position feedback device 2

[pooslinger]

I am controlling a 48in butterfly valve with a 6in dia neumatic cyclinder.I am using two solenoid valves and creative programming to throttel this valve to the desired water flow. I replaced the positioner because it just didnt work for this valve. There is one drawback, I don't have feed back on the physical position. I have to rely on the expected flow at the position it "should" be. I remember somewhere at a trade show a 4-20mA slide rail position feedback device that I can mount paralle to the motion of the valve. Has anybody seen a device like this and where can I buy one. Thanks.

 

[btrueblood]

google for a "string pot" or string (or cable) + potentiometer. Not what you asked for, exactly, but a heckofalot cheaper. Or, roll your own with a constant-force spring mounted to a winding shaft, attached to a multi-turn pot.

 

[VEBill]

If the OP's screen-name ('pooslinger') has anything to do with his question (re: 48-inch valve), then I'd recommend an ultra-reliable solution no matter what the cost.

One wouldn't want any accidents... ;-)

 

[jraef]

The slide-out position feedback device you likely saw was either a Linear Displacement Transducer, abbreviated LDT, or a Linear Voltage Displacement Transducer, abbreviated LVDT. If you Google LDT and LVDT you will hit most of the major suppliers. The difference is only in the technology used to determine the distance movement. LVDTs use a transformer based version that looks at dissimilar voltage feedback, typically used when small distances are involved. LDTs use a technology called "magnetostrictive pulse measurement" that sends a pulse down a wire that makes it "twist" and reflect the pulse back to the transmitter. Then a small magnet on the detection device makes the timing of the reflection different which translates to very a accurate position of the magnet. LDTs can theoretically have almost no distance limit, but are usually available only up to a few feet long off the shelf. It all sound complex, but all of that is transparent to you, you just see a rod and a box at the end with the information you want available to your system.

Here is one of the latter. These are very commonly used in hydraulic valves.

http://www.ametekapt.com/products/sku.cfm?SKU_Id=12066

JRaef.com
"Engineers like to solve problems. If there are no problems handily available, they will create their own problems." Scott Adams
For the best use of Eng-Tips, please click here -> faq731-376

 

[pooslinger]

Thanks jreaf. That was exactly what I was looking for. To VE1BLL you are very perceptive, Thanks for the chuckle.

 

[richs]

Hiya-

Before you wander away, you might look at some of the scales
uses for Digital Read Out. There are some inexpensive
chinese scales that might fit your needs. They have a
wierd digital interface, but have some inexpensive feedback
and display units. You might want to reference the 8"
version found at this site:

http://littlemachineshop.com/products/product_view.php?ProductID=2013

which might provide you the resoltion that you require.

Unfortunately, the DRO displays that are shown do not have
anything other than a visual display. They don't have
a computer interface.

Jraef, what are the prices of the LDT's that you mentioned?
I spent a good 15-20 minutes trying to get a rough
estimate of the cost of some of these units. Second,
what sort of indexing/calibration is required for this
sort of sensor?

With the scales that I have mentioned above, when power is
lost to the devices, there is no "zero index" to them.
In the "interface spec" from the internet from someone
who has "reverse engineered" the interface, the absolute
signed value is set when power is restored to the scale.

Is a similar sort of thing required for the LDTs? Does
the control electronics have to recalibrate itself by
sending down pulses to establish?

cheers,

Rich S.

 

[ScottyUK]

Fisher's 'Fieldview' intelligent positioner is great for this kind of application. We've used them with success on a few awkward valve control applications and have been pleased with the results. They aren't cheap, but they're good.

You could also look at magnetostrictive and LVDT sensors from (e.g.) Balluff, Penny & Giles and Temposonics for position transducers which fit internally to the actuator. I'm a big fan of Balluff's products in hydraulic applications where I need precision and a ruggedised installation. These are fairly expensive too, but if you need precision these are about the best of the mass-market transducer technologies.

Variohm make a variety of resistance-based transducers which sound similar to the type you described. Cost is low relative to the previous suggestions.

http://www.variohm.com/position/index.htm

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

Sometimes I only open my mouth to swap feet...

 

[rmw]

I read the question but not the rest of the thread. So I hope I am not being redundant.

Go to

http://www.jordancontrols.com

to find your feedback device. There are a variety of choices.

rmw

 

[jraef]

Richs,
Look at the original post. He has a 48" valve and the pneumatic actuating cylinder is 6" diameter, not 6" long! The travel distance is probably 2 feet or more.

To give you an idea, here's a similar setup although not pneumatic and only 18" pipe. The pneumatic actuator would go in place of that big spring. He want to be able to detect how far he has moved the actuator to determine the amount of restriction he is putting on the flow with that valve. He would simply put the LDT parallel to the actuator.

To answer your other questions;
About $1000 for a 24" unit
Yes, it recalibrates itself on power up. It takes a few milliseconds. No manual adjustments are necessary, that's one of the beuties of them.

JRaef.com
"Engineers like to solve problems. If there are no problems handily available, they will create their own problems." Scott Adams
For the best use of Eng-Tips, please click here -> faq731-376

 

[ScottyUK]

Although the OP asked about measuring linear position on the actuator, I'd also look at direct mounting to the valve and measure angular displacement of the shaft. There are a few options: potentiometer, RVDT, resolver, encoder, etc. It should be entirely possible to mount the device on a bracket so the shafts are coaxial - a valve that size should have plenty metal to drill & tap a couple of small holes at a non-stressed location of the casting. A resilient self-aligning coupling between the pot (or whatever) would take care of the rest.

It would help to know the following:

Available budget
Intended lifespan
Purpose of feedback (indication only or closed loop control?)
Environmental conditions & susceptability to damage.


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

Sometimes I only open my mouth to swap feet...

 

[zeusfaber]

Before you come to a final decision, think for a moment about how the likely failure modes of each device type might affect you.

Potentiometers are cheap, but they do tend to spend more of their time than you'd like telling you that your actuator is hard up against one of its end stops. You need to know whether this is going to upset you when it happens.

A.

 

[richs]

jraef

You mentioned:
"Look at the original post. He has a 48" valve and the pneumatic actuating cylinder is 6" diameter, not 6" long! The travel distance is probably 2 feet or more."

Opps. my bad. I did read 6" throw.

Yep, gottcha on the mounting of the LDT.

And thanks on the pricing. About what I figured, give or
take 50% to 150% or so......

The little scales mentioned above are dirt cheap, 8" $69.00,
16" $108.00, 1 mil accuracy. They also have 4" and 12"
units similarly "scaled" in price (sorry about the pun).

It opens up a lot of possiblities. Using etched glass and
other linear encoders is an older alternative for the
semiconductor industry. One of the clients I worked with
make laser interferometers, and those are not cheap by
any shape or form ;-)

Since the OP is "programming" the interface, a little more
"creative programming" could be used to interface the
scales. As mentioned, it's not a "standard" interface,
although not too bad. Again as a downside, when power is
lost to the scale, absolute position reference is lost,
requiring a seperate "home" position to recalibrate it.
Probabily shutting down the water supply to the system would
be a "show stopper", but I can certianly think of
alternatives that would circumvent this problem allbeit with
temporary increases or decreases to water flow.

I'm "playing" with a DRO interface design myself, so these
scales are somewhat in the forefront of my mind right now.
I'm using a digital caliper from Harbor Freight for $17.00
as the interface right now (well, as soon as I either get
a good connetor for it or get this little homebuilt
connector I'm using right now more reliable).

Since they are a "hand tool" they seem robust enough and I'm
sure with a suitable housing could be used in an
industrial setting.

I'm not affiliated with either the manufacturer of the
scales nor the littlemachineshop, although I am a customer
of the latter.

It's an interesting alternative. ScottyUK has some of the
real interesting questions that should be answered first.
Indeed, number of units would also be a question asked as
there might be multiple copies of each required, although
it is implied that this is a one off situation.

Cheers,

Rich S.

 

[pooslinger]

The cyclinder has a 28in stroke. The valve is used in a Waste water plant and it controls the flow to a temporary retention basin in a storm event when the influent flow (56 MGD) exceeds the capacity of the final setteling tanks depending how many tanks in service. With all tanks in service max capacity is 33 MGD. A positioner did not give me the reaction time I needed with the mechanical condition of the valve. Solenoid valves gave me the reaction time but had a tendency to over torque the valve making it stick shut during the next storm event. I am limited to 4-20mA because I am using AB PLC5 with an analoge input and a large distance. An LDT wins the prize. Thanks all I am glad I asked.