Shop OBEX P1 Docs P2 Docs Learn Events
Stamps and motor control — Parallax Forums

Stamps and motor control

ArchiverArchiver Posts: 46,084
edited 2000-04-10 16:36 in General Discussion
Hello,

I have an old Hero 2000 robot arm, which was sold by Heathkit. The same arm
was sold separately from the robot as an "arm trainer". Not counting the
gripper, the arm is moved by four identical, 12V gearmotors. One for
shoulder, one for elbow, and two for wrist (the wrist is a differential
mechanism).

I have been trying to create an interface so that I can control the arm with
my PC. I am using the Basic Stamp II for PC interfacing. To drive the
motors, I am using SN754410NE motor driver chips. For PWM speed control, I'm
using the PAK Va from AWC.

Here's the problem: I'm a bit new at motors, and I'm having some troubles.
At low PWM duty cycles, the motor makes a high pitched sound and won't move.
At slightly higher speeds, the tone is still there but the motor starts to
move. What is this tone?

Also, I was under the impression that the SN754410NE had built-in diodes for
backEMF protection. One of the chips exploded (sparks, little divot out of
the top of the IC) today, leading me to believe that something was wrong.

Does anyone have experience working with the Hero 2000 arm? What I'd really
love is a circuit to build that lets the PC or Stamp (or other
microcontroller) drive the arm's motors. I can handle control algorithms,
etc. I'm just having trouble with the hardware interfacing. I'd like to see
someone's
circuit to see where I'm going wrong.

Thanks,
Joe Driscoll

Comments

  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 00:32
    HI,

    Don't know about the SN754410NE chip but the MC33030 chip is very easy to use
    with a STAMP2 . I am using them with an x-Y control system and it is fairly
    simple.
    I found that the STAMP needs expansion for both analog signal and digital data
    handling.
    If you want I can email my prgm ( not yet fully debugged ) and it will give you
    some idea of the complexity.
    The noises are the PWM causing the windings to vibrate. If you are using a PWM
    to control speed
    you must be sure that the inductance of the motor is high enough so that the
    current continues to flow during the off
    time. Try the chip in a simple config before using it in a STAMP control system
    to make sure you have that part working smoothly.

    Also you need diodes to make sure that the stored energy in the inductance has
    somewhere to go.. In the case of full H bridge that means 4 diodes
    for each motor. Usually they are built in to the control chip.

    ( PLEASE NO MORE DIODE RELAY STUFF!!! [noparse]:)[/noparse])
    To: <basicstamps@egroups.com>
    Date sent: Thu, 6 Apr 2000 17:51:40 -0500
    From: "Joseph Driscoll" <driscoll85@h...>
    Send reply to: basicstamps@egroups.com
    Subject: [noparse][[/noparse]basicstamps] Stamps and motor control

    > Hello,
    >
    > I have an old Hero 2000 robot arm, which was sold by Heathkit. The same arm
    > was sold separately from the robot as an "arm trainer". Not counting the
    > gripper, the arm is moved by four identical, 12V gearmotors. One for
    > shoulder, one for elbow, and two for wrist (the wrist is a differential
    > mechanism).
    >
    > I have been trying to create an interface so that I can control the arm with
    > my PC. I am using the Basic Stamp II for PC interfacing. To drive the
    > motors, I am using SN754410NE motor driver chips. For PWM speed control, I'm
    > using the PAK Va from AWC.
    >
    > Here's the problem: I'm a bit new at motors, and I'm having some troubles.
    > At low PWM duty cycles, the motor makes a high pitched sound and won't move.
    > At slightly higher speeds, the tone is still there but the motor starts to
    > move. What is this tone?
    >
    > Also, I was under the impression that the SN754410NE had built-in diodes for
    > backEMF protection. One of the chips exploded (sparks, little divot out of
    > the top of the IC) today, leading me to believe that something was wrong.
    >
    > Does anyone have experience working with the Hero 2000 arm? What I'd really
    > love is a circuit to build that lets the PC or Stamp (or other
    > microcontroller) drive the arm's motors. I can handle control algorithms,
    > etc. I'm just having trouble with the hardware interfacing. I'd like to see
    > someone's
    > circuit to see where I'm going wrong.
    >
    > Thanks,
    > Joe Driscoll
    >
    >
    >
    >


    ron... the motor man...
    Ronald Doctors
    http://www.members.home.net/rdoctors
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 01:50
    Hi,

    Thanks for the reply. Yes, please send me your program -- that would help a
    lot. When you say "you must be sure that the inductance of the motor is high
    enough so that the current continues to flow during the off time", I'm not
    sure I understand that part. With the SN754410NE and the L293D, when the
    enable pin is low, the outputs are at a high impedance state. Is that a
    problem? It seems like the back emf produced during that "off" time needs to
    go somewhere. Maybe this is the problem?

    Thanks,
    Joe Driscoll


    Original Message
    From: Ronald Doctors <rdoctors@h...>
    To: <basicstamps@egroups.com>
    Sent: Thursday, April 06, 2000 6:32 PM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > HI,
    >
    > Don't know about the SN754410NE chip but the MC33030 chip is very easy to
    use
    > with a STAMP2 . I am using them with an x-Y control system and it is
    fairly simple.
    > I found that the STAMP needs expansion for both analog signal and digital
    data handling.
    > If you want I can email my prgm ( not yet fully debugged ) and it will
    give you some idea of the complexity.
    > The noises are the PWM causing the windings to vibrate. If you are using
    a PWM to control speed
    > you must be sure that the inductance of the motor is high enough so that
    the current continues to flow during the off
    > time. Try the chip in a simple config before using it in a STAMP control
    system to make sure you have that part working smoothly.
    >
    > Also you need diodes to make sure that the stored energy in the inductance
    has somewhere to go.. In the case of full H bridge that means 4 diodes
    > for each motor. Usually they are built in to the control chip.
    >
    > ( PLEASE NO MORE DIODE RELAY STUFF!!! [noparse]:)[/noparse])
    > To: <basicstamps@egroups.com>
    > Date sent: Thu, 6 Apr 2000 17:51:40 -0500
    > From: "Joseph Driscoll" <driscoll85@h...>
    > Send reply to: basicstamps@egroups.com
    > Subject: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    > > Hello,
    > >
    > > I have an old Hero 2000 robot arm, which was sold by Heathkit. The same
    arm
    > > was sold separately from the robot as an "arm trainer". Not counting the
    > > gripper, the arm is moved by four identical, 12V gearmotors. One for
    > > shoulder, one for elbow, and two for wrist (the wrist is a differential
    > > mechanism).
    > >
    > > I have been trying to create an interface so that I can control the arm
    with
    > > my PC. I am using the Basic Stamp II for PC interfacing. To drive the
    > > motors, I am using SN754410NE motor driver chips. For PWM speed control,
    I'm
    > > using the PAK Va from AWC.
    > >
    > > Here's the problem: I'm a bit new at motors, and I'm having some
    troubles.
    > > At low PWM duty cycles, the motor makes a high pitched sound and won't
    move.
    > > At slightly higher speeds, the tone is still there but the motor starts
    to
    > > move. What is this tone?
    > >
    > > Also, I was under the impression that the SN754410NE had built-in diodes
    for
    > > backEMF protection. One of the chips exploded (sparks, little divot out
    of
    > > the top of the IC) today, leading me to believe that something was
    wrong.
    > >
    > > Does anyone have experience working with the Hero 2000 arm? What I'd
    really
    > > love is a circuit to build that lets the PC or Stamp (or other
    > > microcontroller) drive the arm's motors. I can handle control
    algorithms,
    > > etc. I'm just having trouble with the hardware interfacing. I'd like to
    see
    > > someone's
    > > circuit to see where I'm going wrong.
    > >
    > > Thanks,
    > > Joe Driscoll
    > >
    > >
    > >
    > >
    >
    >
    > ron... the motor man...
    > Ronald Doctors
    > http://www.members.home.net/rdoctors
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 02:36

    Original Message
    From: Dave Paton <suprdave@p...>
    To: <basicstamps@egroups.com>
    Sent: Thursday, April 06, 2000 8:48 PM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > On 4/6/00 at 5:51 PM, Joseph Driscoll <driscoll85@h...> wrote:
    >
    > > Does anyone have experience working with the Hero 2000 arm?
    >
    > Way back when I assembled one and installed it into a hero 2K for my
    junior high science class. The only things I recall vividly are that:
    >
    > 1. the motors were driven by large heatsinked H-bridges. I suggest getting
    a much more robust controller. 5A @ 12V would not be unreasonable.

    I've measured the current draw, and it's usually around 0.5 amp.


    > 2. failure to monitor the limit switches caused biiiiiiiig problems. The
    cable drive is more fragile than it looks.

    My plan is to monitor the limit switches from the stamp, by running them to
    input lines. Or, I could just have a hardware shutoff, using an OR gate or
    something.


    > I have a vague recollection of encoders on the arm as well, but this might
    be a figment of my imagination. I do remembr clearly the ordeal that getting
    to e arm to 'home' was. It seems the hero 2K SW was very picky.

    Right, there are optical encoders on all of the motors. They output pulses,
    which are squared up via a LM393. I plan to use a PAK VII pulse coprocessor
    (from AWCE) to keep track of pulses. The stamp can then ask the PAK
    periodically for pulse counts, and that way the position of the arm is known
    (assuming it started from a known location).

    I plan to use the stamp as an arm "server": it polls sensors (and the PAK
    VII) and sends that data to a PC. The PC then sends speed and direction
    commands. The control algorithms can then be carried out on the PC.
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 02:48
    On 4/6/00 at 5:51 PM, Joseph Driscoll <driscoll85@h...> wrote:

    > Does anyone have experience working with the Hero 2000 arm?

    Way back when I assembled one and installed it into a hero 2K for my junior high
    science class. The only things I recall vividly are that:

    1. the motors were driven by large heatsinked H-bridges. I suggest getting a
    much more robust controller. 5A @ 12V would not be unreasonable.

    2. failure to monitor the limit switches caused biiiiiiiig problems. The cable
    drive is more fragile than it looks.

    I have a vague recollection of encoders on the arm as well, but this might be a
    figment of my imagination. I do remembr clearly the ordeal that getting to e arm
    to 'home' was. It seems the hero 2K SW was very picky.

    -dave

    Passion must be present. Everything else can be learned.
    WCCR chief engineer, Professional student, geek at large, and I'm 6'4
    dave paton http://homepage.mac.com/dpaton dave at ieee dot org
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 03:32
    Joe:
    The TI datasheet states that external clamp diodes should be used with
    inductive loads. The equivalent circuit they show does not have clamp
    diodes. This is probably what caused your smoke. You need 4 diodes. At
    each motor terminal, connect two diodes: one to the bridge positive supply,
    the other to bridge gnd or negative. All diodes should "point" up toward
    plus supply.

    Ray McArthur

    Original Message
    From: Joseph Driscoll <driscoll85@h...>
    To: <basicstamps@egroups.com>
    Sent: Thursday, April 06, 2000 8:50 PM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > Hi,
    >
    > Thanks for the reply. Yes, please send me your program -- that would help
    a
    > lot. When you say "you must be sure that the inductance of the motor is
    high
    > enough so that the current continues to flow during the off time", I'm not
    > sure I understand that part. With the SN754410NE and the L293D, when the
    > enable pin is low, the outputs are at a high impedance state. Is that a
    > problem? It seems like the back emf produced during that "off" time needs
    to
    > go somewhere. Maybe this is the problem?
    >
    > Thanks,
    > Joe Driscoll
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 03:57
    On 4/6/00 at 8:36 PM, Joseph Driscoll <driscoll85@h...> wrote:

    > > 1. the motors were driven by large heatsinked H-bridges. I suggest getting
    > > a much more robust controller. 5A @ 12V would not be unreasonable.
    >
    > I've measured the current draw, and it's usually around 0.5 amp.

    Er..that's what I meant. Missed the '.'. Still, half an amp at 12V is 6W. I
    doubt most IC drivers could handle that. Besides, an h-bridge isn't very hard to
    build. Also don't forget the starting current will be a little higher. The
    motors are geared way down, but there is still a largish kick needed to start
    them, especially when they are loaded heavily. And then there's the inductive
    kick when you turn it off, or reverse it...

    Of course, I'm fond of overdesigning my controllers by a fcator of 2, so take
    this with a grain or two of salt.

    > My plan is to monitor the limit switches from the stamp, by running them to
    > input lines. Or, I could just have a hardware shutoff, using an OR gate or
    > something.

    Mmm...I'd go for the switches as inputs to the stamp. MUX them if required.
    Using the limit switches as hardware cutoffs will complicate things
    unnecesarially, IMHO, as you'll have to bypass them to restart the motor, and
    with reversing factored in....ugh.

    > > I have a vague recollection of encoders on the arm as well, but this might
    > > be a figment of my imagination. I do remembr clearly the ordeal that
    > > getting the arm to 'home' was. It seems the hero 2K SW was very picky.

    > Right, there are optical encoders on all of the motors. They output pulses,
    > which are squared up via a LM393. I plan to use a PAK VII pulse coprocessor
    > (from AWCE) to keep track of pulses. The stamp can then ask the PAK
    > periodically for pulse counts, and that way the position of the arm is known
    > (assuming it started from a known location).

    Cool. Sounds like a nice imple way to get it to go. It also saves the stamp from
    having to do the counting, and the associated overhead.

    > I plan to use the stamp as an arm "server": it polls sensors (and the PAK
    > VII) and sends that data to a PC. The PC then sends speed and direction
    > commands. The control algorithms can then be carried out on the PC.

    Sounds like fun. Keep us updated!

    -dave

    Passion must be present. Everything else can be learned.
    WCCR chief engineer, Professional student, geek at large, and I'm 6'4
    dave paton http://homepage.mac.com/dpaton dave at ieee dot org
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 04:08
    Joe:
    My last message applies to the L293. I just looked at the data for the
    SN754410NE, and it DOES show clamp diodes built into the IC. Also, they do
    not have the recommendation to use them. What I don't understand is that
    they still used external clamps in their example circuit.

    Ray McArthur

    Original Message
    From: rjmca <rjmca@w...>
    To: <basicstamps@egroups.com>
    Sent: Thursday, April 06, 2000 10:32 PM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > Joe:
    > The TI datasheet states that external clamp diodes should be used with
    > inductive loads. The equivalent circuit they show does not have clamp
    > diodes. This is probably what caused your smoke. You need 4 diodes. At
    > each motor terminal, connect two diodes: one to the bridge positive
    supply,
    > the other to bridge gnd or negative. All diodes should "point" up toward
    > plus supply.
    >
    > Ray McArthur
    >
    >
    Original Message
    > From: Joseph Driscoll <driscoll85@h...>
    > To: <basicstamps@egroups.com>
    > Sent: Thursday, April 06, 2000 8:50 PM
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > Hi,
    > >
    > > Thanks for the reply. Yes, please send me your program -- that would
    help
    > a
    > > lot. When you say "you must be sure that the inductance of the motor is
    > high
    > > enough so that the current continues to flow during the off time", I'm
    not
    > > sure I understand that part. With the SN754410NE and the L293D, when the
    > > enable pin is low, the outputs are at a high impedance state. Is that a
    > > problem? It seems like the back emf produced during that "off" time
    needs
    > to
    > > go somewhere. Maybe this is the problem?
    > >
    > > Thanks,
    > > Joe Driscoll
    >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 04:21
    I'm not an expert on motor drivers, but I think the L293 does NOT have
    diodes. the L293D does -- you have to make sure which one you have.

    Regards,

    Al Williams
    AWC
    * 8 Channels of PWM with the PAK-V at
    http://www.al-williams.com/awce/pak5.htm

    >
    > Joe:
    > My last message applies to the L293. I just looked at the data for the
    > SN754410NE, and it DOES show clamp diodes built into the IC.
    > Also, they do
    > not have the recommendation to use them. What I don't understand is that
    > they still used external clamps in their example circuit.
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 04:53
    Hi,

    I have the L293D and the SN754410NE, both of which have built-in diode
    protection at the outputs. The current draw from the motor I'm using is
    around 0.5A (which is close to max on the L293D, so that's why I use the
    SN754410NE). And still, poof: the chip blew up during PWM operation (not at
    startup). I have a schematic in GIF format if anyone is interested.

    Joe Driscoll


    Original Message
    From: Al Williams <alw@a...>
    To: <basicstamps@egroups.com>
    Sent: Thursday, April 06, 2000 10:21 PM
    Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control


    > I'm not an expert on motor drivers, but I think the L293 does NOT have
    > diodes. the L293D does -- you have to make sure which one you have.
    >
    > Regards,
    >
    > Al Williams
    > AWC
    > * 8 Channels of PWM with the PAK-V at
    > http://www.al-williams.com/awce/pak5.htm
    >
    > >
    > > Joe:
    > > My last message applies to the L293. I just looked at the data for the
    > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > > Also, they do
    > > not have the recommendation to use them. What I don't understand is
    that
    > > they still used external clamps in their example circuit.
    > >
    >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 05:58
    I revisited their site. You're right, Al. Thanks for the info.

    Ray McArthur

    Original Message
    From: Al Williams <alw@a...>
    To: <basicstamps@egroups.com>
    Sent: Thursday, April 06, 2000 11:21 PM
    Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control


    > I'm not an expert on motor drivers, but I think the L293 does NOT have
    > diodes. the L293D does -- you have to make sure which one you have.
    >
    > Regards,
    >
    > Al Williams
    > AWC
    > * 8 Channels of PWM with the PAK-V at
    > http://www.al-williams.com/awce/pak5.htm
    >
    > >
    > > Joe:
    > > My last message applies to the L293. I just looked at the data for the
    > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > > Also, they do
    > > not have the recommendation to use them. What I don't understand is
    that
    > > they still used external clamps in their example circuit.
    > >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 06:10
    Joe:

    You may send me a schematic off-list if you want. No guarantees, but I can
    try. Question: what is the PWM frequency?

    Ray McArthur

    Original Message
    From: Joseph Driscoll <driscoll85@h...>
    To: <basicstamps@egroups.com>
    Sent: Thursday, April 06, 2000 11:53 PM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > Hi,
    >
    > I have the L293D and the SN754410NE, both of which have built-in diode
    > protection at the outputs. The current draw from the motor I'm using is
    > around 0.5A (which is close to max on the L293D, so that's why I use the
    > SN754410NE). And still, poof: the chip blew up during PWM operation (not
    at
    > startup). I have a schematic in GIF format if anyone is interested.
    >
    > Joe Driscoll
    >
    >
    >
    Original Message
    > From: Al Williams <alw@a...>
    > To: <basicstamps@egroups.com>
    > Sent: Thursday, April 06, 2000 10:21 PM
    > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > I'm not an expert on motor drivers, but I think the L293 does NOT have
    > > diodes. the L293D does -- you have to make sure which one you have.
    > >
    > > Regards,
    > >
    > > Al Williams
    > > AWC
    > > * 8 Channels of PWM with the PAK-V at
    > > http://www.al-williams.com/awce/pak5.htm
    > >
    > > >
    > > > Joe:
    > > > My last message applies to the L293. I just looked at the data for
    the
    > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > > > Also, they do
    > > > not have the recommendation to use them. What I don't understand is
    > that
    > > > they still used external clamps in their example circuit.
    > > >
    > >
    > >
    > >
    > >
    > >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 13:56
    Oh, sorry Joe. I was replying to a reply and didn't realize you were the
    originator of the thread. I know we covered this off line already. Whoops.

    Regards,

    Al Williams
    AWC
    *Solderless Stamp Prototyping: http://www.al-williams.com/awce


    >
    Original Message
    > From: Joseph Driscoll [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=PxENTq7vd_PsjAG31F35qo7Sn1vR611Cpd0UG4npeQkQmtS3GzHQ0TrgeSGqjpZR2xAWbm24sAP1F7Du4K47Q88]driscoll85@h...[/url
    > Sent: Thursday, April 06, 2000 10:54 PM
    > To: basicstamps@egroups.com
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > Hi,
    >
    > I have the L293D and the SN754410NE, both of which have built-in diode
    > protection at the outputs. The current draw from the motor I'm using is
    > around 0.5A (which is close to max on the L293D, so that's why I use the
    > SN754410NE). And still, poof: the chip blew up during PWM
    > operation (not at
    > startup). I have a schematic in GIF format if anyone is interested.
    >
    > Joe Driscoll
    >
    >
    >
    Original Message
    > From: Al Williams <alw@a...>
    > To: <basicstamps@egroups.com>
    > Sent: Thursday, April 06, 2000 10:21 PM
    > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > I'm not an expert on motor drivers, but I think the L293 does NOT have
    > > diodes. the L293D does -- you have to make sure which one you have.
    > >
    > > Regards,
    > >
    > > Al Williams
    > > AWC
    > > * 8 Channels of PWM with the PAK-V at
    > > http://www.al-williams.com/awce/pak5.htm
    > >
    > > >
    > > > Joe:
    > > > My last message applies to the L293. I just looked at the
    > data for the
    > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > > > Also, they do
    > > > not have the recommendation to use them. What I don't understand is
    > that
    > > > they still used external clamps in their example circuit.
    > > >
    > >
    > >
    > >
    > >
    > >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 15:13
    At 09:22 AM 4/7/00 -0500, Joseph Driscoll promulgated:
    >Hi,
    >
    >The GIF is on its way to you off list. I have a question for everyone: what
    >happens if you switch the enable pin of the motor driver chip too fast?
    >Maybe the frequency of the PWM signal is too high? Would too fast a
    >switching time fry the driver chip?

    Hi Joe -

    Sure could !

    There is generally a frequency (aka repetition) maximum in the data sheet.

    Sorry - I don't have a data sheet handy.

    It should be under the AC characteristics - even though this is a DC device.

    Regards,

    Bruce

    >
    >Thanks,
    >Joe Driscoll
    >
    >
    >
    Original Message
    >From: rjmca <rjmca@w...>
    >To: <basicstamps@egroups.com>
    >Sent: Friday, April 07, 2000 12:10 AM
    >Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    >> Joe:
    >>
    >> You may send me a schematic off-list if you want. No guarantees, but I
    >can
    >> try. Question: what is the PWM frequency?
    >>
    >> Ray McArthur
    >>
    >>
    Original Message
    >> From: Joseph Driscoll <driscoll85@h...>
    >> To: <basicstamps@egroups.com>
    >> Sent: Thursday, April 06, 2000 11:53 PM
    >> Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >>
    >>
    >> > Hi,
    >> >
    >> > I have the L293D and the SN754410NE, both of which have built-in diode
    >> > protection at the outputs. The current draw from the motor I'm using is
    >> > around 0.5A (which is close to max on the L293D, so that's why I use the
    >> > SN754410NE). And still, poof: the chip blew up during PWM operation (not
    >> at
    >> > startup). I have a schematic in GIF format if anyone is interested.
    >> >
    >> > Joe Driscoll
    >> >
    >> >
    >> >
    Original Message
    >> > From: Al Williams <alw@a...>
    >> > To: <basicstamps@egroups.com>
    >> > Sent: Thursday, April 06, 2000 10:21 PM
    >> > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    >> >
    >> >
    >> > > I'm not an expert on motor drivers, but I think the L293 does NOT have
    >> > > diodes. the L293D does -- you have to make sure which one you have.
    >> > >
    >> > > Regards,
    >> > >
    >> > > Al Williams
    >> > > AWC
    >> > > * 8 Channels of PWM with the PAK-V at
    >> > > http://www.al-williams.com/awce/pak5.htm
    >> > >
    >> > > >
    >> > > > Joe:
    >> > > > My last message applies to the L293. I just looked at the data for
    >> the
    >> > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    >> > > > Also, they do
    >> > > > not have the recommendation to use them. What I don't understand is
    >> > that
    >> > > > they still used external clamps in their example circuit.
    >> > > >
    >> > >
    >> > >
    >> > >
    >> > >
    >> > >
    >> >
    >> >
    >> >
    >>
    >>
    >>
    >>
    >>
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 15:15
    At 09:22 AM 4/7/00 -0500, Joseph Driscoll promulgated:
    >Hi,
    >
    >The GIF is on its way to you off list. I have a question for everyone: what
    >happens if you switch the enable pin of the motor driver chip too fast?
    >Maybe the frequency of the PWM signal is too high? Would too fast a
    >switching time fry the driver chip?
    >
    >Thanks,
    >Joe Driscoll
    >
    Joe -

    Just one more comment -

    The PWM signal should be going to the POWER LEADS, *NOT* the ENABLE lines.
    At least that's what I've generally seen, YMMV depending on the particular
    chip.

    egards,

    Bruce

    >
    >
    Original Message
    >From: rjmca <rjmca@w...>
    >To: <basicstamps@egroups.com>
    >Sent: Friday, April 07, 2000 12:10 AM
    >Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    >> Joe:
    >>
    >> You may send me a schematic off-list if you want. No guarantees, but I
    >can
    >> try. Question: what is the PWM frequency?
    >>
    >> Ray McArthur
    >>
    >>
    Original Message
    >> From: Joseph Driscoll <driscoll85@h...>
    >> To: <basicstamps@egroups.com>
    >> Sent: Thursday, April 06, 2000 11:53 PM
    >> Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >>
    >>
    >> > Hi,
    >> >
    >> > I have the L293D and the SN754410NE, both of which have built-in diode
    >> > protection at the outputs. The current draw from the motor I'm using is
    >> > around 0.5A (which is close to max on the L293D, so that's why I use the
    >> > SN754410NE). And still, poof: the chip blew up during PWM operation (not
    >> at
    >> > startup). I have a schematic in GIF format if anyone is interested.
    >> >
    >> > Joe Driscoll
    >> >
    >> >
    >> >
    Original Message
    >> > From: Al Williams <alw@a...>
    >> > To: <basicstamps@egroups.com>
    >> > Sent: Thursday, April 06, 2000 10:21 PM
    >> > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    >> >
    >> >
    >> > > I'm not an expert on motor drivers, but I think the L293 does NOT have
    >> > > diodes. the L293D does -- you have to make sure which one you have.
    >> > >
    >> > > Regards,
    >> > >
    >> > > Al Williams
    >> > > AWC
    >> > > * 8 Channels of PWM with the PAK-V at
    >> > > http://www.al-williams.com/awce/pak5.htm
    >> > >
    >> > > >
    >> > > > Joe:
    >> > > > My last message applies to the L293. I just looked at the data for
    >> the
    >> > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    >> > > > Also, they do
    >> > > > not have the recommendation to use them. What I don't understand is
    >> > that
    >> > > > they still used external clamps in their example circuit.
    >> > > >
    >> > >
    >> > >
    >> > >
    >> > >
    >> > >
    >> >
    >> >
    >> >
    >>
    >>
    >>
    >>
    >>
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 15:22
    Hi,

    The GIF is on its way to you off list. I have a question for everyone: what
    happens if you switch the enable pin of the motor driver chip too fast?
    Maybe the frequency of the PWM signal is too high? Would too fast a
    switching time fry the driver chip?

    Thanks,
    Joe Driscoll


    Original Message
    From: rjmca <rjmca@w...>
    To: <basicstamps@egroups.com>
    Sent: Friday, April 07, 2000 12:10 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > Joe:
    >
    > You may send me a schematic off-list if you want. No guarantees, but I
    can
    > try. Question: what is the PWM frequency?
    >
    > Ray McArthur
    >
    >
    Original Message
    > From: Joseph Driscoll <driscoll85@h...>
    > To: <basicstamps@egroups.com>
    > Sent: Thursday, April 06, 2000 11:53 PM
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > Hi,
    > >
    > > I have the L293D and the SN754410NE, both of which have built-in diode
    > > protection at the outputs. The current draw from the motor I'm using is
    > > around 0.5A (which is close to max on the L293D, so that's why I use the
    > > SN754410NE). And still, poof: the chip blew up during PWM operation (not
    > at
    > > startup). I have a schematic in GIF format if anyone is interested.
    > >
    > > Joe Driscoll
    > >
    > >
    > >
    Original Message
    > > From: Al Williams <alw@a...>
    > > To: <basicstamps@egroups.com>
    > > Sent: Thursday, April 06, 2000 10:21 PM
    > > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > > > I'm not an expert on motor drivers, but I think the L293 does NOT have
    > > > diodes. the L293D does -- you have to make sure which one you have.
    > > >
    > > > Regards,
    > > >
    > > > Al Williams
    > > > AWC
    > > > * 8 Channels of PWM with the PAK-V at
    > > > http://www.al-williams.com/awce/pak5.htm
    > > >
    > > > >
    > > > > Joe:
    > > > > My last message applies to the L293. I just looked at the data for
    > the
    > > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > > > > Also, they do
    > > > > not have the recommendation to use them. What I don't understand is
    > > that
    > > > > they still used external clamps in their example circuit.
    > > > >
    > > >
    > > >
    > > >
    > > >
    > > >
    > >
    > >
    > >
    >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 15:44
    Hi,

    I'm looking at the SN754410NE datasheet, and they have some timing diagrams
    using a "test circuit" that had a pulse high-time of 10 uS. But I can't tell
    if that is a minimum by looking at the page. On another page, where min,
    max, and typical values are given, the rise time, fall time, and pulse
    duration values are blank. I'm using the PAK Va from AWC for PWM. It has two
    modes of operation, equal area mode and proportional PWM mode. Its data
    sheet says that when using equal area mode, the time period is about 512 uS,
    and proportional PWM mode uses a 2 uS time base. Sound like equal mode is
    slower, and therefore safer in my application? What driver chips have others
    used with the PAK Va? How were things connected?

    Thanks,
    Joe Driscoll


    Original Message
    From: Bruce Bates <bvbates@u...>
    To: <basicstamps@egroups.com>
    Sent: Friday, April 07, 2000 9:13 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > At 09:22 AM 4/7/00 -0500, Joseph Driscoll promulgated:
    > >Hi,
    > >
    > >The GIF is on its way to you off list. I have a question for everyone:
    what
    > >happens if you switch the enable pin of the motor driver chip too fast?
    > >Maybe the frequency of the PWM signal is too high? Would too fast a
    > >switching time fry the driver chip?
    >
    > Hi Joe -
    >
    > Sure could !
    >
    > There is generally a frequency (aka repetition) maximum in the data sheet.
    >
    > Sorry - I don't have a data sheet handy.
    >
    > It should be under the AC characteristics - even though this is a DC
    device.
    >
    > Regards,
    >
    > Bruce
    >
    > >
    > >Thanks,
    > >Joe Driscoll
    > >
    > >
    > >
    Original Message
    > >From: rjmca <rjmca@w...>
    > >To: <basicstamps@egroups.com>
    > >Sent: Friday, April 07, 2000 12:10 AM
    > >Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > >> Joe:
    > >>
    > >> You may send me a schematic off-list if you want. No guarantees, but I
    > >can
    > >> try. Question: what is the PWM frequency?
    > >>
    > >> Ray McArthur
    > >>
    > >>
    Original Message
    > >> From: Joseph Driscoll <driscoll85@h...>
    > >> To: <basicstamps@egroups.com>
    > >> Sent: Thursday, April 06, 2000 11:53 PM
    > >> Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >>
    > >>
    > >> > Hi,
    > >> >
    > >> > I have the L293D and the SN754410NE, both of which have built-in
    diode
    > >> > protection at the outputs. The current draw from the motor I'm using
    is
    > >> > around 0.5A (which is close to max on the L293D, so that's why I use
    the
    > >> > SN754410NE). And still, poof: the chip blew up during PWM operation
    (not
    > >> at
    > >> > startup). I have a schematic in GIF format if anyone is interested.
    > >> >
    > >> > Joe Driscoll
    > >> >
    > >> >
    > >> >
    Original Message
    > >> > From: Al Williams <alw@a...>
    > >> > To: <basicstamps@egroups.com>
    > >> > Sent: Thursday, April 06, 2000 10:21 PM
    > >> > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >> >
    > >> >
    > >> > > I'm not an expert on motor drivers, but I think the L293 does NOT
    have
    > >> > > diodes. the L293D does -- you have to make sure which one you have.
    > >> > >
    > >> > > Regards,
    > >> > >
    > >> > > Al Williams
    > >> > > AWC
    > >> > > * 8 Channels of PWM with the PAK-V at
    > >> > > http://www.al-williams.com/awce/pak5.htm
    > >> > >
    > >> > > >
    > >> > > > Joe:
    > >> > > > My last message applies to the L293. I just looked at the data
    for
    > >> the
    > >> > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > >> > > > Also, they do
    > >> > > > not have the recommendation to use them. What I don't understand
    is
    > >> > that
    > >> > > > they still used external clamps in their example circuit.
    > >> > > >
    > >> > >
    > >> > >
    > >> > >
    > >> > >
    > >> > >
    > >> >
    > >> >
    > >> >
    > >>
    > >>
    > >>
    > >>
    > >>
    > >
    > >
    > >
    > >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 15:57
    You mean the logic Vcc of the driver chip should be switched? Or do you mean
    the driver channel inputs? I've heard of the latter, but I've seen many
    cases of pulsing the enable lines (but maybe they were using a slower PWM
    frequency). Why is pulsing the enable lines bad? Is it because when the
    enable is low, the outputs are at high-Z?

    Joe


    Original Message
    From: Bruce Bates <bvbates@u...>
    To: <basicstamps@egroups.com>
    Sent: Friday, April 07, 2000 9:15 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control

    > Joe -
    >
    > Just one more comment -
    >
    > The PWM signal should be going to the POWER LEADS, *NOT* the ENABLE lines.
    > At least that's what I've generally seen, YMMV depending on the particular
    > chip.
    >
    > egards,
    >
    > Bruce
    >
    > >
    > >
    Original Message
    > >From: rjmca <rjmca@w...>
    > >To: <basicstamps@egroups.com>
    > >Sent: Friday, April 07, 2000 12:10 AM
    > >Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > >> Joe:
    > >>
    > >> You may send me a schematic off-list if you want. No guarantees, but I
    > >can
    > >> try. Question: what is the PWM frequency?
    > >>
    > >> Ray McArthur
    > >>
    > >>
    Original Message
    > >> From: Joseph Driscoll <driscoll85@h...>
    > >> To: <basicstamps@egroups.com>
    > >> Sent: Thursday, April 06, 2000 11:53 PM
    > >> Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >>
    > >>
    > >> > Hi,
    > >> >
    > >> > I have the L293D and the SN754410NE, both of which have built-in
    diode
    > >> > protection at the outputs. The current draw from the motor I'm using
    is
    > >> > around 0.5A (which is close to max on the L293D, so that's why I use
    the
    > >> > SN754410NE). And still, poof: the chip blew up during PWM operation
    (not
    > >> at
    > >> > startup). I have a schematic in GIF format if anyone is interested.
    > >> >
    > >> > Joe Driscoll
    > >> >
    > >> >
    > >> >
    Original Message
    > >> > From: Al Williams <alw@a...>
    > >> > To: <basicstamps@egroups.com>
    > >> > Sent: Thursday, April 06, 2000 10:21 PM
    > >> > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >> >
    > >> >
    > >> > > I'm not an expert on motor drivers, but I think the L293 does NOT
    have
    > >> > > diodes. the L293D does -- you have to make sure which one you have.
    > >> > >
    > >> > > Regards,
    > >> > >
    > >> > > Al Williams
    > >> > > AWC
    > >> > > * 8 Channels of PWM with the PAK-V at
    > >> > > http://www.al-williams.com/awce/pak5.htm
    > >> > >
    > >> > > >
    > >> > > > Joe:
    > >> > > > My last message applies to the L293. I just looked at the data
    for
    > >> the
    > >> > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > >> > > > Also, they do
    > >> > > > not have the recommendation to use them. What I don't understand
    is
    > >> > that
    > >> > > > they still used external clamps in their example circuit.
    > >> > > >
    > >> > >
    > >> > >
    > >> > >
    > >> > >
    > >> > >
    > >> >
    > >> >
    > >> >
    > >>
    > >>
    > >>
    > >>
    > >>
    > >
    > >
    > >
    > >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 16:33
    Just to clarify:

    1) The PAK-V and the PAK-Va are identical except the old PAK-V required a
    shunt resistor across its timing element. The PAK-Va does not. There are no
    more PAK-Vs, just the PAK-Va, but we have several volume customers who would
    be upset if we just changed how the chip works and didn't give it a new
    name. However, informally, we just call them PAK-V.

    2) In either mode, the time base is 2uS. Here's how it works.

    Proportional mode (like the Stamp) figures the least amount of work
    required. At 50%, for example, there is a 2uS on time and a 2uS off time.
    For 33% there is a 2uS on time and 4uS off time. The proportion of the
    pulses is correct, but the total frequency varies depending on the duty
    cycle. So 50% is 250kHz and 33% is 166.7kHz. This mode charges a capacitor
    as fast as possible for a given duty cycle.

    Equal Area mode always uses 256 time periods (512uS or 1.95kHz). 50% is
    256uS on and 256uS off. 1/256 is 2uS on and 510uS off. This mode is often
    used in controlling LEDs. Makes for a nice scope display too.

    The mode is set for the whole chip. So all 8 PWM outputs are in one mode or
    the other.

    I know from your previous e-mails that the Stamp's PWM is not driving your
    motors any better than the PAK-V. I wonder if you try driving the supply pin
    if that will make a difference?

    Regards,

    Al Williams
    AWC
    * Expand your Stamp I/O with a PAK-III or IV:
    http://www.al-williams.com/awce/pak3.htm




    >
    Original Message
    > From: Joseph Driscoll [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=EPcTaQR2fLlHk8jpF-e73hTe-nQ9RaY-e4XCVcppGi8M_fq-LLZGhyyl2WUGbbB_bCnVMA31mmpxxcjX]driscoll85@h...[/url
    > Sent: Friday, April 07, 2000 9:45 AM
    > To: basicstamps@egroups.com
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > Hi,
    >
    > I'm looking at the SN754410NE datasheet, and they have some
    > timing diagrams
    > using a "test circuit" that had a pulse high-time of 10 uS. But I
    > can't tell
    > if that is a minimum by looking at the page. On another page, where min,
    > max, and typical values are given, the rise time, fall time, and pulse
    > duration values are blank. I'm using the PAK Va from AWC for PWM.
    > It has two
    > modes of operation, equal area mode and proportional PWM mode. Its data
    > sheet says that when using equal area mode, the time period is
    > about 512 uS,
    > and proportional PWM mode uses a 2 uS time base. Sound like equal mode is
    > slower, and therefore safer in my application? What driver chips
    > have others
    > used with the PAK Va? How were things connected?
    >
    > Thanks,
    > Joe Driscoll
    >
    >
    >
    Original Message
    > From: Bruce Bates <bvbates@u...>
    > To: <basicstamps@egroups.com>
    > Sent: Friday, April 07, 2000 9:13 AM
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > At 09:22 AM 4/7/00 -0500, Joseph Driscoll promulgated:
    > > >Hi,
    > > >
    > > >The GIF is on its way to you off list. I have a question for everyone:
    > what
    > > >happens if you switch the enable pin of the motor driver chip too fast?
    > > >Maybe the frequency of the PWM signal is too high? Would too fast a
    > > >switching time fry the driver chip?
    > >
    > > Hi Joe -
    > >
    > > Sure could !
    > >
    > > There is generally a frequency (aka repetition) maximum in the
    > data sheet.
    > >
    > > Sorry - I don't have a data sheet handy.
    > >
    > > It should be under the AC characteristics - even though this is a DC
    > device.
    > >
    > > Regards,
    > >
    > > Bruce
    > >
    > > >
    > > >Thanks,
    > > >Joe Driscoll
    > > >
    > > >
    > > >
    Original Message
    > > >From: rjmca <rjmca@w...>
    > > >To: <basicstamps@egroups.com>
    > > >Sent: Friday, April 07, 2000 12:10 AM
    > > >Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > >
    > > >
    > > >> Joe:
    > > >>
    > > >> You may send me a schematic off-list if you want. No
    > guarantees, but I
    > > >can
    > > >> try. Question: what is the PWM frequency?
    > > >>
    > > >> Ray McArthur
    > > >>
    > > >>
    Original Message
    > > >> From: Joseph Driscoll <driscoll85@h...>
    > > >> To: <basicstamps@egroups.com>
    > > >> Sent: Thursday, April 06, 2000 11:53 PM
    > > >> Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > >>
    > > >>
    > > >> > Hi,
    > > >> >
    > > >> > I have the L293D and the SN754410NE, both of which have built-in
    > diode
    > > >> > protection at the outputs. The current draw from the motor
    > I'm using
    > is
    > > >> > around 0.5A (which is close to max on the L293D, so that's
    > why I use
    > the
    > > >> > SN754410NE). And still, poof: the chip blew up during PWM operation
    > (not
    > > >> at
    > > >> > startup). I have a schematic in GIF format if anyone is interested.
    > > >> >
    > > >> > Joe Driscoll
    > > >> >
    > > >> >
    > > >> >
    Original Message
    > > >> > From: Al Williams <alw@a...>
    > > >> > To: <basicstamps@egroups.com>
    > > >> > Sent: Thursday, April 06, 2000 10:21 PM
    > > >> > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > >> >
    > > >> >
    > > >> > > I'm not an expert on motor drivers, but I think the L293 does NOT
    > have
    > > >> > > diodes. the L293D does -- you have to make sure which
    > one you have.
    > > >> > >
    > > >> > > Regards,
    > > >> > >
    > > >> > > Al Williams
    > > >> > > AWC
    > > >> > > * 8 Channels of PWM with the PAK-V at
    > > >> > > http://www.al-williams.com/awce/pak5.htm
    > > >> > >
    > > >> > > >
    > > >> > > > Joe:
    > > >> > > > My last message applies to the L293. I just looked at the data
    > for
    > > >> the
    > > >> > > > SN754410NE, and it DOES show clamp diodes built into the IC.
    > > >> > > > Also, they do
    > > >> > > > not have the recommendation to use them. What I don't
    > understand
    > is
    > > >> > that
    > > >> > > > they still used external clamps in their example circuit.
    > > >> > > >
    > > >> > >
    > > >> > >
    > > >> > >
    > > >> > >
    > > >> > >
    > > >> >
    > > >> >
    > > >> >
    > > >>
    > > >>
    > > >>
    > > >>
    > > >>
    > > >
    > > >
    > > >
    > > >
    > >
    > >
    > >
    > >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 18:42
    I'll send it off line.. the diodes 4 need to go directly from the motor
    to the supply pos and neg. all 4 diodes go arrow up toward the rail. They are
    not ON except during back emf.
    They need to be rated at 2 x the rail max volts and at least equal to max amps.


    To: <basicstamps@egroups.com>
    Date sent: Thu, 6 Apr 2000 19:50:42 -0500
    From: "Joseph Driscoll" <driscoll85@h...>
    Send reply to: basicstamps@egroups.com
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control

    > Hi,
    >
    > Thanks for the reply. Yes, please send me your program -- that would help a
    > lot. When you say "you must be sure that the inductance of the motor is high
    > enough so that the current continues to flow during the off time", I'm not
    > sure I understand that part. With the SN754410NE and the L293D, when the
    > enable pin is low, the outputs are at a high impedance state. Is that a
    > problem? It seems like the back emf produced during that "off" time needs to
    > go somewhere. Maybe this is the problem?
    >
    > Thanks,
    > Joe Driscoll
    >
    >
    >
    Original Message
    > From: Ronald Doctors <rdoctors@h...>
    > To: <basicstamps@egroups.com>
    > Sent: Thursday, April 06, 2000 6:32 PM
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > HI,
    > >
    > > Don't know about the SN754410NE chip but the MC33030 chip is very easy to
    > use
    > > with a STAMP2 . I am using them with an x-Y control system and it is
    > fairly simple.
    > > I found that the STAMP needs expansion for both analog signal and digital
    > data handling.
    > > If you want I can email my prgm ( not yet fully debugged ) and it will
    > give you some idea of the complexity.
    > > The noises are the PWM causing the windings to vibrate. If you are using
    > a PWM to control speed
    > > you must be sure that the inductance of the motor is high enough so that
    > the current continues to flow during the off
    > > time. Try the chip in a simple config before using it in a STAMP control
    > system to make sure you have that part working smoothly.
    > >
    > > Also you need diodes to make sure that the stored energy in the inductance
    > has somewhere to go.. In the case of full H bridge that means 4 diodes
    > > for each motor. Usually they are built in to the control chip.
    > >
    > > ( PLEASE NO MORE DIODE RELAY STUFF!!! [noparse]:)[/noparse])
    > > To: <basicstamps@egroups.com>
    > > Date sent: Thu, 6 Apr 2000 17:51:40 -0500
    > > From: "Joseph Driscoll" <driscoll85@h...>
    > > Send reply to: basicstamps@egroups.com
    > > Subject: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > > > Hello,
    > > >
    > > > I have an old Hero 2000 robot arm, which was sold by Heathkit. The same
    > arm
    > > > was sold separately from the robot as an "arm trainer". Not counting the
    > > > gripper, the arm is moved by four identical, 12V gearmotors. One for
    > > > shoulder, one for elbow, and two for wrist (the wrist is a differential
    > > > mechanism).
    > > >
    > > > I have been trying to create an interface so that I can control the arm
    > with
    > > > my PC. I am using the Basic Stamp II for PC interfacing. To drive the
    > > > motors, I am using SN754410NE motor driver chips. For PWM speed control,
    > I'm
    > > > using the PAK Va from AWC.
    > > >
    > > > Here's the problem: I'm a bit new at motors, and I'm having some
    > troubles.
    > > > At low PWM duty cycles, the motor makes a high pitched sound and won't
    > move.
    > > > At slightly higher speeds, the tone is still there but the motor starts
    > to
    > > > move. What is this tone?
    > > >
    > > > Also, I was under the impression that the SN754410NE had built-in diodes
    > for
    > > > backEMF protection. One of the chips exploded (sparks, little divot out
    > of
    > > > the top of the IC) today, leading me to believe that something was
    > wrong.
    > > >
    > > > Does anyone have experience working with the Hero 2000 arm? What I'd
    > really
    > > > love is a circuit to build that lets the PC or Stamp (or other
    > > > microcontroller) drive the arm's motors. I can handle control
    > algorithms,
    > > > etc. I'm just having trouble with the hardware interfacing. I'd like to
    > see
    > > > someone's
    > > > circuit to see where I'm going wrong.
    > > >
    > > > Thanks,
    > > > Joe Driscoll
    > > >
    > > >
    > > >
    > > >
    > >
    > >
    > > ron... the motor man...
    > > Ronald Doctors
    > > http://www.members.home.net/rdoctors
    > >
    > >
    > >
    > >
    >
    >
    >


    ron... the motor man...
    Ronald Doctors
    http://www.members.home.net/rdoctors
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-07 22:29
    Joe:
    I have to study your circuit & info, but a general answer about frequency is
    yes, it can make smoke if it's too high. PWM is used to minimize power
    dissipation in the control transistors, because the transistors are switched
    "on" or "off". Power dissipated is Vce x Ic. When the transistor is "on",
    Vce is low, so power is small. When the switch is "off", power is low
    because current is zero. However, during switching "on" to "off" or vice
    versa, power dissipation can be quite high (especially with an inductive
    load). If switching frequency is low enough, these intervals of high power
    dissipation occur a small percentage of total time, so the average power is
    small. Another problem is that diodes have a "recovery time", which means
    that they momentarily are like a "short" when going from conduction to
    non-conduction. This causes current spikes in the switching transistors
    which becomes more important as frequency rises.

    The question of what is "too high" a frequency depends on the speed of the
    transistors & diodes used. Off hand I don't know. Maybe we will have an
    idea after studying your stuff. I wonder if Ron Doctors has experience with
    this?

    Ray McArthur.

    Original Message
    From: Joseph Driscoll <driscoll85@h...>
    To: <basicstamps@egroups.com>
    Sent: Friday, April 07, 2000 10:22 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > Hi,
    >
    > The GIF is on its way to you off list. I have a question for everyone:
    what
    > happens if you switch the enable pin of the motor driver chip too fast?
    > Maybe the frequency of the PWM signal is too high? Would too fast a
    > switching time fry the driver chip?
    >
    > Thanks,
    > Joe Driscoll
    >
    >
    >
    Original Message
    > From: rjmca <rjmca@w...>
    > To: <basicstamps@egroups.com>
    > Sent: Friday, April 07, 2000 12:10 AM
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > Joe:
    > >
    > > You may send me a schematic off-list if you want. No guarantees, but I
    > can
    > > try. Question: what is the PWM frequency?
    > >
    > > Ray McArthur
    > >
    > >
    Original Message
    > > From: Joseph Driscoll <driscoll85@h...>
    > > To: <basicstamps@egroups.com>
    > > Sent: Thursday, April 06, 2000 11:53 PM
    > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > > > Hi,
    > > >
    > > > I have the L293D and the SN754410NE, both of which have built-in diode
    > > > protection at the outputs. The current draw from the motor I'm using
    is
    > > > around 0.5A (which is close to max on the L293D, so that's why I use
    the
    > > > SN754410NE). And still, poof: the chip blew up during PWM operation
    (not
    > > at
    > > > startup). I have a schematic in GIF format if anyone is interested.
    > > >
    > > > Joe Driscoll
    > > >
    > > >
    > > >
    Original Message
    > > > From: Al Williams <alw@a...>
    > > > To: <basicstamps@egroups.com>
    > > > Sent: Thursday, April 06, 2000 10:21 PM
    > > > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > >
    > > >
    > > > > I'm not an expert on motor drivers, but I think the L293 does NOT
    have
    > > > > diodes. the L293D does -- you have to make sure which one you have.
    > > > >
    > > > > Regards,
    > > > >
    > > > > Al Williams
    > > > > AWC
    > > > > * 8 Channels of PWM with the PAK-V at
    > > > > http://www.al-williams.com/awce/pak5.htm
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-09 02:25
    Joe:

    The data for the SN754410 calls out 400-900ns enable/disable time. This
    means that if you are applying PWM to the enable line in proportional mode,
    the frequency is way too high. Al Williams states that this mode outputs in
    the range of 200kHz. But Al also points out that equal area mode is ~ 2kHz,
    which is much more reasonable for your chips. Could you operate in this
    lower frequency mode?

    Ray McArthur

    Original Message
    From: Joseph Driscoll <driscoll85@h...>
    To: <basicstamps@egroups.com>
    Sent: Friday, April 07, 2000 10:22 AM
    Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control


    > Hi,
    >
    > The GIF is on its way to you off list. I have a question for everyone:
    what
    > happens if you switch the enable pin of the motor driver chip too fast?
    > Maybe the frequency of the PWM signal is too high? Would too fast a
    > switching time fry the driver chip?
    >
    > Thanks,
    > Joe Driscoll
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-09 03:55
    Well, 900nS on and 900nS off is 1.8uS. The PAK works at 2uS per half
    cycle -- 4uS per cycle or 250kHz. So I think that is OK.

    Regards,

    Al Williams
    AWC
    * Microcontroller Projects with Basic Stamps:
    www.al-williams.com/awce/sbook.htm


    >
    Original Message
    > From: rjmca [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=ADECN2VpSCzZk-QMXGKncAgrpvCwYEuZifuRdFWK8nPb9MpcMsSTpmWzbKt5VOiKjmHmt-7ImpSf]rjmca@w...[/url
    > Sent: Saturday, April 08, 2000 8:26 PM
    > To: basicstamps@egroups.com
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > Joe:
    >
    > The data for the SN754410 calls out 400-900ns enable/disable time. This
    > means that if you are applying PWM to the enable line in
    > proportional mode,
    > the frequency is way too high. Al Williams states that this mode
    > outputs in
    > the range of 200kHz. But Al also points out that equal area mode
    > is ~ 2kHz,
    > which is much more reasonable for your chips. Could you operate in this
    > lower frequency mode?
    >
    > Ray McArthur
    >
    >
    Original Message
    > From: Joseph Driscoll <driscoll85@h...>
    > To: <basicstamps@egroups.com>
    > Sent: Friday, April 07, 2000 10:22 AM
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > Hi,
    > >
    > > The GIF is on its way to you off list. I have a question for everyone:
    > what
    > > happens if you switch the enable pin of the motor driver chip too fast?
    > > Maybe the frequency of the PWM signal is too high? Would too fast a
    > > switching time fry the driver chip?
    > >
    > > Thanks,
    > > Joe Driscoll
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-09 06:08
    Al:
    I am afraid that the switching losses will be very high when the time spent
    transitioning from hi to lo and back is an appreciable part of the total
    duty cycle. Is there a problem running at the lower frequency? If you have
    a chance, read my message of 4/7 at 5:30PM and see if you agree with it.

    Thanks,
    Ray McArthur

    Original Message
    From: Al Williams <alw@a...>
    To: <basicstamps@egroups.com>
    Sent: Saturday, April 08, 2000 10:55 PM
    Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control


    > Well, 900nS on and 900nS off is 1.8uS. The PAK works at 2uS per half
    > cycle -- 4uS per cycle or 250kHz. So I think that is OK.
    >
    > Regards,
    >
    > Al Williams
    > AWC
    > * Microcontroller Projects with Basic Stamps:
    > www.al-williams.com/awce/sbook.htm
    >
    >
    > >
    Original Message
    > > From: rjmca [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=T_TNx--ywac1zdd1INVKdayt8-CMaYxVQ9wAfGzeEZkxiGt1xknCDVHH5Cq3LnQAI0Rm1n0rr3Fw]rjmca@w...[/url
    > > Sent: Saturday, April 08, 2000 8:26 PM
    > > To: basicstamps@egroups.com
    > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > > Joe:
    > >
    > > The data for the SN754410 calls out 400-900ns enable/disable time. This
    > > means that if you are applying PWM to the enable line in
    > > proportional mode,
    > > the frequency is way too high. Al Williams states that this mode
    > > outputs in
    > > the range of 200kHz. But Al also points out that equal area mode
    > > is ~ 2kHz,
    > > which is much more reasonable for your chips. Could you operate in this
    > > lower frequency mode?
    > >
    > > Ray McArthur
    > >
    > >
    Original Message
    > > From: Joseph Driscoll <driscoll85@h...>
    > > To: <basicstamps@egroups.com>
    > > Sent: Friday, April 07, 2000 10:22 AM
    > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > > > Hi,
    > > >
    > > > The GIF is on its way to you off list. I have a question for everyone:
    > > what
    > > > happens if you switch the enable pin of the motor driver chip too
    fast?
    > > > Maybe the frequency of the PWM signal is too high? Would too fast a
    > > > switching time fry the driver chip?
    > > >
    > > > Thanks,
    > > > Joe Driscoll
    > >
    > >
    > >
    > >
    >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-09 06:37
    That's true too. Actually, you can scale back the clock element and reduce
    your baud rate by the same factor. This also reduces current consumption to
    the chip. So a 25Mhz resonator (or crystal) will cut your baud rate in half
    and set the time base to 4uS.

    Regards,

    Al Williams
    AWC
    * Solderless Stamp prototyping at http://www.al-williams.com/awce


    >
    Original Message
    > From: rjmca [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=E1MvJwF7-3meP_okLo2fuy27sy2c2e65DgLVAadGLbKm5hjcH3B54S0TSTi2O8qWcVSkBAqhXHC2FA]rjmca@w...[/url
    > Sent: Sunday, April 09, 2000 12:09 AM
    > To: basicstamps@egroups.com
    > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > Al:
    > I am afraid that the switching losses will be very high when the
    > time spent
    > transitioning from hi to lo and back is an appreciable part of the total
    > duty cycle. Is there a problem running at the lower frequency?
    > If you have
    > a chance, read my message of 4/7 at 5:30PM and see if you agree with it.
    >
    > Thanks,
    > Ray McArthur
    >
    >
    Original Message
    > From: Al Williams <alw@a...>
    > To: <basicstamps@egroups.com>
    > Sent: Saturday, April 08, 2000 10:55 PM
    > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    >
    >
    > > Well, 900nS on and 900nS off is 1.8uS. The PAK works at 2uS per half
    > > cycle -- 4uS per cycle or 250kHz. So I think that is OK.
    > >
    > > Regards,
    > >
    > > Al Williams
    > > AWC
    > > * Microcontroller Projects with Basic Stamps:
    > > www.al-williams.com/awce/sbook.htm
    > >
    > >
    > > >
    Original Message
    > > > From: rjmca [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=E1MvJwF7-3meP_okLo2fuy27sy2c2e65DgLVAadGLbKm5hjcH3B54S0TSTi2O8qWcVSkBAqhXHC2FA]rjmca@w...[/url
    > > > Sent: Saturday, April 08, 2000 8:26 PM
    > > > To: basicstamps@egroups.com
    > > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > >
    > > >
    > > > Joe:
    > > >
    > > > The data for the SN754410 calls out 400-900ns enable/disable
    > time. This
    > > > means that if you are applying PWM to the enable line in
    > > > proportional mode,
    > > > the frequency is way too high. Al Williams states that this mode
    > > > outputs in
    > > > the range of 200kHz. But Al also points out that equal area mode
    > > > is ~ 2kHz,
    > > > which is much more reasonable for your chips. Could you
    > operate in this
    > > > lower frequency mode?
    > > >
    > > > Ray McArthur
    > > >
    > > >
    Original Message
    > > > From: Joseph Driscoll <driscoll85@h...>
    > > > To: <basicstamps@egroups.com>
    > > > Sent: Friday, April 07, 2000 10:22 AM
    > > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > >
    > > >
    > > > > Hi,
    > > > >
    > > > > The GIF is on its way to you off list. I have a question
    > for everyone:
    > > > what
    > > > > happens if you switch the enable pin of the motor driver chip too
    > fast?
    > > > > Maybe the frequency of the PWM signal is too high? Would too fast a
    > > > > switching time fry the driver chip?
    > > > >
    > > > > Thanks,
    > > > > Joe Driscoll
    > > >
    > > >
    > > >
    > > >
    > >
    > >
    > >
    > >
    >
    >
    >
  • ArchiverArchiver Posts: 46,084
    edited 2000-04-10 16:36
    Hi,

    I'll try things in equal area mode and see what happens. I'll be away from
    the arm for a week or so, but I'll let you know when I've tried things.

    Thanks,
    Joe Driscoll


    >From: "Al Williams" <alw@a...>
    >Reply-To: basicstamps@egroups.com
    >To: <basicstamps@egroups.com>
    >Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    >Date: Sun, 9 Apr 2000 00:37:47 -0500
    >
    >That's true too. Actually, you can scale back the clock element and reduce
    >your baud rate by the same factor. This also reduces current consumption to
    >the chip. So a 25Mhz resonator (or crystal) will cut your baud rate in half
    >and set the time base to 4uS.
    >
    >Regards,
    >
    >Al Williams
    >AWC
    >* Solderless Stamp prototyping at http://www.al-williams.com/awce
    >
    >
    > >
    Original Message
    > > From: rjmca [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=s5GumWWCPH2R8A8RWbxD1GJF2ze-xvrDOa1_YXdYNF9QA0ihwHXTE8x5LWa-N02mmuq5JQDi7hwJ2l3w]rjmca@w...[/url
    > > Sent: Sunday, April 09, 2000 12:09 AM
    > > To: basicstamps@egroups.com
    > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > > Al:
    > > I am afraid that the switching losses will be very high when the
    > > time spent
    > > transitioning from hi to lo and back is an appreciable part of the total
    > > duty cycle. Is there a problem running at the lower frequency?
    > > If you have
    > > a chance, read my message of 4/7 at 5:30PM and see if you agree with
    >it.
    > >
    > > Thanks,
    > > Ray McArthur
    > >
    > >
    Original Message
    > > From: Al Williams <alw@a...>
    > > To: <basicstamps@egroups.com>
    > > Sent: Saturday, April 08, 2000 10:55 PM
    > > Subject: RE: [noparse][[/noparse]basicstamps] Stamps and motor control
    > >
    > >
    > > > Well, 900nS on and 900nS off is 1.8uS. The PAK works at 2uS per half
    > > > cycle -- 4uS per cycle or 250kHz. So I think that is OK.
    > > >
    > > > Regards,
    > > >
    > > > Al Williams
    > > > AWC
    > > > * Microcontroller Projects with Basic Stamps:
    > > > www.al-williams.com/awce/sbook.htm
    > > >
    > > >
    > > > >
    Original Message
    > > > > From: rjmca [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=s5GumWWCPH2R8A8RWbxD1GJF2ze-xvrDOa1_YXdYNF9QA0ihwHXTE8x5LWa-N02mmuq5JQDi7hwJ2l3w]rjmca@w...[/url
    > > > > Sent: Saturday, April 08, 2000 8:26 PM
    > > > > To: basicstamps@egroups.com
    > > > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > > >
    > > > >
    > > > > Joe:
    > > > >
    > > > > The data for the SN754410 calls out 400-900ns enable/disable
    > > time. This
    > > > > means that if you are applying PWM to the enable line in
    > > > > proportional mode,
    > > > > the frequency is way too high. Al Williams states that this mode
    > > > > outputs in
    > > > > the range of 200kHz. But Al also points out that equal area mode
    > > > > is ~ 2kHz,
    > > > > which is much more reasonable for your chips. Could you
    > > operate in this
    > > > > lower frequency mode?
    > > > >
    > > > > Ray McArthur
    > > > >
    > > > >
    Original Message
    > > > > From: Joseph Driscoll <driscoll85@h...>
    > > > > To: <basicstamps@egroups.com>
    > > > > Sent: Friday, April 07, 2000 10:22 AM
    > > > > Subject: Re: [noparse][[/noparse]basicstamps] Stamps and motor control
    > > > >
    > > > >
    > > > > > Hi,
    > > > > >
    > > > > > The GIF is on its way to you off list. I have a question
    > > for everyone:
    > > > > what
    > > > > > happens if you switch the enable pin of the motor driver chip too
    > > fast?
    > > > > > Maybe the frequency of the PWM signal is too high? Would too fast
    >a
    > > > > > switching time fry the driver chip?
    > > > > >
    > > > > > Thanks,
    > > > > > Joe Driscoll
    > > > >
    > > > >
    > > > >
    > > > >
    > > >
    > > >
    > > >
    > > >
    > >
    > >
    > >
    >
    >
    >
    >

    ______________________________________________________
    Get Your Private, Free Email at http://www.hotmail.com
Sign In or Register to comment.