Shop OBEX P1 Docs P2 Docs Learn Events
Motor Control with 4-20mA signal — Parallax Forums

Motor Control with 4-20mA signal

bhaanbhaan Posts: 37
edited 2005-07-14 13:25 in BASIC Stamp
I am trying to control a motor with a 4-20mA signal.· I am aware of the PWM feature of the BS2, however that is not a "true" analog signal and I would rather use other methods.· On the schematics of the actuator I am trying to control it shows "y+" and·"y-"· terminals that I have to connect to.· Can I use a a digital Potentiometer to control the actuator, and if so, is the "y-" terminal supposed to be connected to ground?· I do not have a great deal of experience with electronics so this is a little confusing to me.· Any ideas or advice would be greatly appreciated.

Thanks,
Brian

I have a pdf of the technical specs of the actuator, if it would help I could post the pdf.

Comments

  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-01 17:34
    Yes please include the pdf, the more info we have, the better we can help you.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  • bhaanbhaan Posts: 37
    edited 2005-07-01 20:27
    The pdf is located at:·· http://www.sbt.siemens.com/HVP/Combustion/products/Documents/155517.pdf

    My circuit board in the motor is the: AGA56.41A17

    The bottom of page 9 and then page 10 has the actual schematics.· The 4-20mA signal is the one I have to use and is the part I am not sure how to do.
  • OrionOrion Posts: 236
    edited 2005-07-01 21:54
    The -y terminal should go to your signal gnd and y+ will be the input for your 4-20 signal. Most HVAC actuators I have seen sink the current from the signal. You should be able to use a digital pot to drive the actuator, but ill let some else post you a circuit. The input impedance on that one looks to be 300ohm.

    That thing looks like a honeywell clone, complete with all the cam and micoswitch nonsense. If you’re looking for better actuator with programmable limits, signals check out the MFT series from Belimo. www.belimo.com. If you need some pm me, I work for a platinum dealer and can good prices.
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-01 22:02
    Ok I've done a brief once-through the specs. I was under the impression that a digipot would work, but looking at the specs of the AGA56.41A.. driver board on page 24 concerns me. If understand the specs correctly the current source supplying 4-20mA must have an output impedance less than 300Ω, your not going to achieve this with the digipot without throwing away the majority of the digipot's range (they start at 1k&#937[noparse];)[/noparse] and having to supply a special voltage to the digipot to achieve the 4-20mA range. Let me think/look at it some more, the digipot may still be the best solution, Im just not sure at this point. Orion seems to have some experience with what you are trying to accomplish, between us hopefully your questions can be answered.
  • OrionOrion Posts: 236
    edited 2005-07-02 07:23
    Here is some links I found trolling around on Google:


    http://www.maxim-ic.com/appnotes.cfm/appnote_number/807 - see figure 6

    http://www.maxim-ic.com/appnotes.cfm/appnote_number/722

    http://www.discovercircuits.com/PDF-FILES/NewPDF/CURRENTSOURCE1.pdf - replace 10k pot with a digital?

    http://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=9944 - Precision Programmable Current Sources Use Digital Pots

    Post Edited (Orion) : 7/2/2005 5:01:51 PM GMT
  • bhaanbhaan Posts: 37
    edited 2005-07-02 16:02
    I really appreciate your ongoing help Orion and Paul.· I am a very inexperienced person when it comes to electronics, however, I was thrown into having to control this particular·actuator.· I will start studying the links Orion posted, they seem pretty interesting.· I was also wondering if maybe an Input/Output module would be the answer to getting a quality 4-20mA signal?· Cost isn't a big concern yet at this point, I just would rather not have to result in buying a PLC.
  • OrionOrion Posts: 236
    edited 2005-07-02 17:31
    bhaan said...

    · I was also wondering if maybe an Input/Output module would be the answer to getting a quality 4-20mA signal?·
    I wouldn't worry too much about getting a “quality” signal.· That actuator and most others designed for similar uses only have ~256 positions over 90 degrees that they can be driven too.· The type of valve or damper its driving will have the biggest impact on the flow control.· Each type of valve/damper will have different flow characteristic that will have to be taken in account if you need precise linear flow control.· If you can get 8-bit resolution on your 4-20ma signal you should be fine, let the software deal with the rest.· What is your application for this actuator?
    ·

    Post Edited (Orion) : 7/2/2005 5:41:31 PM GMT
  • bhaanbhaan Posts: 37
    edited 2005-07-02 18:04
    The actuator controls a butterfly valve.· Once I am able to control the actuator I will get into studying the valve more, but controlling the actuator is, for me,·the hardest part of the project I am working on.·

    I am controlling a test furnace to follow a specific temperature curve and send all·the data to·a PC. ·I have already written software to communicate with the stamp and save all aquired data to a spreadsheet, but I still have to calibrate the PID part of the software after I get everything figured out.· I'll have the software do all calculations and then have it determine a setting to send to the BS2 so it can control the actuator.·

    Would D/A converters be the answer to sending my 4-20mA signal?· I was looking for one·at maxim-ic but I was a little confused with all the different options they offer.

    Paul said something about output impedance earlier, I am unclear on what output impedance is?·
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-02 20:44
    Output impeadance is what the equivalent resistance looking into the source (the circuit supplying the 4-20mA) from the perspective of the load (the actuator). In the digipot scenario the output impedance is the value of the resistor you use to generate the current. So if you are using a 1kΩ digipot, you shouldn't use a value greater than 300Ω to meet the requirements. This requirement is pretty low, meaning your load is probably reactive (inductive or capacitive), a high source impedance means a longer settling time (slower reaction) when you change the control value.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
  • OrionOrion Posts: 236
    edited 2005-07-02 22:04
    Wouldn’t using a circuit like the below help the source impedance (using an amp and a transistor)? Sorry I'm not an analog design nut.

    http://www.discovercircuits.com/PDF-FILES/NewPDF/CURRENTSOURCE1.pdf

    Another option, using a transducer card like this one (PWA-2A)
    http://www.kele.com/olcat/OT8/PWA.pdf
    to convert from a digital pulse to 4-20ma. That one will drive up to a 650ohm load. I have used these many times to interface to 4-20ma devices. The pwm signal they accept has a long cycle time, so a stamp could do other times in the mean time. Also, I believe this card holds its output when no input signal is present, stamp could signal once and continue. That is something I could test next week if we have one in stock, and you’re interested.

    I still think it would be better to find a simple circuit that you could generate the signal with a digital pot or DAC. Why have controls driving controls.....

    BTW My company also is a Kele dealer and can get you a big cut off the list price.
  • OrionOrion Posts: 236
    edited 2005-07-03 16:41
    Well learning a few things about digital pots, I see that most don't seem to be able to have much more than 5v on there resistor outputs. This won't work with the circuit from above. Continued searching through the vast land of google I found this:
    http://www.chipdocs.com/pndecoder/number/694.html

    Check out the AD694 - http://www.analog.com/UploadedFiles/Data_Sheets/390142776AD694_b.pdf

    This could be driven with a 0-2v or 0-10v signal to drive 4-20ma current loop.

    Paul,

    My though on the impedance figure on page 24 is they are talking about the input of the actuator. i.e. the input is going to look like a <=300ohm resistor in the loop to the transmitter.

    Anyway I have a digikey order I'm working up, and I'll add a AD694 and give it a try.

    Post Edited (Orion) : 7/3/2005 5:22:38 PM GMT
  • bhaanbhaan Posts: 37
    edited 2005-07-05 13:35
    Orion, what kind of price range is that output transducer in?· Do you see any problems with using it?· Is the output impedance in range?· I couldn't·find that in the specs.· Again, I really appreciate your help in solving this problem, I don't think I could get too far without all·the·advice and help.
  • OrionOrion Posts: 236
    edited 2005-07-06 02:55
    what kind of price range is that output transducer in?

    $10.00 on digikey

    Do you see any problems with using it?

    No, it may not work though.

    Is the output impedance in range?

    Output impedance is not even in the ballpark, though I searched all over for something that was and found nothing yet.

    In eight years of commercial controls work and I have never had much of a problem with current loops and compatibility. Two main things to know: what sources/sinks the loop power and to min/max voltage level of the devices. I think the datasheet of your actuator is talking about the input impedance and not the output. It wouldn't be the first time I was wrong though. I ordered a few of the AD694s and will try them on some devices we have at our shop. They should be in my hands this weekend unless the holiday screws things up. If I get it working I'll post you a schematic, and you can give it a try. The only thing I have seen toast sensors and controllers is two sourcing devices on the same loop. I wouldn't worry about smoking something. Worst case you may have a ten dollar IC that takes days to drive your actuator from open to close. At that point it may be easier to change out the input card on the actuator for one that will take a voltage input.
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-06 06:31
    Thanks Orion, you jogged my memory, current loop was the trigger. You are right about the output buffer improving performance. I think bhaan was asking about the price of the boards. The chip would work, here is an article for a circuit which accepts a pwm signal to drive the current loop and has an open line feedback.·
  • bhaanbhaan Posts: 37
    edited 2005-07-06 18:10
    Would this work?· I found this when searching google.
    http://www.advantech.com/products/Model_Detail.asp?model_id=1-D6FJV&bu=#

    For now, I think I·am going to try to get some components to test the circuit Paul posted, although I can't seem to find anywhere to buy the TL032 ampliphiers shown.· Digikey doesn't stock that item.· Orion, let me know if your test with the·AD694 works.
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-06 20:36
    Yes that will work.

    I found digkey has the TL032 in 8-DIP and 8-SOIC.
  • OrionOrion Posts: 236
    edited 2005-07-09 20:04
    Well I got the AD694 and tested it out on some·speed drives·and an actuator, works great.· Attached is a schematic of how I had it wired.· I used a PCF8591 to drive the 0-2v input, but just about anything will work.· Also below the circuit is a diagram of how I would try it with the actuator you have.
    3060 x 2310 - 540K
  • bhaanbhaan Posts: 37
    edited 2005-07-13 00:00
    Orion, what kind of command did the BS2 send to the PCF8591(PULSOUT?)?· I have never used a DAC before.· I have my BS2 doing other things so I would like to be able to just send a setting once and·do other things until it has to modify the actuator position.· ··Thanks for testing the AD694, much appreciated.

    I finally wired up Paul's circuit and it works great, but I'd like to not have to use the PWM command so that I don't have to loop it.
  • OrionOrion Posts: 236
    edited 2005-07-13 01:59
    I thought pwm would be a pain for what you’re trying to do. A Dac or digipot will do what you want. There is a ton of info on the Parallax site for using a handful of the dacs out there. The 8591 has 4 analog inputs and one analog out via an I2C interface, may be over kill for just driving that actuator. I just used it because I have it in a current project. Heres a great write up from the master JW on the 8591 http://www.parallax.com/dl/docs/cols/nv/vol3/col/nv79.pdf. That code is for the bs2p and higher that support the I2C commands. There was examples in the forums a while back with bs2 code, sorry I'm currently out of town and forced to use dialup so you will have to do the searching. Basically any dac or digipot will let you set them and they will hold there outputs, enabling the stamp to move on and do other things. When I get back to my home base I can post the code if you still need, but the NV link should be all you need to get going.
  • bhaanbhaan Posts: 37
    edited 2005-07-13 13:22
    Thanks for the link Orion, you answered my question. ·I was just curious in general how you had·communicated with the DAC.· I didn't really need the code(not yet anyways [noparse];)[/noparse] ), but I just wanted to make sure I could set it and move on to doing other things before I bought the components.·
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-13 13:41
    It depends on which DAC you choose; parallel, I2C and SPI all are possible interfaces to choose from.
  • bhaanbhaan Posts: 37
    edited 2005-07-13 14:38
    Again, my inexperience in electronics has come up with another question.· For the circuit Paul posted a while back (circuit), it recommended sending pulses in order to vary the voltage, and I can get this to work using the PWM command.· But, would I be able to use a digital potentiometer to give the circuit a·varied voltage rather than use the PWM command?· I don't think the output impedance would change for the overall circuit·since the current is coming from the transistor, right???



    Post Edited (bhaan) : 7/13/2005 2:45:00 PM GMT
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-13 14:55
    The article from·EDN requires a PWM input, which as Orion said, can cause problems when using a Stamp, since you couldn't do anything else while generating the PWM.

    The circuit Orion posted using the AD694 has a high impeadance input for the 0-2V input signal. This means the output impedance of the voltage source isn't a big issue, so you can use a digipot to set the input voltage.

    After re-analyzing the EDN article, I am fairly certain that removing the 4.7 µF capacitor from the input of the circuit will adapt the circuit for use with an analog input for the control circuit, enabling use of a digipot. Actually come to think of it, you can use an analog signal for the input of the circuit as it exists. This is because the 4.7 µF capacitor integrates the PWM signal into an analog voltage. If you supply a constant voltage in (not rapidly changing) the capacitor has no effect on the input voltage (capacitors "resist" changes in voltages across its terminals by storing or releasing electrons from it's plates, if the voltage across the terminals is not changing, the charge on the capacitor stays the same and therefore does not affect the voltage on it's positive terminal), it is just fed straight to the op-amp. The only effect of the 4.7 µF capacitor when using an analog input is when the input voltage is changed the capacitor causes the input voltage to change more slowly. This will cause the output current to also change more slowly. This may be a desired quality for you, it depends on how the device your controlling reacts to a rapidly changing input current. Making the capacitor value smaller will cause the output to change more quickly, increasing it will cause the output to change more slowly.

    Post Edited (Paul Baker) : 7/13/2005 3:16:12 PM GMT
  • bhaanbhaan Posts: 37
    edited 2005-07-13 19:35
    The output impedance of the circuit in the·EDN article is of no concern correct?· Since my actuator requires an output impedance of less than 300 Ohms, I just want to make sure before I hook it up to the actuator.· The circuit seems to be working great with a potentiometer, I think I'll leave the capacitor in there because it does give it a nice smooth output change.
  • Paul BakerPaul Baker Posts: 6,351
    edited 2005-07-13 20:46
    The output impedance you are concerned with is the resistance looking into Q3 from the view-point of RL, this is a low enough for your application (don't ask me to calculate what it is, thats fairly difficult for this particular circuit). The reason a low impedance output is required is in the world of IC circuits 20 mA is a fairly large load, in contrast the current going into the + terminal of the left most op-amp is in the µA or even nA region, this is roughly a 1000 to 1,000,000 times less current than the resistive load is consuming, and as such you don't have to worry about the impedance of the voltage you are supplying to the circuit.
  • OrionOrion Posts: 236
    edited 2005-07-14 01:40
    Looks like your well on your way, just a word of caution.
    ·
    That actuator is pretty big so I’m assuming that it’s a gas valve on a large furnace.· Natural gas can be a very impressive force if misused.· Is this valve part of a piece of a package unit?· If so you best understand how all the safeties and wiring they have in place before you take control of the valve.· Boilers can crack if put to a high rate of fire and the water is too cold,· furnace heat exchangers can crack/catch fire·releasing carbon monoxide into the air.· Just be careful and understand how the equipment operates before you change its control.· If any safeties trip on that unit, your valve should fail close, even if there is a 2-position valve before it.· Gas valves can stick.
    ·
  • bhaanbhaan Posts: 37
    edited 2005-07-14 13:25
    Yeah the actuator does control a large gas valve on a very large furnace.· The furnace has been professionally designed and approved.· The inside of the furnace is about 8'x8'x9', and is heated by 4 large burners.· The good thing is that there is a burner management system, along with safety shutoff solenoids on all gas lines, and manual shutoffs as well.· The burner management system has full control of everything, including the actuator,·and can overide any signals I send to the actuator.· The only thing the BS2 does is control the actuator to follow a temperature curve once the BMS releases the actuator to be controlled by the BS2.· I have written a VB.net program to load an excel file of the temperature curve, communicate with the BS2 and make proper calculations and adjustments to follow the curve, then log and display all temperature data and save it to an excel file.· The hardest part was/is the 4-20mA signal, but all the help I got·was much appreciated.


    Post Edited (bhaan) : 7/14/2005 1:34:09 PM GMT
Sign In or Register to comment.