Shop OBEX P1 Docs P2 Docs Learn Events
Voltage to current converters (4-20mA transmitter) — Parallax Forums

Voltage to current converters (4-20mA transmitter)

ArchiverArchiver Posts: 46,084
edited 2003-09-24 19:44 in General Discussion
I've tried the 4-20mA transmitter show in this link ->
http://www.elecdesign.com/Files/29/2985/Figure_02.gif but for some
reason I can't get down to 4mA, even with zero and span set at
absolute minimums. I'm thinking its because I'm using either a
PN2907 or MPSA64 transistors and they're more efficient than the
SK100. Don't know for sure because I simply can't find a datasheet
for the SK100 anywhere...

I can get down to 4mA by increasing P2 to a 1k pot but that results
in a lower maximum load resistance.

Any tips on other ways to get this puppy down to 4mA?

The other side of this coin is that I'd like to get the input signal
to be 0-5V instead of 0-1. By playing with the span I can get it to
about 0-2.5v for 4-20.

I'm also looking for a decent explanation of voltage to current
converters. I've figured out negative and positive voltage followers
but beyond that my book jumps into AC circuits. I've checked
out "Design formulas simplify classic V/I converter" in the EDN
design ideas, but the whole thing is wrong. One formula uses R4 in
it and there's no R4 in the bloomin schematic. It also says current
output is independant of RL, but its not. Using the example numbers,
I measure before RL and I get say 10mA, after RL lower than 10mA -
even with only a 50ohm resistor and 15v supply!

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2003-09-21 00:33
    >I've tried the 4-20mA transmitter show in this link ->
    >http://www.elecdesign.com/Files/29/2985/Figure_02.gif but for some
    >reason I can't get down to 4mA, even with zero and span set at
    >absolute minimums. I'm thinking its because I'm using either a
    >PN2907 or MPSA64 transistors and they're more efficient than the
    >SK100. Don't know for sure because I simply can't find a datasheet
    >for the SK100 anywhere...
    >
    >I can get down to 4mA by increasing P2 to a 1k pot but that results
    >in a lower maximum load resistance.
    >
    >Any tips on other ways to get this puppy down to 4mA?
    >
    >The other side of this coin is that I'd like to get the input signal
    >to be 0-5V instead of 0-1. By playing with the span I can get it to
    >about 0-2.5v for 4-20.
    >
    >I'm also looking for a decent explanation of voltage to current
    >converters. I've figured out negative and positive voltage followers
    >but beyond that my book jumps into AC circuits. I've checked
    >out "Design formulas simplify classic V/I converter" in the EDN
    >design ideas, but the whole thing is wrong. One formula uses R4 in
    >it and there's no R4 in the bloomin schematic. It also says current
    >output is independant of RL, but its not. Using the example numbers,
    >I measure before RL and I get say 10mA, after RL lower than 10mA -
    >even with only a 50ohm resistor and 15v supply!

    I believe the problem stems from the LM324 op-amp that is used in
    that circuit. The circuit design demands that both the inputs and
    the outputs swing up near the positive supply rails. For example,
    with 4ma current, the voltage across P2, the 100 ohm resistor, would
    be only 0.4 volt. So, if the power supply V+ is 12 volts, then the
    voltage at pin 6 of the op-amp is 11.6 volts. But according to the
    LM324 data sheet, the input common mode range only extends up to
    within 1.5 volts of the V+. No good circuit using the LM324 should
    really attempt to operate with that pin 6 voltage above 10.5 volts
    when V+=12 volts. You probably had better luck with the 1kohm
    resistor because with 500 ohms in the circuit, there would be a two
    volt drop, and that is well within the input common mode
    specification. (but that resistor also throws off the span
    adjustment, as you noted)

    Another problem the design has is the limit to how high the LM324 can
    swing at its output. In order to cut off the transistor fully, the
    base-to emitter voltage of the transistor has to be less than about
    0.6 volts. However, the spec sheet for the LM324 only claims that the
    op-amp can swing within about 1 volt below the V+. That 1 volt
    would leave the transistor turned on. Quite often op-amps do in fact
    perform better than their spec sheet data, but you can never count on
    it, and the performance is likely to be marginal at best.

    You might ask, why is such a circuit to be found in prestigious
    Electronic Design magazine? hehe! It is unlikely that the submitter
    ever actually built it, or else they would have run aground on those
    problems. And the editors of the magazine use a panel of experts who
    only look at those designs quickly, and I can see how they would be
    recognize it as a clever circuit. The thing about op-amp circuits is
    that there are many many practical details that have to be
    considered, and the input and output common mode range is one of
    those. The choice of op-amp has nothing directly to do with the
    cleverness of the circuit.

    One solution is to use what is called a rail to rail input and output
    op-amp. These days there are many of those available. For example,
    one I keep in my stock is the LM6482. With one of those the circuit
    should work as advertised.

    Be aware that the offset in that circuit depends on V+, so V+ has to
    be regulated.

    There are other solutions, keeping the LM324. One of these would be
    to disconnect both P2 (100 ohm pot) and P1 (20k pot) from V+, then
    connect the input of a 10 volt regulator to V+ and connect its 10
    volt output to the junction between the top of P1 and P2. And leave
    the op-amp powered from V+. The regulator establish a reference
    point below V+ that puts everything within the common mode voltage
    range of the LM324. The output compliance would then be about 9
    volts. And the output would not depend on the supply voltage so long
    as the regulator is putting out its 10 volts.

    I don't think the transistors you are using have much to do with the
    problems you are seeing. The most accurate results come when you use
    a high gain transistor, or a darlington, or best of all, a p-channel
    mosfet. You see, the op-amp circuit is monitoring the emitter
    current of the transistor, but what actually gets to the load is the
    collector current. The emitter current is the sum of the base
    current + the collector current, so base current counts as an error.
    The elecdesign article mentions that. There is zero base current
    error in a mosfet.

    -- Tracy
  • ArchiverArchiver Posts: 46,084
    edited 2003-09-21 03:35
    >
    Original Message
    > From: Tracy Allen [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=ULyqADAbLyQCwwEiQlkA7ZWt2yd7RUPY_7TxJFilsGWA2GJFvKP0w4SWjQMibOdsYQjPIlN0u5gMf0E2U8WX]tracy@e...[/url
    > Sent: September 20, 2003 4:33 PM
    > To: basicstamps@yahoogroups.com
    > Subject: Re: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    > transmitter)

    Ah, I was thinking the 324 could go to within 1mV of the rails. Looking over
    my other op-amps though, that was the OP2340! Unfortunately I have that in
    SOIC only (can't breadboard it). I also have some TLC2262 - true rail to
    rail... again, soic. I really should remember to order dip versions of every
    chip I buy to breadboard with.

    I just tried again using a pair of CA3160E (10mV from rails). Still no fly,
    can't get below 8.2mA. I never would have thought it would be so hard to
    find a 4-20mA transmitter circuit that actually works...
  • ArchiverArchiver Posts: 46,084
    edited 2003-09-21 04:23
    > >
    Original Message
    >> From: Tracy Allen [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=ITyUJf_Wy9lirhEWLK2Z_Z_rIzpYnzsQU6GOuoY3RN5pp-MNr0d8Vdr1XRUIldQyw3cbQ6mTgtM0Xzo]tracy@e...[/url
    >> Sent: September 20, 2003 4:33 PM
    >> To: basicstamps@yahoogroups.com
    >> Subject: Re: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    >> transmitter)
    >
    >Ah, I was thinking the 324 could go to within 1mV of the rails. Looking over
    >my other op-amps though, that was the OP2340! Unfortunately I have that in
    >SOIC only (can't breadboard it). I also have some TLC2262 - true rail to
    >rail... again, soic. I really should remember to order dip versions of every
    >chip I buy to breadboard with.
    >
    >I just tried again using a pair of CA3160E (10mV from rails). Still no fly,
    >can't get below 8.2mA. I never would have thought it would be so hard to
    >find a 4-20mA transmitter circuit that actually works...



    The CA3160 swings close to both rails at the output, but the input
    can't get anywhere near the (+) supply. For this circuit it has to
    be both. I'm not familiar with the OP2340. The LM324 can swing close
    to ground and the input common mode range includes ground, but not
    the (+) supply.

    I keep the LMC6482 & LMC6582 on hand for r-r input and output.

    My own voltage to current converters
    <http://www.emesystems.com/pdfs/VICO_ops.pdf> have to be very
    accurate, so I use a CAZ op amp with less than 5 microvolts of input
    offset. It is not r-r, so I use a scheme with a voltage regulator
    like I described to you in a previous email. They do work! I have
    another circuit based on an LM10, that does 4-20 milliamps for 0-1
    volt input, in loop powered applications where the sensor itself
    draws less than 3 milliamps.

    Get ahold of some soic-8 to dip-8 adapter circuit boards!

    -- Tracy
  • ArchiverArchiver Posts: 46,084
    edited 2003-09-21 16:34
    You Sir, are a font of knowledge [noparse]:D[/noparse]

    I didn't know about the input voltage limits of op amps. Thats one more
    mystery de-mystified! Thanks much for the LMC6482 lead. It looks like a good
    replacement for the CA3160 on my pH meters as well. 10+ Teraohm input
    impedance and a LOT more ESD tolerance. I've got a bunch on order at Digikey
    now (and a couple DIPS!).

    It turned out that I did have dip versions of the OPA2340 but they're low
    voltage - 5v max. I set up the circuit again using only 5V and ground just
    to see how it would react. I did indeed manage to get below 4mA but now I
    have just about no ability to set a current range. To get above zero, p1 has
    to be just about at zero volts out. I'll try again when the LMC6482's
    arrive.

    I'll see if I can find something to give me 10V regulated tomorrow. Only
    have 9 and 12 volt regulators at the moment. Should be an adjustable at the
    local store that I can use. Then I'll see how the LM324 works with that
    change. At the same time I can try the split supply circuit that was in the
    same article as this single supply one. It also requires 10V regulated.

    Actually, just to make sure I'm not getting this backwards. P1 (20k pot to +
    input) is the zero and P2 (100 ohm pot to the collector) is the span.
    Correct?

    Heh, after I get this working I break out the BS2 and BOE then go through
    the PID section of the industrial controls course and try to figure out how
    to combine the two [noparse]:D[/noparse]


    >
    Original Message
    > From: Tracy Allen [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=q663ZTT5ShukHfVszC-vF7GZ9MGM7hCmStH02R0vYiPJfEIQA-sQ6W_6YhAJRcUmHwLW-EmqM1-4EU4]tracy@e...[/url
    > Sent: September 20, 2003 8:23 PM
    > To: basicstamps@yahoogroups.com
    > Subject: RE: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    > transmitter)
    >
    >
    > > >
    Original Message
    > >> From: Tracy Allen [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=q663ZTT5ShukHfVszC-vF7GZ9MGM7hCmStH02R0vYiPJfEIQA-sQ6W_6YhAJRcUmHwLW-EmqM1-4EU4]tracy@e...[/url
    > >> Sent: September 20, 2003 4:33 PM
    > >> To: basicstamps@yahoogroups.com
    > >> Subject: Re: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    > >> transmitter)
    > >
    > >Ah, I was thinking the 324 could go to within 1mV of the rails.
    > Looking over
    > >my other op-amps though, that was the OP2340! Unfortunately I
    > have that in
    > >SOIC only (can't breadboard it). I also have some TLC2262 - true rail to
    > >rail... again, soic. I really should remember to order dip
    > versions of every
    > >chip I buy to breadboard with.
    > >
    > >I just tried again using a pair of CA3160E (10mV from rails).
    > Still no fly,
    > >can't get below 8.2mA. I never would have thought it would be so hard to
    > >find a 4-20mA transmitter circuit that actually works...
    >
    >
    >
    > The CA3160 swings close to both rails at the output, but the input
    > can't get anywhere near the (+) supply. For this circuit it has to
    > be both. I'm not familiar with the OP2340. The LM324 can swing close
    > to ground and the input common mode range includes ground, but not
    > the (+) supply.
    >
    > I keep the LMC6482 & LMC6582 on hand for r-r input and output.
    >
    > My own voltage to current converters
    > <http://www.emesystems.com/pdfs/VICO_ops.pdf> have to be very
    > accurate, so I use a CAZ op amp with less than 5 microvolts of input
    > offset. It is not r-r, so I use a scheme with a voltage regulator
    > like I described to you in a previous email. They do work! I have
    > another circuit based on an LM10, that does 4-20 milliamps for 0-1
    > volt input, in loop powered applications where the sensor itself
    > draws less than 3 milliamps.
    >
    > Get ahold of some soic-8 to dip-8 adapter circuit boards!
    >
    > -- Tracy
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the
    > Subject and Body of the message will be ignored.
    >
    >
    > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2003-09-21 20:19
    Hi Pat,

    The LMC6482 is one member of a family of r-r CMOS op amps from
    National Semi. Family members optimize different qualities, input
    accuracy, power supply, bandwidth etc. For example, the LMC6462
    operates with only 50 microamps supply current, but the
    gain-bandwidth is reduced to 50 kilohertz, compared with the LMC6462
    with 1.6 milliamps and 1 megahertz GBW. It is better than the old
    CA3160! All members of the family share the typical 0.02 picoamps
    input current to go with that >10 teraohm input impedance. The
    offset voltage is on the order of a millivolt, but the drift is
    pretty good at ~1 microvolt per degree C.

    For the circuit suggestion, the absolute value of the regulated
    voltage is unimportant. 9 volts could do for test purposes. It does
    reduce the compliance at the output. With a 9 volts regulator, the
    output could swing up to 6.5 volts or so. You have to allow 1.25
    volts across the span pot P2 plus at least one volt for the
    transistor from collector to emitter. (The base current error
    increases as the transistor approaches saturation.)

    In the circuit http://www.elecdesign.com/Files/29/2985/Figure_02.gif,
    yes, the 20k pot P1 is the zero, and 100 ohm pot P2 is the span.
    While that is true, the calibration is not as easy as they make it
    sound in the caption to the figure. The adjustments interact and the
    zero will be very sensitive.

    The first op-amp stage has a gain of -1 volt per volt, and 1 volt
    change needs to produce an output current change of 16 milliamps.
    (That is the calibration they state in the caption.) So the
    adjustment of P2 will be very close to 1 volt/0.016 amp = 62.5 ohms.
    To achieve a current of 4 milliamps, the voltage across the 62.5 ohms
    must be 0.25 volt, and that is, 0.25 volt less than the 9 volt
    regulator. So, 8.75 volts at the output of the first stage. The
    first stage gain is x2 for the voltage from the wiper of the zeroing
    pot, P2. So that voltage will need to be set at very very close to
    8.75/2 = 4.375 volts. To put it another way, when the input is
    shorted, that 4.375 volts will be amplified to 8.75 volts at the
    output of the first stage and that will be converted into the 4
    milliamp output current. You can see why it is sensitive to the 9
    volt reference supply. The 20k pot better be a 20 turn precision
    type, or better yet, a pot centered in a voltage divider made with
    fixed resistors. When the input is brought up to 1 volt full scale
    input, then there will be 1.25 volts across the 62.5 ohm span
    resistor, and the output current should be 1.25/62.5 = 20 milliamps.

    I see I mistyped a URL in the last email. It should be...
    <http://www.emesystems.com/pdfs/VICO-ops.pdf>

    -- Tracy


    >You Sir, are a font of knowledge [noparse]:D[/noparse]
    >
    >I didn't know about the input voltage limits of op amps. Thats one more
    >mystery de-mystified! Thanks much for the LMC6482 lead. It looks like a good
    >replacement for the CA3160 on my pH meters as well. 10+ Teraohm input
    >impedance and a LOT more ESD tolerance. I've got a bunch on order at Digikey
    >now (and a couple DIPS!).
    >
    >It turned out that I did have dip versions of the OPA2340 but they're low
    >voltage - 5v max. I set up the circuit again using only 5V and ground just
    >to see how it would react. I did indeed manage to get below 4mA but now I
    >have just about no ability to set a current range. To get above zero, p1 has
    >to be just about at zero volts out. I'll try again when the LMC6482's
    >arrive.
    >
    >I'll see if I can find something to give me 10V regulated tomorrow. Only
    >have 9 and 12 volt regulators at the moment. Should be an adjustable at the
    >local store that I can use. Then I'll see how the LM324 works with that
    >change. At the same time I can try the split supply circuit that was in the
    >same article as this single supply one. It also requires 10V regulated.
    >
    >Actually, just to make sure I'm not getting this backwards. P1 (20k pot to +
    >input) is the zero and P2 (100 ohm pot to the collector) is the span.
    >Correct?
    >
    >Heh, after I get this working I break out the BS2 and BOE then go through
    >the PID section of the industrial controls course and try to figure out how
    >to combine the two [noparse]:D[/noparse]
    >
    >
    >>
    Original Message
    >> From: Tracy Allen [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=DjDzcdwtov8lXSBe2WmXwB4PNxToi2ZY8MSv1ok0cA83Yu5J1KplKWZBLOduFM1i3wVbMs2LPf7RHMJzou5-]tracy@e...[/url
    >> Sent: September 20, 2003 8:23 PM
    >> To: basicstamps@yahoogroups.com
    >> Subject: RE: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    > > transmitter)
    > >
    > >
    > > > >
    Original Message
    > > >> From: Tracy Allen [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=DjDzcdwtov8lXSBe2WmXwB4PNxToi2ZY8MSv1ok0cA83Yu5J1KplKWZBLOduFM1i3wVbMs2LPf7RHMJzou5-]tracy@e...[/url
    > > >> Sent: September 20, 2003 4:33 PM
    > > >> To: basicstamps@yahoogroups.com
    > > >> Subject: Re: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    >> >> transmitter)
    >> >
    >> >Ah, I was thinking the 324 could go to within 1mV of the rails.
    >> Looking over
    >> >my other op-amps though, that was the OP2340! Unfortunately I
    >> have that in
    >> >SOIC only (can't breadboard it). I also have some TLC2262 - true rail to
    >> >rail... again, soic. I really should remember to order dip
    >> versions of every
    >> >chip I buy to breadboard with.
    >> >
    >> >I just tried again using a pair of CA3160E (10mV from rails).
    >> Still no fly,
    >> >can't get below 8.2mA. I never would have thought it would be so hard to
    >> >find a 4-20mA transmitter circuit that actually works...
    >>
    >>
    >>
    >> The CA3160 swings close to both rails at the output, but the input
    > > can't get anywhere near the (+) supply. For this circuit it has to
    > > be both. I'm not familiar with the OP2340. The LM324 can swing close
    > > to ground and the input common mode range includes ground, but not
    >> the (+) supply.
    >>
    >> I keep the LMC6482 & LMC6582 on hand for r-r input and output.
    >>
    >> My own voltage to current converters
    > > <http://www.emesystems.com/pdfs/VICO_ops.pdf> have to be very
    >> accurate, so I use a CAZ op amp with less than 5 microvolts of input
    >> offset. It is not r-r, so I use a scheme with a voltage regulator
    >> like I described to you in a previous email. They do work! I have
    >> another circuit based on an LM10, that does 4-20 milliamps for 0-1
    >> volt input, in loop powered applications where the sensor itself
    >> draws less than 3 milliamps.
    >>
    >> Get ahold of some soic-8 to dip-8 adapter circuit boards!
    >>
    >> -- Tracy
    >>
    >> To UNSUBSCRIBE, just send mail to:
    >> basicstamps-unsubscribe@yahoogroups.com
    >> from the same email address that you subscribed. Text in the
    >> Subject and Body of the message will be ignored.
    >>
    >>
    >> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >>
    >>
    >>
    >
    >
    >
    >To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    >from the same email address that you subscribed. Text in the
    >Subject and Body of the message will be ignored.
    >
    >
    >Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
  • ArchiverArchiver Posts: 46,084
    edited 2003-09-24 07:26
    Received the LMC6482s today so I set up the circuit again. I set P2 to 62.5
    ohms and the output of the first amplifier stage to Vcc - 0.25V. Slapped on
    the ammeter and got a reading of 3.87mA! A bit of twiddling with the two
    pots and voila, 4-20mA current driver! Just to be sure I measured the
    current shorted to ground and then through a 470 ohm resistor. Exactly the
    same readings both times!

    Thank you for the help Tracy Allen, you're a great boon to this group!

    Tomorrow I start on the PID section of the industrial controls pdf on the
    parallax site. I've read through it already and understand about 98% of what
    I read. Have to see it in action to be sure though.
  • ArchiverArchiver Posts: 46,084
    edited 2003-09-24 16:07
    Actually, one more thing on this current driver. I wanted to change the
    input range to 0-5 volts (Stamp PWM output). 0-1 would give me 51 steps of
    PWM output which isn't very good for an actuator.

    I read read my opamp chapter (for the 256th time [noparse]:D[/noparse] ) and decided I can't
    really do anything to the circuit itself to change the input range from
    there. So the simple solution comes to mind, divide the PWM output to 1/5th.
    Do the RC output as per normal but add a resistor (multi-turn pot) to ground
    between the pin and the capacitor.

    Is there a better way to do this?


    The simplest answer would be a voltage divider (100k and about 12K seems to
    do the trick).

    >
    Original Message
    > From: PatM [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=0KIe-w_C1U0jv6lgsfCfjzdMJTHeTlLg8HyTmt5VjzCl6IR0qqrliOHxMfawbvhYr0bfL9zW]pmeloy@s...[/url
    > Sent: September 23, 2003 11:26 PM
    > To: basicstamps@yahoogroups.com
    > Subject: RE: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    > transmitter)
    >
    >
    > Received the LMC6482s today so I set up the circuit again. I set
    > P2 to 62.5
    > ohms and the output of the first amplifier stage to Vcc - 0.25V.
    > Slapped on
    > the ammeter and got a reading of 3.87mA! A bit of twiddling with the two
    > pots and voila, 4-20mA current driver! Just to be sure I measured the
    > current shorted to ground and then through a 470 ohm resistor. Exactly the
    > same readings both times!
    >
    > Thank you for the help Tracy Allen, you're a great boon to this group!
    >
    > Tomorrow I start on the PID section of the industrial controls pdf on the
    > parallax site. I've read through it already and understand about
    > 98% of what
    > I read. Have to see it in action to be sure though.
    >
    >
    >
    > To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    > from the same email address that you subscribed. Text in the
    > Subject and Body of the message will be ignored.
    >
    >
    > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2003-09-24 19:44
    The 4-20ma circuit has an input impedance of 10 kohms, so the PWM has
    to be bufffered first. The thing about the Stamp PWM to analog is it
    needs to look into a high impedance. Here is an additional op-amp as
    a buffer:

    1/2 LMC6482
    100k |\ ;----buffer to 4-20 circuit
    Stamp P0 ---/\/\--o
    |+\ |
    PWM | | >--o-/\/\-o-/\/\-;
    | ;-|-/ | 40.2k 10k |
    0.1uf ===== | |/ | R1 R2 |
    | | | Vss
    | `
    '
    Vss


    R1 and R2 can be chosen with a ratio of R2/(R1+R2)=1/5, then into
    another buffer (another 1/2 your dual LMC6482 op-amp), and then into
    the 4-20 circuit, adjusted for 0->1 volt input as it is now.

    Or, here is an alternative:

    1/2 LMC6482
    100k |\
    Stamp P0 ---/\/\--o
    |+\
    PWM | | >--o-/\/\--- 4-20 circuit (-) input
    | ;-|-/ | 49.9k
    0.1uf ===== | |/ |
    | | |
    | `
    '
    Vss

    Use 49.9 kohms, right into the (-) input of the 4-20 circuit.
    http://www.elecdesign.com/Files/29/2985/Figure_02.gif (Replacing the
    10kohm resistor that is there now). That makes the inverting gain of
    that circuit -1/5 instead of -1, so the scaling becomes 0 to 5 volts,
    as you want, instead of the original 0->1 volt.

    That also changes the zero point calculation, because with an
    inverting gain of -1/5, the non-inverting gain becomes 1.2 instead of
    2.0. You still need 8.75 volts at the input to the second stage, to
    produce the 4 ma offset. So now the wiper of the "zero" pot P1 has
    to be set at about 8.75/1.2 = 7.29 volts.

    -- Tracy Allen
    http://www.emesystems.com




    >Actually, one more thing on this current driver. I wanted to change the
    >input range to 0-5 volts (Stamp PWM output). 0-1 would give me 51 steps of
    >PWM output which isn't very good for an actuator.
    >
    >I read read my opamp chapter (for the 256th time [noparse]:D[/noparse] ) and decided I can't
    >really do anything to the circuit itself to change the input range from
    >there. So the simple solution comes to mind, divide the PWM output to 1/5th.
    >Do the RC output as per normal but add a resistor (multi-turn pot) to ground
    >between the pin and the capacitor.
    >
    >Is there a better way to do this?
    >
    >
    >The simplest answer would be a voltage divider (100k and about 12K seems to
    >do the trick).
    >
    >>
    Original Message
    >> From: PatM [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=bkImIHVT0w2JR2DNJ5l_Dtqdv04jFxKIj1jtphwPBAKOEO91n0x0nkxzL98-QPMZNYgnzE0]pmeloy@s...[/url
    >> Sent: September 23, 2003 11:26 PM
    >> To: basicstamps@yahoogroups.com
    >> Subject: RE: [noparse][[/noparse]basicstamps] Voltage to current converters (4-20mA
    >> transmitter)
    >>
    >>
    >> Received the LMC6482s today so I set up the circuit again. I set
    >> P2 to 62.5
    >> ohms and the output of the first amplifier stage to Vcc - 0.25V.
    >> Slapped on
    >> the ammeter and got a reading of 3.87mA! A bit of twiddling with the two
    >> pots and voila, 4-20mA current driver! Just to be sure I measured the
    >> current shorted to ground and then through a 470 ohm resistor. Exactly the
    >> same readings both times!
    >>
    >> Thank you for the help Tracy Allen, you're a great boon to this group!
    >>
    >> Tomorrow I start on the PID section of the industrial controls pdf on the
    >> parallax site. I've read through it already and understand about
    >> 98% of what
    >> I read. Have to see it in action to be sure though.
    >>
    >>
    >>
    >> To UNSUBSCRIBE, just send mail to:
    >> basicstamps-unsubscribe@yahoogroups.com
    >> from the same email address that you subscribed. Text in the
    >> Subject and Body of the message will be ignored.
    >>
    >>
    >> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
    >>
    >>
    >>
    >
    >
    >
    >To UNSUBSCRIBE, just send mail to:
    > basicstamps-unsubscribe@yahoogroups.com
    >from the same email address that you subscribed. Text in the
    >Subject and Body of the message will be ignored.
    >
    >
    >Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
Sign In or Register to comment.