flow control valves

[BillA]

well, here's how the pros do it at ~$3k per
http://cgi.ebay.com/ws/eBayISAPI.dll...ME:B:EOAB:US:6

it will be interesting to see what accuracy and repeatability I'll be able to hold
- automated testing is ever closer

[Tempus]

wow, that was cheap at 80 bucks.

That for your home testing setup ?

[BillA]

lol
naw, I try and leave this stuff at Swiftech

[9mmCensor]

your a ebay shark. nice find(s).

[Tempus]

Originally Posted by unregistered

lol
naw, I try and leave this stuff at Swiftech

hehe. I thought I was the only one trolling ebay for extra gear at work.

[brucoman]

great economical buy!
ok, they are 120vac actuators, but what are the positioners controlled by? 4-20ma? I can only see the 120 input lines, no controls...then again are these ancient?

[Kobuchi]

Looks like the actual valve is easy to exchange... for more appropriate size and type.

Nice units.

[BillA]

current model, takes digging
4 - 20mA

[brucoman]

Originally Posted by Kobuchi

Looks like the actual valve is easy to exchange... for more appropriate size and type.

Nice units.

usually the trims are the most expensive parts!

[brucoman]

Originally Posted by unregistered

current model, takes digging
4 - 20mA

very nice & proper
near everything we use is conrolled/transmits by a 4-20, works well even in the biggest industrial plants

[jlrii]

Confused here....what are you going to use those for? I don't know much about process automation but those actuators look huge for the size of the valve body. Magnetic actuators perhaps? Are you using feedback? If so how acurate must it be? I may have some older Staefa (Siemens) magnetic valves hanging around....will look at sizes. They are very acurate and originaly worked on 0-20v phase cut, but a wigit is available to convert them to 0-10v (24v power). Unsure if they have any have feed-back. But the actuators are a hell of a lot more managable. Do you plan on monitoring flow and using it to control the valve w/ a pid? If so what resolution do you need, is 8 bit ok or do you need 16 bit?


Edit:Huh...Invensys bought out Foxboro, I was wondering what happened to them, They used to be just around the corner from me, thought they went belly up.

[BillA]

hey jlrii, glad to hear from someone who has some experience
the actuators are "MV-1100 series linear electric valve actuator" consisting of a motor, gearbox w/feedback gears, control/feedback amplifier, etc.
http://www.jordancontrols.com/pdf/Bu...00_MV-1500.pdf
these came out of Oak Ridge, no telling their original use but the valves are for metering; they are similar to a needle valve but have only the slightest taper - perhaps this was to skirt the 1% min valve change

my intended use is as flow control in the range of 0.2 to 5gpm, so I suspect I'll have go get a more appropriate valve(S)
my present task is ginning up a 4-20mA adjustable source via GPIB, they are not cheap

suggestions ?

EDIT
and if you go to Foxboro they say the flow control division was sold Jan 5 to another co
??
I got this info from Jordan

[BillA]

I use a variety of mag flow meters, all of which have a 4-20mA output
using that and LabVIEW I intend to have a semi-automated program that steps through the predetermined flow rates and when equilibrium (as I define it) is reached to step to the next flow rate
- I have a solenoid setup now but would prefer self-adjusting valves

I don't think it is a problem of bit resolution, the limiting factor will be the repeatability of the valve setting

[jlrii]

While I know what GPIB is I am not very familar with the devices available. Are you currently controlling the solenoid valves with GPIB, if so consider allowing a separate pid controller to monitor flow and control the valve and allow GPIB to provide the setpoint via a resistor / voltage network switched by GPIB (dunno if that would offset costs). The devices I deal with are based in building automation systems so I am somewhat confined by their proprietary system as far as direct comunication with GPIB (and my knowelge) is concerned. What types of physical output are you currently capable of? Where you planning on providing a strict position for the valve or allowing flow to control it? I am assuming it needs to be controlled by flow due to the different Cv's of blocks tested.

[BillA]

- the solenoids, each with its oun preset metering valve, are simply switched on and off via a relay card in a scanner controlled via GPIB
- the adjustable metering valve is the other way to do the same thing but using the 4-20mA output of a flow meter and a PID controller set up in software (LabVIEW) to then adjust the valve to the particular setpoint, in terms of flow

- there are numerous scanner cards, which can be for current, voltage, TCs, RTDs, AC for the solenoids, etc.; each of which is connected to the appropriate instrument, which provides the datum back via GPIB to LabVIEW

I'm not so concerned about input and output types, unless something requires a variable resistance input
-> for this Groth is building me a 'digital pot' with a LabVIEW vi, a bit of a chore but coming along

[jlrii]

With a PID controlling the valve repeatability, as far as the valve being in a certain position for a given output is somewhat moot as the position will adjust to obtain the desired flow due to the integral derivative calc, as do fouling and the cv of the block being tested. Repeatability will depend on only the accurcay of the flow meter and loop tuning. An older card which I might be able to "procure" should be able to use a digital pot as a setpoint. Using one of these controllers you could output the flow to another output 0f 0- 10v (easy to tailor output) and monitor it w/ LabView. Resolution is only 8 bit though (meaning only 255 counts), control is via PI...not as nice as PID but they control well. Pband as well as integral are adjustable and the loop can be tuned "live". The base and gain of the inputs are adjustable as well making the inputs pretty much universal, accepting 0-5v, 4-20ma (2-5v w/ a 250Ώ resitor across the inputs) or a resistance input. Wigits you might be interested in exist here in calouge/ section/ output transducers they are fairly resonable consiring the prices asked for this type of stuff. The rtd's / thermistors etc are probaly not acurate enough for your aplication though.

[jlrii]

I was thinking some today and wondered if you are still using the same chiller setup to maintain the block inlet temp at a given set point. As far as how the loop is set up: Does the chiller constantly recirc through a reservoir and a separate pump controls the block loop? If not consider a 3 way valve. Constant temps will be easier to maintain with consistent flow through the chiller. I looked at what I have left for magnetic valves, unfortunately the smallest I have is a 1'' fpt with an iron body.....nuf said there, not exactly a good match. Very accurate valves though, I came across the demo unit for the valves. It consists of a controller its head end board and one of the valves which is cut away to expose its innards. The bottom is drilled out to allow insertion of a dial indicator, which is part of the display. I played with it a little and found for a given input the valve position is repeatable within a couple of thousands of an inch for what it's worth.

[jlrii]

Huh....killed this thread....how about this solution, costs $111.00 @ list from KELE you can probably do better than list though.




The first would not compensate for block differant block cv's

The second solution would provide full PID control via LabVIEW over valve position, as long as you are monitoring flow.

[BillA]

very nice jlrii
and low cost too
Thanks

[jlrii]

Hey, just wondering....have you tried using a speed control an an ac pump. Controls are available that use a 0-10v signal to vary the speed of psc motors....fewer components that way. However most motors don't like to be slowed to less than 50-60% of their full output , motors designed for it will go to 30-35%. Dunno if that range would work for your application.