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

Re: Concerns about current draw in a micromouse.

Expand Messages
  • Ben Estroff
    A few moment s research yielded the L293, functionally identical but with a 2A maximum. Should that be enough for a lightweight micromouse? I think the battery
    Message 1 of 18 , Oct 2, 2007
    • 0 Attachment
      A few moment's research yielded the L293, functionally identical but
      with a 2A maximum.
      Should that be enough for a lightweight micromouse?
      I think the battery will cut out before the current draw gets much
      higher than that.

      On 10/2/07, Ben Estroff <birdinhand@...> wrote:
      > Hello, everyone.
      >
      > Despite Murphy's law trying to stop me from getting this robot done in
      > the last days before the competition, I'm still fighting.
      >
      > However, we've recently had to go to new motors in order to achieve
      > any kind of decent speed, and I'm worried about the ability of the
      > H-bridge to source the required amount of current.
      > Here's the new motor type.
      > http://www.robotshop.ca/home/products/robot-parts/motors/gear-motors/spur-gearmotors/lynxmotion-ghm-10-gear-head-motor.html
      >
      > I'm using the SN754410 TI Dual-H bridge IC currently, and that can
      > easily handle the current of our older motors (the MO-201 found here:
      > http://www.oceancontrols.com.au/motors/dc/dc_motors.htm)
      > Even under full load, but the new Lynxmotion motors will far surpass
      > the current draw under a full load.
      >
      > We're intending to use only IR sensors (rangefinders and encoders) so
      > that we don't hit a wall, and the robot should be extremely light, but
      > I'm still concerned that the SN754410 will simply go up in smoke when
      > executing quick turns.
      >
      > Are there any other suggested h-bridges in the same vein as the SN754410?
      >
      >
      >
      > Also, any good sources (in the northwest) for a mini-tamiya connector
      > that doesn't already have pigtails? I've only found one place in
      > Wisconsin that seems to be selling the individual parts.
      >
      > Thanks in advance!
      >
    • dan michaels
      ... The L293D is pin compatible with the 754410, but it s current drive is much less. L293D = 600-mA, 754410 = 1-Amp. I think there is also an smt part, L293E
      Message 2 of 18 , Oct 2, 2007
      • 0 Attachment
        --- In SeattleRobotics@yahoogroups.com, "Ben Estroff"
        <birdinhand@...> wrote:
        >


        The L293D is pin compatible with the 754410, but it's current drive
        is much less. L293D = 600-mA, 754410 = 1-Amp.

        I think there is also an smt part, L293E or some such, with possibly
        different spec.


        > A few moment's research yielded the L293, functionally identical but
        > with a 2A maximum.
        > Should that be enough for a lightweight micromouse?
        > I think the battery will cut out before the current draw gets much
        > higher than that.
        >
        > On 10/2/07, Ben Estroff <birdinhand@...> wrote:
        > > Hello, everyone.
        > >
        > > Despite Murphy's law trying to stop me from getting this robot
        done in
        > > the last days before the competition, I'm still fighting.
        > >
        > > However, we've recently had to go to new motors in order to
        achieve
        > > any kind of decent speed, and I'm worried about the ability of the
        > > H-bridge to source the required amount of current.
        > > Here's the new motor type.
        > > http://www.robotshop.ca/home/products/robot-parts/motors/gear-
        motors/spur-gearmotors/lynxmotion-ghm-10-gear-head-motor.html
        > >
        > > I'm using the SN754410 TI Dual-H bridge IC currently, and that can
        > > easily handle the current of our older motors (the MO-201 found
        here:
        > > http://www.oceancontrols.com.au/motors/dc/dc_motors.htm)
        > > Even under full load, but the new Lynxmotion motors will far
        surpass
        > > the current draw under a full load.
        > >
        > > We're intending to use only IR sensors (rangefinders and
        encoders) so
        > > that we don't hit a wall, and the robot should be extremely
        light, but
        > > I'm still concerned that the SN754410 will simply go up in smoke
        when
        > > executing quick turns.
        > >
        > > Are there any other suggested h-bridges in the same vein as the
        SN754410?
        > >
        > >
        > >
        > > Also, any good sources (in the northwest) for a mini-tamiya
        connector
        > > that doesn't already have pigtails? I've only found one place in
        > > Wisconsin that seems to be selling the individual parts.
        > >
        > > Thanks in advance!
        > >
        >
      • Peter Harrison
        ... The L293 comes in several flavours. The L293D has flyback diodes in the output, the others don t. You will need the diodes. As far as I know, all the 293s
        Message 3 of 18 , Oct 2, 2007
        • 0 Attachment
          Ben Estroff wrote:
          > A few moment's research yielded the L293, functionally identical but
          > with a 2A maximum.
          > Should that be enough for a lightweight micromouse?
          > I think the battery will cut out before the current draw gets much
          > higher than that.
          >
          > On 10/2/07, Ben Estroff <birdinhand@...> wrote:
          >> Hello, everyone.
          >>
          >> Despite Murphy's law trying to stop me from getting this robot done in
          >> the last days before the competition, I'm still fighting.
          >>
          >> However, we've recently had to go to new motors in order to achieve
          >> any kind of decent speed, and I'm worried about the ability of the
          >> H-bridge to source the required amount of current.
          >> Here's the new motor type.
          >> http://www.robotshop.ca/home/products/robot-parts/motors/gear-motors/spur-gearmotors/lynxmotion-ghm-10-gear-head-motor.html
          >>
          >> I'm using the SN754410 TI Dual-H bridge IC currently, and that can
          >> easily handle the current of our older motors (the MO-201 found here:
          >> http://www.oceancontrols.com.au/motors/dc/dc_motors.htm)
          >> Even under full load, but the new Lynxmotion motors will far surpass
          >> the current draw under a full load.
          >>
          >> We're intending to use only IR sensors (rangefinders and encoders) so
          >> that we don't hit a wall, and the robot should be extremely light, but
          >> I'm still concerned that the SN754410 will simply go up in smoke when
          >> executing quick turns.
          >>
          >> Are there any other suggested h-bridges in the same vein as the SN754410?

          The L293 comes in several flavours. The L293D has flyback diodes in the
          output, the others don't. You will need the diodes. As far as I know,
          all the 293s have the same current rating which is 1A, the same as the
          SN754410. The 2A figure may be the total current through both bridges.

          I believe you can safely stack these on top of each other to increase
          the current handling. Bear in mind that the bottom one of a stacked pair
          will not be able to dissipate as much heat.

          The 293 and its cousins are thermally protected and will shut down
          rather than release the magic smoke.

          If you think the batteries will sag before delivering 2A they may not
          have enough capacity to deliver current spikes to your motors anyway.
          NiCd and NiMh cells of any reasonable capacity (say 600mAh) will be able
          to provide much more than 2A before giving in.

          The motors you describe are very powerful and it is likely the wheels
          will spin rather than stall if your mouse is under 2kg in weight and on
          a smooth surface. Of course, if you get the carpet wound round them,
          something will have to give :)

          Pete Harrison
        • Jon Hylands
          On Tue, 02 Oct 2007 23:01:27 -0000, dan michaels ... I think he s thinking of the L298, which is rated for 2 amps per channel... The TI
          Message 4 of 18 , Oct 2, 2007
          • 0 Attachment
            On Tue, 02 Oct 2007 23:01:27 -0000, "dan michaels" <oric_dan@...>
            wrote:

            > The L293D is pin compatible with the 754410, but it's current drive
            > is much less. L293D = 600-mA, 754410 = 1-Amp.

            I think he's thinking of the L298, which is rated for 2 amps per channel...

            The TI chip won't go up in smoke if you exceed the current - it will just
            go into thermal shutdown.

            Later,
            Jon

            --------------------------------------------------------------
            Jon Hylands Jon@... http://www.huv.com/jon

            Project: Micro Raptor (Small Biped Velociraptor Robot)
            http://www.huv.com/blog
          • Ben Estroff
            Yes, it seemed confusing that the SN754410 was an improved functional replacement for the L293. Since my protoboard space is at a premium, any opinions on
            Message 5 of 18 , Oct 2, 2007
            • 0 Attachment
              Yes, it seemed confusing that the SN754410 was an "improved functional
              replacement" for the L293.
              Since my protoboard space is at a premium, any opinions on the number of
              SN754410s I should put on in parallel?
              My hazy recollections of Kirchoff's current equations suggest that it might
              work, but I'd rather just use a single chip.
              I'll be putting everything in sockets, because I'm assuming the worst will
              happen and things will need replacement.

              This is the battery I'm intending to use:
              http://www.all-battery.com/index.asp?PageAction=VIEWPROD&ProdID=508
              but only if I can get a male and female Tamiya so I can charge the darn
              thing and hook it to the board.

              I'd much rather crimp to the battery's leads.


              On 10/2/07, Jon Hylands <jon@...> wrote:
              >
              > On Tue, 02 Oct 2007 23:01:27 -0000, "dan michaels" <oric_dan@...<oric_dan%40yahoo.com>
              > >
              > wrote:
              >
              > > The L293D is pin compatible with the 754410, but it's current drive
              > > is much less. L293D = 600-mA, 754410 = 1-Amp.
              >
              > I think he's thinking of the L298, which is rated for 2 amps per
              > channel...
              >
              > The TI chip won't go up in smoke if you exceed the current - it will just
              > go into thermal shutdown.
              >
              > Later,
              > Jon
              >
              > ----------------------------------------------------------
              > Jon Hylands Jon@... <Jon%40huv.com> http://www.huv.com/jon
              >
              > Project: Micro Raptor (Small Biped Velociraptor Robot)
              > http://www.huv.com/blog
              >
              >


              [Non-text portions of this message have been removed]
            • Randy M. Dumse
              Ben Estroff talked of: Tuesday, October 02, 2007 7:16 PM ... Two. In the data sheet it tells you which halves of the bridge to parallel. Better to parallel
              Message 6 of 18 , Oct 2, 2007
              • 0 Attachment
                Ben Estroff talked of: Tuesday, October 02, 2007 7:16 PM
                > Since my protoboard space is at a premium, any opinions on
                > the number of SN754410s I should put on in parallel?

                Two.

                In the data sheet it tells you which halves of the bridge to
                parallel. Better to parallel that way, and do one bridge per
                chip, than to stack two and parallel between different chips.

                You can parallel two chips, but it only works through an odd
                sharing of thermal cutouts. Bipolar transistors do not parallel
                well, and one tries to take the full load. Of course if its
                internal protection saves it, the other gets the load, and they
                sort of switch back and forth carrying too much current, and
                heating and cooling alternatively.

                Did you look at the specs of your motors?

                12vdc 416rpm, Gear Head Motor / Part no. GHM-10
                Torque = 50.83 oz. in.
                Weight = 2.08 oz.
                Reduction = 30 :1
                Length = 43mm (motor and gear)
                Length = 17.5mm (shaft only)
                Diameter = 27mm (motor and gear)
                Diameter = 4mm (shaft) (0.157")
                Current = 123mA (at 12v no load)
                Current = 2.05A (at 12v locked shaft)

                So you need 2A capacity as a minimum. Two 754410's will do it. I
                think they are listed as 1.2A temporary load. Heat sinks tied to
                the chip and the middle ground pins would be a very good idea.

                Your better replacement would probably be a Infineon TLE-5026-2.
                This is the part we use in our NMIH-0050 H-bridge. 5A if
                properly heat sinked. Easily does about 2A open air. So this is
                probably the smallest and best protected bridge you can get for
                your motor and supply.

                Randy
                www.newmicros.com
              • Ben Estroff
                Hm, I d like to use the Infineon product, but surface mount is a bit beyond my capability to solder. To verify: you re suggesting just using one full bridge
                Message 7 of 18 , Oct 2, 2007
                • 0 Attachment
                  Hm, I'd like to use the Infineon product, but surface mount is a bit beyond
                  my capability to solder.
                  To verify: you're suggesting just using one full bridge per chip, correct?

                  And what heat sinks should I be looking at for the SN754410s? I'm going to
                  be placing an order for IC sockets (for the SNs and the inverter and pretty
                  much every component save the clock) with digikey because this project's
                  seemingly cursed, and I don't want to take any chances at this late date.
                  Might as well pick up the heatsinks from there, if possible.

                  On 10/2/07, Randy M. Dumse <rmd@...> wrote:
                  >
                  > Ben Estroff talked of: Tuesday, October 02, 2007 7:16 PM
                  > > Since my protoboard space is at a premium, any opinions on
                  > > the number of SN754410s I should put on in parallel?
                  >
                  > Two.
                  >
                  > In the data sheet it tells you which halves of the bridge to
                  > parallel. Better to parallel that way, and do one bridge per
                  > chip, than to stack two and parallel between different chips.
                  >
                  > You can parallel two chips, but it only works through an odd
                  > sharing of thermal cutouts. Bipolar transistors do not parallel
                  > well, and one tries to take the full load. Of course if its
                  > internal protection saves it, the other gets the load, and they
                  > sort of switch back and forth carrying too much current, and
                  > heating and cooling alternatively.
                  >
                  > Did you look at the specs of your motors?
                  >
                  > 12vdc 416rpm, Gear Head Motor / Part no. GHM-10
                  > Torque = 50.83 oz. in.
                  > Weight = 2.08 oz.
                  > Reduction = 30 :1
                  > Length = 43mm (motor and gear)
                  > Length = 17.5mm (shaft only)
                  > Diameter = 27mm (motor and gear)
                  > Diameter = 4mm (shaft) (0.157")
                  > Current = 123mA (at 12v no load)
                  > Current = 2.05A (at 12v locked shaft)
                  >
                  > So you need 2A capacity as a minimum. Two 754410's will do it. I
                  > think they are listed as 1.2A temporary load. Heat sinks tied to
                  > the chip and the middle ground pins would be a very good idea.
                  >
                  > Your better replacement would probably be a Infineon TLE-5026-2.
                  > This is the part we use in our NMIH-0050 H-bridge. 5A if
                  > properly heat sinked. Easily does about 2A open air. So this is
                  > probably the smallest and best protected bridge you can get for
                  > your motor and supply.
                  >
                  > Randy
                  > www.newmicros.com
                  >
                  >
                  >


                  [Non-text portions of this message have been removed]
                • Randy M. Dumse
                  Ben Estroff queried: Wednesday, October 03, 2007 12:20 AM ... We use the leaded part. It is the size of a TO220. Just more leads. ... Not enough context to
                  Message 8 of 18 , Oct 3, 2007
                  • 0 Attachment
                    Ben Estroff queried: Wednesday, October 03, 2007 12:20 AM
                    > Hm, I'd like to use the Infineon product, but surface mount
                    > is a bit beyond my capability to solder.

                    We use the leaded part. It is the size of a TO220. Just more
                    leads.

                    > To verify: you're suggesting just using one full bridge per
                    > chip, correct?

                    Not enough context to answer this simply, because I don't know
                    which chips you are asking the question about.

                    One TLE=5206-2 per motor. Yes.

                    Two 754410's per motor. Yes. I looked at the data sheet to try
                    to find which half-bridges should be paralleled, but couldn't
                    find what I was pretty sure I'd seen before. So I know there may
                    be a prefered set to parallel, but I can't say which.
                    http://focus.ti.com/lit/ds/slrs007b/slrs007b.pdf

                    You might pick up some information with a Google search on:
                    parallel 754410.

                    Randy
                    www.newmicros.com
                  • Peter Harrison
                    ... They look nice. If anyone else goes searching, Infineon have them under throttle control bridges and the part number is TLE 5026-2. Do you know of a handy
                    Message 9 of 18 , Oct 3, 2007
                    • 0 Attachment
                      Randy M. Dumse wrote:

                      > Your better replacement would probably be a Infineon TLE-5026-2.
                      > This is the part we use in our NMIH-0050 H-bridge. 5A if
                      > properly heat sinked. Easily does about 2A open air. So this is
                      > probably the smallest and best protected bridge you can get for
                      > your motor and supply.

                      They look nice. If anyone else goes searching, Infineon have them under
                      throttle control bridges and the part number is TLE 5026-2.

                      Do you know of a handy supplier?
                    • Ben Estroff
                      That does look much better. Only problem is that I ve been fighting the PWM module on this PIC18F4431 and have yet to get it outputting complementary PWM
                      Message 10 of 18 , Oct 3, 2007
                      • 0 Attachment
                        That does look much better. Only problem is that I've been fighting the PWM
                        module on this PIC18F4431 and have yet to get it outputting complementary
                        PWM signals. My current setup relies on the simple patch of throwing the
                        signal through an inverter, something that the Infineon TLE 5206-2 seems
                        require, unless my application falls into 'locked anti-phase', but it's not
                        clear how that feature is implemented/wired. I can certainly provide the
                        dedicated PWM ports, but not ports that have to switch between PWM and
                        constant high. The PIC's registers can't handle that level of custom
                        controls.

                        http://www.infineon.com/dgdl/tle5206-2_20010619.pdf?folderId=db3a304412b407950112b43820d56a4a&fileId=db3a304412b407950112b438215b6a4b
                        For the datasheet. l'd need to get on ordering this right away, so all
                        discussion of what I should be doing will be very helpful.

                        And yes, a supplier would be good. Better still if I can get samples of the
                        product, trying to build on a shoestring (self-imposed limitation) but the
                        time has come to get this thing done.

                        On 10/3/07, Randy M. Dumse <rmd@...> wrote:
                        >
                        > Ben Estroff queried: Wednesday, October 03, 2007 12:20 AM
                        > > Hm, I'd like to use the Infineon product, but surface mount
                        > > is a bit beyond my capability to solder.
                        >
                        > We use the leaded part. It is the size of a TO220. Just more
                        > leads.
                        >
                        > > To verify: you're suggesting just using one full bridge per
                        > > chip, correct?
                        >
                        > Not enough context to answer this simply, because I don't know
                        > which chips you are asking the question about.
                        >
                        > One TLE=5206-2 per motor. Yes.
                        >
                        > Two 754410's per motor. Yes. I looked at the data sheet to try
                        > to find which half-bridges should be paralleled, but couldn't
                        > find what I was pretty sure I'd seen before. So I know there may
                        > be a prefered set to parallel, but I can't say which.
                        > http://focus.ti.com/lit/ds/slrs007b/slrs007b.pdf
                        >
                        > You might pick up some information with a Google search on:
                        > parallel 754410.
                        >
                        > Randy
                        > www.newmicros.com
                        >
                        >
                        >


                        [Non-text portions of this message have been removed]
                      • Randy M. Dumse
                        ... We use Arrow www.arrow.com but checking the have no stock. Randy www.newmicros.com
                        Message 11 of 18 , Oct 3, 2007
                        • 0 Attachment
                          Peter Harrison asked:
                          > Do you know of a handy supplier?

                          We use Arrow www.arrow.com but checking the have no stock.

                          Randy
                          www.newmicros.com
                        • Ben Estroff
                          Dang. The UW s got a representative who sent me samples of the SN754410s in the first place. Well, worth a shot calling them again to maybe arrange it, the lab
                          Message 12 of 18 , Oct 3, 2007
                          • 0 Attachment
                            Dang. The UW's got a representative who sent me samples of the SN754410s in
                            the first place.
                            Well, worth a shot calling them again to maybe arrange it, the lab has other
                            applications for these H-Bridges beyond the current project.

                            On 10/3/07, Randy M. Dumse <rmd@...> wrote:
                            >
                            > Peter Harrison asked:
                            > > Do you know of a handy supplier?
                            >
                            > We use Arrow www.arrow.com but checking the have no stock.
                            >
                            > Randy
                            > www.newmicros.com
                            >
                            >
                            >


                            [Non-text portions of this message have been removed]
                          • Randy M. Dumse
                            Ben Estroff remarked: Wednesday, October 03, 2007 2:14 PM ... Sorry, a moment of self promotion: That s why we so much prefer the DSP processors in our Pod
                            Message 13 of 18 , Oct 3, 2007
                            • 0 Attachment
                              Ben Estroff remarked: Wednesday, October 03, 2007 2:14 PM
                              > That does look much better. Only problem is that I've been
                              > fighting the PWM module on this PIC18F4431 and have yet to
                              > get it outputting complementary PWM signals.

                              Sorry, a moment of self promotion: That's why we so much prefer
                              the DSP processors in our 'Pod line of micros. They can easily
                              generate the PWM modes you need. While more complicated
                              hardware, actual use is so much easier, because the hardware
                              does so much of this sort of management for you. It is a proper
                              micro designed for motion control, and it excels at it. Plus
                              it's a 40 MIP's processor, so you have plenty left over to do
                              other tasks.

                              > My current setup relies on the simple patch of throwing
                              > the signal through an inverter, something that the
                              > Infineon TLE 5206-2 seems require, unless my application
                              > falls into 'locked anti-phase', but it's not clear how
                              > that feature is implemented/wired.

                              You make locked antiphase by using one PWM output and using an
                              inverter, just what you were talking about, actually. Then at
                              50% PWM, the wheel is still, higher forward, lower backwards (or
                              opposite depending on wiring choices et. al.).

                              You could call my office and ask them to sell you a couple at $5
                              plus shipping. It's not normally what we do, kind of like self
                              competition for our finished products, but we do help folks that
                              way sometimes. Call early, like 7A to 3PM central, ask for "BC".
                              I am leaving for a trip for a week to UNI on university matters,
                              so I won't be there. I'm copying BC so he knows this is my
                              suggestion.

                              Randy
                              www.newmicros.com
                            • Randy M. Dumse
                              Ben Estroff ... Ah. Pam tells me Digikey has them as TLE5206-2. In stock at $6.20. Randy www.newmicros.com
                              Message 14 of 18 , Oct 3, 2007
                              • 0 Attachment
                                Ben Estroff
                                > Dang.

                                Ah. Pam tells me Digikey has them as TLE5206-2. In stock at
                                $6.20.

                                Randy
                                www.newmicros.com
                              • Ben Estroff
                                Sweet. And that s great news, regarding the anti-lock phase mode. Thank you very much, Randy. ... [Non-text portions of this message have been removed]
                                Message 15 of 18 , Oct 3, 2007
                                • 0 Attachment
                                  Sweet.
                                  And that's great news, regarding the anti-lock phase mode.
                                  Thank you very much, Randy.


                                  On 10/3/07, Randy M. Dumse <rmd@...> wrote:
                                  >
                                  > Ben Estroff
                                  > > Dang.
                                  >
                                  > Ah. Pam tells me Digikey has them as TLE5206-2. In stock at
                                  > $6.20.
                                  >
                                  > Randy
                                  > www.newmicros.com
                                  >
                                  >
                                  >


                                  [Non-text portions of this message have been removed]
                                • Next Kiwi
                                  The ultimate Hbridge chip to use these days looks to be the BTS7960 (half bridge) See http://www.robotpower.com/products/simple-h_info.html Cheers Keith
                                  Message 16 of 18 , Oct 3, 2007
                                  • 0 Attachment
                                    The ultimate Hbridge chip to use these days looks to be the
                                    BTS7960 (half bridge)

                                    See http://www.robotpower.com/products/simple-h_info.html

                                    Cheers
                                    Keith


                                    [Non-text portions of this message have been removed]
                                  • Ben Estroff
                                    Since I m anticipating this all going downhill fast, I m wanting to use sockets to replace things like the PIC, sensors, and smoothing capacitors should
                                    Message 17 of 18 , Oct 5, 2007
                                    • 0 Attachment
                                      Since I'm anticipating this all going downhill fast, I'm wanting to use
                                      sockets to replace things like the PIC, sensors, and smoothing capacitors
                                      should something blow.
                                      But I'm guessing I can't do that with the TLE5602, due to the current draw.

                                      Should I use sockets at all, or just get desoldering wick?

                                      On 10/3/07, Next Kiwi <next@...> wrote:
                                      >
                                      > The ultimate Hbridge chip to use these days looks to be the
                                      > BTS7960 (half bridge)
                                      >
                                      > See http://www.robotpower.com/products/simple-h_info.html
                                      >
                                      > Cheers
                                      > Keith
                                      >
                                      > [Non-text portions of this message have been removed]
                                      >
                                      >
                                      >


                                      [Non-text portions of this message have been removed]
                                    Your message has been successfully submitted and would be delivered to recipients shortly.