High accuracy sensors?

[wyfflemunky]

I'm currently in the preliminary design phase of a high accuracy positioning device. I need a position sensor with a resolution equal to or better than .001 inches. Contact is not an issue, but I'd like to consider non-contact options. I've looked around, and found a few things. Ultrasound sensors seem to not be up to the task I am totally new to this, and I have never done any work in electromechanical control systems before. Could someone point me to some useful reference sites, or give me a few pointers? I would greatly appreciate it.

 

[automatic2]

Consider that a cheap ol' printer or scanner can have resolutions of 1200dpi, then consider that it could take a hundred feet to stop a car at 60mph.

Consider the first example. If you go 'home', then step away from that reference position, you can enjoy some fine resolution. If on the other hand you are considering the second example, you'll want a more complex decel, reference resolve, index to position, all wrapped in control algorithm.

Tell us more about the characteristics of the 'position' and something about the inertial load you would like to get to the position.

 

[ScottyUK]

Wyfflemunky,

I suggest you take a look at the Micropulse range of transducers by Balluff. We are using them on a hydraulic servo system and have had excellent results, achieving repeatability of about 0.0015 and good dynamic response. The transducer has a better accuracy specification than we can make use of because we are limited by the hydraulics. We ditched the original RVDT and linkage because it was a pig to set up and had long term drift problems. Both problems are now thankfully in the past.

If you have the patience to set them up, LVDTs are also very good. Take a look at Penny & Giles, for example. There are numerous manufacturers of these transducers.




----------------------------------

If we learn from our mistakes,
I'm getting a great education!

 

[buzzp]

ScottyUK,
I have use Baluff sensors as well. Just had a funny story about their sensors. We were using one in a hydraulic cyclinder (new design) and I was trying to prove the pulses on the bench by applying the appropriate supply, etc. They kept failing. I could never get anything out of sensor. After 3 failures (includes waiting for delivery and of course it was a hot project) and everyone questioning my ability to hook this thing up right. I finally figured out that the pigtail connector was being soldered on backwards at the factory. And some of the other pins don't like 12VDC applied to them so it was constantly failing the units. They felt like idiots but after this was resolved it worked like a champ and have had no problems.
One other consideration for the poster is laser sensors. However, we need more info. Range of sensing? What are we trying to determine the position of? etc.

 

[nbucska]

If you can afford it, try laser interferometer

<nbucska@pcperipherals DOT com> subj: eng-tips

 

[wyfflemunky]

The range is small. The entire device is contained in a 6x6x6 inch cube. The range of measurement is maximum 2.5 inches. It moves in 3 dimensions, so I'll need to fit three of these displacement sensors inside along with everything else. In other words, smaller = better. Money is an issue, but I think we have a lot to work with so expensive sensors are not necessarily out.

Thanks a ton already, you guys have given me material for a hundred google searches so far.

 

[sreid]

Heidenhain makes a steel tape incremental encoder (LIDA400) that has a 20 micrometer pitch (5 micrometer edges). They have a new one with a small read head. They have built in limit switches and a home pulse.

 

[automatic2]

are you looking for the position, or are you sending something to a position.

 

[wyfflemunky]

I'm sending something to a specific position, and using the sensors as feedback and to ensure it's positioned accurately.

 

[automatic2]

in many cases, sending an object to a position has higher accuracy than detecting it's position. If enough resolution is built into the drive, very high accuracies are possible. Much of this has to do with a stable reference. As the object departs it's 'home' position, a reference is established. In seeking it's location, the detection device(s) must have similarily stable references.



What type of transport/transmission will you be using to place the object?.