spare), but no wiring diagram.
Hurst Mfg Corp Instrument motor Model No LAS Part No 3602-001
12 Volts DC
Hurst is a division of Emerson Electric. ...
I found the spec. It's still in current manufacture. With light load, you can get about 1/4"/sec. Email if you want it, explaining why you can't buy one on your budget.
To be truthful, I don't yet know what kind of (actual) response time may necessary for this application, or what level of control I will be able to acheive with the material flow. I do know that the tunning element is a 1/8" diameter pin of tool steel that needs to be inserted 1/2 of way into the
3-3/4" deep waveguide from the waveguide inner surface level. No forces to counteract except friction, gravity or a spring. Frictional forces should be negligible, but the inertial load depends on how fast or how far I have to compensate for a match or mismatch.
I am considering using an aluminum funnel type device with a needle valve, gravity feed and external vibrator to control the flow of grain sized heavy materials, but fear that this may not be as effective a control for lighter massed fine powders. Perhaps a vaporized slurry will work better in this instance to keep flow relatively constant.
As far as positional accuracy, the actual position is for reference to well characterized flows only. What I am really after is to maximize the power to the plasma and material flow by minimizing any reflected power that may occur for "whatever" reason. I "think" that this is going to depend on the material flow, but I won't truly know for sure until after I have built the machine and have done some experiments.
I am trying now to think of some worst case scenarios, and how they might be mitigated before I have any real experimental data to consider.
Very Good.
3.75 is the width or depth of the waveguide, or the maximum travel that the pin could possibly make (worst case). In reality 1/2 that distance is the position that creates the maximum disturbance in the microwave radiation field, which is all that is really necessary (1.875 inches plus a little extra for overshoot and assembly tolerances). But then again I may want to change the frequency from 2.45 GHz to 900 MHz in the future and be able to use as many of the same parts as possible.
Yes. One of these might work with the addition of a lever mechanisim to extend the stroke length.
Ah yes. I remember these from laser light shows. I'll take a look.
Linear actuators or voice coils seem to be the general consensus at this time. Anyway I have yet to fab an isolator, which needs to be completed before I can even turn the machine on. That is another project all in itself, so there is plenty of time for me to think about what I am going to do here.
Probably more appropriate, although when it gets to the plasma I fully expect everything to be vaporized... But also before that with the correct nozzle, it could be "atomized" as well.
No. The linear actuator is the only plan at this time to drive the stub tuners.
Thanks Jerry, but I have a couple of these here as well. I'm placing my bets that I can control the feedstock well enough that the linear acuator with the stepper motor will have enough granularity and speed to keep an accurate hold on the tuning paramater. I feel pretty good about this right now. As soon as it cools off outside (102 Deg F.), I am going back out to the garage to cut some more steel for it.
Actually I have both stub tuners and sliding shorts in the design. The sliding shorts terminate each end of the waveguide, and are moveable to make sure that the end reflections are exacly 1/4 wavelength from the microwave tube element and the focusing nozzle. Once these have been adjusted manually and set, they are locked into place and will never move after that. Three stub tuners are located between the magnetron element and the circulator. One of these will be selected to be driven by the linear actuator. The other two are manually operated. The idea here is that the two manual tuners can be set for rough or coarse adjustments depending on the feedstock requirements, and the third stub tuner will be driven with the linear actuator for fine / variable adjustments while processing.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.