Loading ...
Sorry, an error occurred while loading the content.

Re: My first head design: input please

Expand Messages
  • Robert Schwartz
    Hmmmm I m getting nothing out of it, no motor, no LEDs. I don t know if it is something wrong with the circuit, or my soldering. If somebody else can test
    Message 1 of 23 , Jun 2, 2005
    • 0 Attachment
      Hmmmm
      I'm getting nothing out of it, no motor, no LEDs. I don't know if it
      is something wrong with the circuit, or my soldering. If somebody
      else can test the circuit out that would be great. Meanwhile, based
      on Martin McKee's suggestions, I took away the voltage dividing
      resistors reducing the part count down to 4 (LM324, Pot, 2 Cds).
    • Wilf Rigter
      Right parts, wrong connections. In fact, be careful as you can short out the battery with the pot (smoke!) I have placed the corrected version in your folder.
      Message 2 of 23 , Jun 2, 2005
      • 0 Attachment
        Right parts, wrong connections. In fact, be careful as you can short out the battery with the pot (smoke!)  I have placed the corrected version in your folder. Your original version 1.0 was fine too. I added the LED to ensure that the maximum input voltage is ~1.5V less than the opamp positive supply. That is necesary because the LM324 has a limited input range. Also added the power supply pin numbers for reference.  Note the much smaller file size using the GIF format. 
         
        regards
         
        wilf
         
         
        ----- Original Message -----
        Sent: Thursday, June 02, 2005 2:24 PM
        Subject: [beam] Re: My first head design: input please

        Hmmmm
        I'm getting nothing out of it, no motor, no LEDs.  I don't know if it
        is    something wrong with the circuit, or my soldering.  If somebody
        else can test the circuit out that would be great.  Meanwhile, based
        on Martin McKee's suggestions, I took away the voltage dividing
        resistors reducing the part count down to 4 (LM324, Pot, 2 Cds).


      • Robert Schwartz
        Hey guys, I revised my design a bit more now with transistor drivers. I designed the bridge myself, but then, after I had designed it, recognized it as a
        Message 3 of 23 , Jun 7, 2005
        • 0 Attachment
          Hey guys,
          I revised my design a bit more now with transistor drivers. I
          designed the bridge myself, but then, after I had designed it,
          recognized it as a standard 4 transistor bridge :). This can be more
          easily turned into a 2DOF head because each DOF only requires 2
          Op-Amps, and the LM324 has 4. Total Part Count: 12 + 11 for an
          additional DOF. The file is in my folder.

          Robert Schwartz
        • Robert Schwartz
          Would using a comparator have these problems? I ve been using an op-amp basicly as a comparator with a higher output capability. If I used the lm339, would I
          Message 4 of 23 , Jun 8, 2005
          • 0 Attachment
            Would using a comparator have these problems? I've been using an
            op-amp basicly as a comparator with a higher output capability. If I
            used the lm339, would I be able to use one resistor on the output, and
            then one for each transitors' base on the right? I want to keep the
            part count to a minumum.
          • Martin Jay McKee
            Rail to rail outputs cure all ill s... well not really, but the problems that have been mentioned are a result of the output capabilities of the op amp, not
            Message 5 of 23 , Jun 8, 2005
            • 0 Attachment
              Rail to rail outputs cure all ill's... well not really, but the problems
              that have been mentioned are a result of the output capabilities of the op
              amp, not the fact that it "is' an op-amp. Remember that with out a feedback
              network an op-amp basically is a comparator. If you replaced the LM324 with
              an op-amp (or comparator, look at TI, Fairchild Semiconductor, etc.) that
              has rail to rail outputs some things would definitely be easier, but you
              still will have to use proper base resistors to avoid frying things.

              Martin Jay McKee
            • roboanalogtom
              I hate to say this but....you don t need resistors between the op amps and the transistors. I m not sure how to upload files but if someone will be kind enough
              Message 6 of 23 , Jun 8, 2005
              • 0 Attachment
                I hate to say this but....you don't need resistors between the op
                amps and the transistors. I'm not sure how to upload files but if
                someone will be kind enough to let me know I'll upload a schematic
                that I use for simple servos that I've cycled 150,000 times under
                load with no problems and without current limiting resistors. I've
                got a CD with 19 videos that I mailed to Wolfgang, one of the videos
                shows a 2 DOF head that tracks and smacks around a styrofoam ball
                using the identical circuit in the schematic (if Wolfgang puts the
                video up, I'll describe how the head works. It's pretty neat.)

                Why don't things fry? The op amps have current limiting resistors
                built in. When I tried using digital chips as comparators things
                started getting a little hot.

                Under heavy load the output of the op amp will reach ground voltage.
                Try setting an op amp to a low output and connect it to a variable
                resistor to ground.

                I've done this with the lm358, lm1458, TL082, and the 5532. To my
                surprise the lowly LM358 gave the best performance.

                Another suprise is when I measured the voltages in the op amp
                outputs on my digital scope, the high voltage was just above .7
                volts and the low was 0 volts, due to saturation. I would agree with
                everyone that things should fry and if I hadn't built and tested the
                circuit I'd have a hard time believing it but after actually
                building and testing the circuit....

                Do I just email the circuit to the group or is there another way of
                uploading files?
              • Wilf Rigter
                Hi Tom, you can create a personal folder in the group file section and upload your schematic files and, if selected, a group announcement can be automatically
                Message 7 of 23 , Jun 8, 2005
                • 0 Attachment
                  Hi Tom, you can create a personal folder in the group file section and upload your schematic files and, if selected, a group announcement can be automatically generated with a link to the new file.    Because of limited space, check that the drawing uses the most compact format possible e.g. 2 color, GIF.
                   
                  Thanks for the motor driver comments Tom. Your input is highly appreciated.

                  Before I reply, I want to just mention that this message is directed at the group as a whole to encourage participation on many levels.
                   
                  Please note that I often comment on a design and suggest changes while keeping the circuit as close to the original as possible. In addition, I may suggest a "if I were in your shoes" variation.  That does not necesarily mean that I would have designed such a circuit in the first place but rather show how ideas can be used as an inspiration for other ideas.
                   
                  I hope that in helping beamers perfect their own designs, this process of proposal, discussion, comments and suggestions will be a positive learning experience. Exploring new ideas and designing your own circuits/mechanics/robots is a big part of what this list is about. 
                   
                  Now specifically to reply to your comment.
                   
                  You are probably jumping ahead of me when you say that "no resistors are needed between the opamps and the transistors" but are we are talking about the original single opamp hybrid bridge design (V1.3)?
                   
                  If so then both Jo Charles and I spotted a relatively innocent looking problem that will most certainly fry two of the transistors. Hence our respective suggestions that base resistors are needed between the opamps and the transistors.
                   
                  In design V1.3, the problem is on the right side of the bridge which consists of two complementary transistors used in the common emitter configuration. These NPN and PNP transistors have their emitters connected to the 0V and +V rails respectively. They also have their bases connected together. This creates a short between the power rails through the two series connected forward biased base emitter junctions. Unless the power supply is current limited, the current will destroy those two transistors.
                   
                  In V1.4, I suggest incremental changes to address the short circuit problem with the original design. Simply adding base resistors to the right hand transistors is not quite enough as there is the LM324 problem of insufficient positive swing which means that the right PNP transistor would never turn off. Adding two LEDs in series with the transistor bases effectively increases the turn on voltage of these two transistors to about 2V to be "more" compatible with the opamp positive output swing.
                  The LEDs also serve as motor rotation indication.
                   
                  I say "more" compatible because the voltage drop due to heavy loading of the opamp output by the emitter follower side of the bridge is another potential problem which can cause both transistors of the common emitter side of the bridge to turn on at the same time.  Keeping the single base resistor on the emitter follower side of the bridge  reduces the voltage drop on the opamp output to avoid that potential problem.     
                   
                  In V1.5, the second inverting opamp is added and I suggested adding negative feedback in V1.5 to limit the voltage gain so that with the light (perfectly) balanced there will be 0V across the motor winding to reduce power. In V1.5, both sides of bridge are changed to a simple current boosting emitter follower design. In that case no base resistors are required at all.
                   
                  This simpler design comes at a price: The maximum voltage across the motor load will equal to the opamp output swing plus the base emitter drops of the transistors or about Vsupply-3V. Not great performance if you use a typical solar supply of less than 6v.
                   
                  The ideal BEAM motor driver should have a saturated output (<200mV drop) and operate down to a supply of 3V or less. Complementary very low gate threshold powerFETs are a good choice but are not easy to find. The power smart bridge comes close but is relatively complicated.    
                   
                  That's all the time I have right now. I hope more later!
                   
                  Tom,  sorry for the painstaking detail and going over old ground but I want to make sure we are talking about the same design.
                   
                  wilf
                   
                  ----- Original Message -----
                  Sent: Wednesday, June 08, 2005 3:07 PM
                  Subject: [beam] Re: My first head design: input please

                  I hate to say this but....you don't need resistors between the op
                  amps and the transistors. I'm not sure how to upload files but if
                  someone will be kind enough to let me know I'll upload a schematic
                  that I use for simple servos that I've cycled 150,000 times under
                  load with no problems and without current limiting resistors. I've
                  got a CD with 19 videos that I mailed to Wolfgang, one of the videos
                  shows a 2 DOF head that tracks and smacks around a styrofoam ball
                  using the identical circuit in the schematic (if Wolfgang puts the
                  video up, I'll describe how the head works. It's pretty neat.)

                  Why don't things fry? The op amps have current limiting resistors
                  built in. When I tried using digital chips as comparators things
                  started getting a little hot.

                  Under heavy load the output of the op amp will reach ground voltage.
                  Try setting an op amp to a low output and connect it to a variable
                  resistor to ground.

                  I've done this with the lm358, lm1458, TL082, and the 5532. To my
                  surprise the lowly LM358 gave the best performance.

                  Another suprise is when I measured the voltages in the op amp
                  outputs on my digital scope, the high voltage was just above .7
                  volts and the low was 0 volts, due to saturation. I would agree with
                  everyone that things should fry and if I hadn't built and tested the
                  circuit I'd have a hard time believing it but after actually
                  building and testing the circuit....

                  Do I just email the circuit to the group or is there another way of
                  uploading files?


                • Joseph Charles
                  I breadboarded your circuit Robert, using 1k base resistors on the right side of the bridge, with both an LM358 op-amp and an LM393 comparator with a 2k2 pull
                  Message 8 of 23 , Jun 9, 2005
                  • 0 Attachment
                    I breadboarded your circuit Robert, using 1k base resistors on the right side of the bridge, with both an LM358 op-amp and an LM393 comparator with a 2k2 pull up resistor on the output, but got much the same result : the high output went only to about 3.5V (using a 5V supply) and, as Wilf said, the right hand PNP would not turn off.  I certainly got the motor to turn both ways, one way (with a high from the output) was a lot slower than the other (a low from the output).
                     
                    I didn't have much time, so couldn't experiment (a pity, 'cos it's my favourite passtime!) say with comparator pull-up resistors and the like.
                     
                    What with all the mucking around with resistors and such, you're probably better off with two op-amps driving an emitter follower bridge (Wilf's v1.5) if you have a high enough Vcc. With 5V supply you'd get around 3.5V across the motor.
                     
                    Apart from the light sensing components, this setup gives you a 1DOF head with one chip (LM358 or half an LM324) and 4 transistors, or a 2DOF head with one chip (LM324) and 8 transistors. Not too bad.
                     
                    Jo


                    Robert Schwartz <bob0513@...> wrote:
                    Would using a comparator have these problems?  I've been using an
                    op-amp basicly as a comparator with a higher output capability.  If I
                    used the lm339, would I be able to use one resistor on the output, and
                    then one for each transitors' base on the right?  I want to keep the
                    part count to a minumum.



                    Send instant messages to your online friends http://au.messenger.yahoo.com

                  • Robert Schwartz
                    Swapping out the transistor bridge with a 74AC240 should solve the problems I think. Since it switches at half Vcc, 3.5V is sufficient to switch, and logic
                    Message 9 of 23 , Jun 9, 2005
                    • 0 Attachment
                      Swapping out the transistor bridge with a 74AC240 should solve the
                      problems I think. Since it switches at half Vcc, 3.5V is sufficient
                      to switch, and logic low should also be low enough (I can always add a
                      diode to make a lower voltage anyways). This should also be more
                      efficient and simplify construction. One half of the bridge needs to
                      be inverting, the other side needs to be noninverting. My newest
                      design also leaves 2 open inverters and the enable lines which may be
                      useful for locking on.
                    • roboanalogtom
                      Hey Wilf, Thanks for showing me how to upload files. Please delete the first file I uploaded, when I opened it I discovered a mistake in the schematic. Sorry
                      Message 10 of 23 , Jun 9, 2005
                      • 0 Attachment
                        Hey Wilf,

                        Thanks for showing me how to upload files. Please delete the first
                        file I uploaded, when I opened it I discovered a mistake in the
                        schematic.

                        Sorry for coming across overbearing, I get excited and no disrespect
                        was intended.

                        I was pointing out in my post that one can rely on the internal
                        current limiting resistors in small signal op amps and that they're
                        rated for continuous duty. They're designed for this sort of abuse,
                        when I use op amps as led drivers I don't bother with resistors,
                        most op amps give out 12ma or so at a 5 volt power supply. You need
                        about 8 volts to get the full rating of 20ma. I have yet to blow out
                        an op amp (I did one time, but that was with a 3000 volt 6 uF
                        capacitor and a triggered spark gap. It failed in a very nasty
                        explosion!). I wasn't adressing nor did I refer to the npn/pnp
                        bridge or any other part of the circuitry, just the fact that you
                        can use the internal resistors.


                        You stated the heavily loaded op amps will limit the low output. Can
                        you elaborate or give a circuit example. I want to make sure we're
                        meaning the same thing; I just played with a LM358 and find that a
                        heavy load lowers the op amp low voltage output instead of limiting
                        it. The voltage drop starts to occur at 720 ohms (to ground) and the
                        more I load it the lower the voltage gets. I find that I can dial in
                        the low voltage output with the proper impedence. The floating
                        output is 1.84 volts at 5 volts supply for this particular chip.

                        -Tom
                      • J Wolfgang Goerlich
                        I am keeping an eye out for the OISM videos and will upload them. Am I correct in assuming that the head uses gear motors as opposed to servo motors? I hope
                        Message 11 of 23 , Jun 9, 2005
                        • 0 Attachment
                          I am keeping an eye out for the OISM videos and will upload them. Am I
                          correct in assuming that the head uses gear motors as opposed to servo
                          motors?

                          I hope so. I would like to make a Turbot using these ideas. With lots
                          of gear motors and next to no free servos, such a circuit would be
                          great.

                          J Wolfgang Goerlich

                          Tom Jenner wrote:
                          > I've got a CD with 19 videos that I mailed to Wolfgang, one of the
                          > videos shows a 2 DOF head that tracks and smacks around a styrofoam
                          > ball using the identical circuit in the schematic (if Wolfgang puts
                          > the video up, I'll describe how the head works. It's pretty neat.)
                        • roboanalogtom
                          Yup, I used DC gearhead motors that you can get at Solarbotics or at Pololu. You can buy directly from the manafactuer (particularly if you don t live in North
                          Message 12 of 23 , Jun 9, 2005
                          • 0 Attachment
                            Yup,
                            I used DC gearhead motors that you can get at Solarbotics or at Pololu.
                            You can buy directly from the manafactuer (particularly if you don't
                            live in North America) at www.wse.com.hk. Check this site out, they
                            have cheap DC gearhead motors. The motors I use cost US$1.04 in units
                            of 10,000 and I think US$2-2.50 or so in units of 10. In North America
                            I recommend Solarbotics, they support us, we should support them. You
                            can clearly see the motors in action. It tracks and hits a styrofoam
                            ball on a string. After people see the video, I'll post a high
                            resolution photo of the motor so people can see how I mounted the
                            potentiometer and circuit board to make a super cheap servo. Warm up
                            your hot glue gun!

                            People mentioned that they wanted to see an OISM walk so there are
                            several videos of 2 servo light follower walkers. There's also an 8
                            servo walker turning in its own radius, 8 servo walker in a chaotic
                            dance, 8 servo on a stand with the camera panning around so you can get
                            a good look at phase coupling and different phase spaces, 8 servo on a
                            stand with the legs being amplitude modulated, a different shot of that
                            trippy 2 servo light follower with touch sensor, 3 servo- 10
                            transistor "sea snake" that uses a single transistor CPG(!), robotic
                            hopping beer cans (with a lot of younger viewers you may consider not
                            posting this one, but it sure is funny!), a good close up of a slow
                            lm567 based crawler that has low pass filters, a couple videos with my
                            newer surface mount developement boards that demonstrates plug and play
                            robotics, an oscillating leg for a biped (good sine waves), and a few
                            more.

                            They went off in the mail on Wednesday so Friday or Saturday you should
                            have them.

                            -Tom


                            --- In beam@yahoogroups.com, "J Wolfgang Goerlich" <jwgoerlich@g...>
                            wrote:
                            > I am keeping an eye out for the OISM videos and will upload them. Am
                            I
                            > correct in assuming that the head uses gear motors as opposed to
                            servo
                            > motors?
                            >
                            > I hope so. I would like to make a Turbot using these ideas. With lots
                            > of gear motors and next to no free servos, such a circuit would be
                            > great.
                            >
                            > J Wolfgang Goerlich
                            >
                            > Tom Jenner wrote:
                            > > I've got a CD with 19 videos that I mailed to Wolfgang, one of the
                            > > videos shows a 2 DOF head that tracks and smacks around a styrofoam
                            > > ball using the identical circuit in the schematic (if Wolfgang puts
                            > > the video up, I'll describe how the head works. It's pretty neat.)
                          • Wilf Rigter
                            If you use a LM324 to drive 74AC inputs (motor driver) then use a 10K pull up resistor to increase the positive swing. Or else use a LM393 or LM339 comparator
                            Message 13 of 23 , Jun 9, 2005
                            • 0 Attachment
                              If you use a LM324 to drive  74AC inputs (motor driver) then use a 10K pull up resistor to increase the positive swing.
                               
                              Or else use a LM393 or LM339 comparator with pull up resistor which will give you a rail to rail output swing when driving a 74AC input. 
                               
                              You will find that the 74AC chip will run somewhat cooler and will have slightly lower forward voltage drop which is worth the extra resistor. 
                               
                              wilf
                               
                               
                               
                              ----- Original Message -----
                              Sent: Thursday, June 09, 2005 11:47 AM
                              Subject: [beam] Re: My first head design: input please

                              Swapping out the transistor bridge with a 74AC240 should solve the
                              problems I think.  Since it switches at half Vcc, 3.5V is sufficient
                              to switch, and logic low should also be low enough (I can always add a
                              diode to make a lower voltage anyways).  This should also be more
                              efficient and simplify construction.  One half of the bridge needs to
                              be inverting, the other side needs to be noninverting.  My newest
                              design also leaves 2 open inverters and the enable lines which may be
                              useful for locking on.


                            • Wilf Rigter
                              Hi Tom, Yes, I often get excited and jump the gun when presenting new ideas which frequently results in follow up corrections. The simple servo corrected
                              Message 14 of 23 , Jun 9, 2005
                              • 0 Attachment
                                Hi Tom,
                                 
                                Yes, I often get excited and jump the gun when presenting "new" ideas which frequently results in follow up corrections.
                                 
                                The simple servo "corrected" circuit, using the window comparator front end and emitter follower bridge, could be adapted to Robert's head application but unless you use a separate low voltage motor supply, it  may suffer from poor efficiency.  I am curious to see where the input signal comes from!
                                 
                                Drive LEDs from opamp outputs without a series resistor?  Spoken like a true beamer. 
                                Over the last few years, we have discussed and posted a number of circuits on this list that use opamps and comparators just this way, operating in the current limited output mode, to drive LEDs and charge batteries without external series resistors. 
                                 
                                I can't wait to see your ideas in action in the videos you sent to Wolf. 
                                 
                                best regards
                                 
                                wilf
                                 
                                 
                                ----- Original Message -----
                                Sent: Thursday, June 09, 2005 2:06 PM
                                Subject: [beam] Re: My first head design: input please

                                Hey Wilf,

                                Thanks for showing me how to upload files. Please delete the first
                                file I uploaded, when I opened it I discovered a mistake in the
                                schematic.

                                Sorry for coming across overbearing, I get excited and no disrespect
                                was intended.

                                I was pointing out in my post that one can rely on the internal
                                current limiting resistors in small signal op amps and that they're
                                rated for continuous duty. They're designed for this sort of abuse,
                                when I use op amps as led drivers I don't bother with resistors,
                                most op amps give out 12ma or so at a 5 volt power supply. You need
                                about 8 volts to get the full rating of 20ma. I have yet to blow out
                                an op amp (I did one time, but that was with a 3000 volt 6 uF
                                capacitor and a triggered spark gap. It failed in a very nasty
                                explosion!). I wasn't adressing nor did I refer to the npn/pnp
                                bridge or any other part of the circuitry, just the fact that you
                                can use the internal resistors.


                                You stated the heavily loaded op amps will limit the low output. Can
                                you elaborate or give a circuit example. I want to make sure we're
                                meaning the same thing; I just played with a LM358 and find that a
                                heavy load lowers the op amp low voltage output instead of limiting
                                it. The voltage drop starts to occur at 720 ohms (to ground) and the
                                more I load it the lower the voltage gets. I find that I can dial in
                                the low voltage output with the proper impedence. The floating
                                output is 1.84 volts at 5 volts supply for this particular chip.

                                -Tom



                              • Wilf Rigter
                                Oops, forgot to elaborate on the opamp loading problem. Let s take a circuit you have on hand, for example your simple servo corrected circuit, which is
                                Message 15 of 23 , Jun 9, 2005
                                • 0 Attachment
                                  Oops, forgot to elaborate on the opamp loading problem. 
                                   
                                  Let's take a circuit you have on hand, for example your simple servo corrected circuit, which is similar to the circuit we were discussing.
                                   
                                  With a motor installed in the emitter follower bridge,  measure the voltage between the opamp output pins (eg LM358 pin 1 and pin 7)
                                   
                                  When the motor is unloaded, the bridge/motor current is low, the opamp output current is low and the voltage between  pin 1 and 7 is 3.5V
                                   
                                  Now stall the motor to increase the bridge/motor current, the opamp opamp current will increase and the voltage between pin 1 and 7 will drop to 2.5V.
                                   
                                  If you were to also measure the voltage across the motor, you will have 2.5V with the motor unloaded and 1.5V when the motor is stalled which is what sparked my comment about efficiency.
                                   
                                  I had to assume a steady state (no fancy PWM) input condition to the opamps that would make the motor rotate continuously in one or the other direction and that you are using non-exotic (PN2222/PN2907) type transistors and that you are using one of the smaller WellStar motors with a winding resistance of about 10 ohms.    
                                   
                                  Conlusion:  heavy loading (increased output current) of an opamp causes an internal voltage drop inside the opamp that reduces the opamp output voltage swing. In other words there will be a larger voltage difference between the most positive opamp output level and the 5V supply or the most negative opamp output voltage level and GND respectively . 
                                   
                                  This case in point applies to the circuit we were discussing. 
                                   
                                  There are other circuits in which an external  pull up resistor is used to increase the positive output swing of an opamp, as I suggested to Robert for use with his opamp circuit driving the inputs of a 74AC motor driver.  Not to be confused with output "loading",  this pull up resistor is used to increase the positive output voltage and acually reduces opamp output current by providing a parallel current path to 5V.  Incidentally when the opamp output is pulled low, the pull up resistor will load the output and decrease the negative voltage swing and therefore increase the voltage difference between the most negative opamp output voltage level and GND.  
                                   
                                  Hope that clarifies it a little.
                                   
                                  best
                                   
                                  wilf 
                                   
                                   
                                  ----- Original Message -----
                                  Sent: Thursday, June 09, 2005 2:06 PM
                                  Subject: [beam] Re: My first head design: input please

                                  Hey Wilf,
                                  snip

                                  You stated the heavily loaded op amps will limit the low output. Can
                                  you elaborate or give a circuit example. I want to make sure we're
                                  meaning the same thing; I just played with a LM358 and find that a
                                  heavy load lowers the op amp low voltage output instead of limiting
                                  it. The voltage drop starts to occur at 720 ohms (to ground) and the
                                  more I load it the lower the voltage gets. I find that I can dial in
                                  the low voltage output with the proper impedence. The floating
                                  output is 1.84 volts at 5 volts supply for this particular chip.

                                  -Tom



                                • Joseph Charles
                                  Sorry! My brain must have been out to lunch (a midnight snack?) last night. The 3.5V (actually ~ 3.2V with my setup) is across the op-amp output pins. You only
                                  Message 16 of 23 , Jun 10, 2005
                                  • 0 Attachment
                                    Sorry! My brain must have been out to lunch (a midnight snack?) last night.
                                     
                                    The 3.5V (actually ~ 3.2V with my setup) is across the op-amp output pins. You only get around 1.8V across the motor (as Wilf has already mentioned). And that's unloaded. Not so good.
                                     
                                    I'm pleased you've found a more efficient driver Robert, and thanks for the original circuit - I've had fun playing with it.
                                     
                                    Jo

                                    Joseph Charles <jodicalhon@...> wrote:
                                    What with all the mucking around with resistors and such, you're probably better off with two op-amps driving an emitter follower bridge (Wilf's v1.5) if you have a high enough Vcc. With 5V supply you'd get around 3.5V across the motor.
                                     
                                     

                                    Send instant messages to your online friends http://au.messenger.yahoo.com

                                  • roboanalogtom
                                    Hey Wilf, Thanks for the clarification. Variable loading is a problem, they work great for heads though. My solution is to find a wall wart with the right
                                    Message 17 of 23 , Jun 11, 2005
                                    • 0 Attachment
                                      Hey Wilf,

                                      Thanks for the clarification.

                                      Variable loading is a problem, they work great for heads though. My
                                      solution is to find a wall wart with the right voltage and adjust
                                      the deadband as needed to avoid unwanted oscillations but I imagine
                                      the efficiency including motor/gearbox efficiency is around 10-15%.

                                      Due to poor performance I use the MC33030P dc servo chip for
                                      continuously variable analog servos (for walkers at least) that you
                                      can get at Digikey. Wha't nice is that it has braking logic on chip
                                      and with external amplifiers you can turn a cheap cordless
                                      screwdriver into a fairly precise servo with robust planetary gears.

                                      -Tom


                                      --- In beam@yahoogroups.com, "Wilf Rigter" <wrigter@d...> wrote:
                                      > Oops, forgot to elaborate on the opamp loading problem.
                                      >
                                      > Let's take a circuit you have on hand, for example your simple
                                      servo corrected circuit, which is similar to the circuit we were
                                      discussing.
                                      >
                                      > With a motor installed in the emitter follower bridge, measure
                                      the voltage between the opamp output pins (eg LM358 pin 1 and pin 7)
                                      >
                                      > When the motor is unloaded, the bridge/motor current is low, the
                                      opamp output current is low and the voltage between pin 1 and 7 is
                                      3.5V
                                      >
                                      > Now stall the motor to increase the bridge/motor current, the
                                      opamp opamp current will increase and the voltage between pin 1 and
                                      7 will drop to 2.5V.
                                      >
                                      > If you were to also measure the voltage across the motor, you will
                                      have 2.5V with the motor unloaded and 1.5V when the motor is stalled
                                      which is what sparked my comment about efficiency.
                                      >
                                      > I had to assume a steady state (no fancy PWM) input condition to
                                      the opamps that would make the motor rotate continuously in one or
                                      the other direction and that you are using non-exotic
                                      (PN2222/PN2907) type transistors and that you are using one of the
                                      smaller WellStar motors with a winding resistance of about 10
                                      ohms.
                                      >
                                      > Conlusion: heavy loading (increased output current) of an opamp
                                      causes an internal voltage drop inside the opamp that reduces the
                                      opamp output voltage swing. In other words there will be a larger
                                      voltage difference between the most positive opamp output level and
                                      the 5V supply or the most negative opamp output voltage level and
                                      GND respectively .
                                      >
                                      > This case in point applies to the circuit we were discussing.
                                      >
                                      > There are other circuits in which an external pull up resistor is
                                      used to increase the positive output swing of an opamp, as I
                                      suggested to Robert for use with his opamp circuit driving the
                                      inputs of a 74AC motor driver. Not to be confused with
                                      output "loading", this pull up resistor is used to increase the
                                      positive output voltage and acually reduces opamp output current by
                                      providing a parallel current path to 5V. Incidentally when the
                                      opamp output is pulled low, the pull up resistor will load the
                                      output and decrease the negative voltage swing and therefore
                                      increase the voltage difference between the most negative opamp
                                      output voltage level and GND.
                                      >
                                      > Hope that clarifies it a little.
                                      >
                                      > best
                                      >
                                      > wilf
                                      >
                                      >
                                      > ----- Original Message -----
                                      > From: roboanalogtom
                                      > To: beam@yahoogroups.com
                                      > Sent: Thursday, June 09, 2005 2:06 PM
                                      > Subject: [beam] Re: My first head design: input please
                                      >
                                      >
                                      > Hey Wilf,
                                      > snip
                                      >
                                      > You stated the heavily loaded op amps will limit the low output.
                                      Can
                                      > you elaborate or give a circuit example. I want to make sure
                                      we're
                                      > meaning the same thing; I just played with a LM358 and find that
                                      a
                                      > heavy load lowers the op amp low voltage output instead of
                                      limiting
                                      > it. The voltage drop starts to occur at 720 ohms (to ground) and
                                      the
                                      > more I load it the lower the voltage gets. I find that I can
                                      dial in
                                      > the low voltage output with the proper impedence. The floating
                                      > output is 1.84 volts at 5 volts supply for this particular chip.
                                      >
                                      > -Tom
                                      >
                                      >
                                      >
                                      >
                                      >
                                      >
                                      > -------------------------------------------------------------------
                                      -----------
                                      >
                                      > Yahoo! Groups Links
                                      >
                                      > a.. To visit your group on the web, go to:
                                      > http://groups.yahoo.com/group/beam/
                                      >
                                      > b.. To unsubscribe from this group, send an email to:
                                      > beam-unsubscribe@yahoogroups.com
                                      >
                                      > c.. Your use of Yahoo! Groups is subject to the Yahoo! Terms
                                      of Service.
                                    Your message has been successfully submitted and would be delivered to recipients shortly.