e-stop buttons and circuits for big robots?

[jim The STEAM Clown - Jim Burnham]

Emergency Stop buttons and circuits?

I'm in the process of having students work on 2 different electric wheel chair based robots. These are both 12 volt golf cart battery system.  I'm in the process of ordering some motor controllers and want to make sure I have a safe and effective Emergency stop.

How have you implemented this on some of your bigger Robots?

I was thinking of having the students research the following, but I'm having a bit of a hard time finding good resources my self. 

1) E-Stop button that has a e-stop signal (pull up/down) direct to a motor controller (if we get on that has this input) 

2) E-Stop button that is connected to a small micro (ARDUINO/esp32) that controlled the motor control signals and if pressed will basically send/change any motor control signals to a "stop all motors" command.

3) similar to #2 but basically have a heartbeat /time out signal that if it ever is missed, the robot controller executes a "stop all motors" 

Do any of you interrupt the battery power to the motors with some type of power relay?

Any circuit diagrams and advice would be appreciated.

-- Jim

[David Hough]

A proper, safe last-resort e-stop should take out the power to the
motors with a big mechanical switch or relay of some sort. Anything that
relies on software is just first-line protection. You really need a big
button somewhere that a person can press that will directly disable the
motors regardless of whether software or wifi or anything else is working.

Dave

[jim The STEAM Clown - Jim Burnham]

yes, I agree... so maybe something like this these 200 AMP automotive relays - https://www.amazon.com/dp/B096QGR24M/ref=redir_mobile_desktop?_encoding=UTF8&aaxitk=a83fde7d6c19d0ee82bebd8adba8d749&hsa_cr_id=9177184970301&pd_rd_plhdr=t&pd_rd_r=e3a1a9a5-38e6-4b0f-b793-d034ec580209&pd_rd_w=4C7eN&pd_rd_wg=XNMe5&ref_=sbx_be_s_sparkle_mcd_asin_2_title

But I'm also looking at a software solution too.  What else do folks use?

[Gmail]

In my case it would need to be a big PRETTY button. 🤣

Thomas

-

Want to learn more about ROBOTS?

https://www.youtube.com/user/thomasfromriverside

On Oct 26, 2021, at 10:51 AM, David Hough <goo...@uks.org.uk> wrote:

A proper, safe last-resort e-stop should take out the power to the

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/06d6c167-1253-a3ab-37e4-46d858074879%40uks.org.uk.

[Jeremy Lizakowski]

All three can be useful.

The most basic (and reliable) should be prioritized, but designing safety into the system with positive control is helpful and sets a good example.

Jeremy

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/f76887de-ee0b-4570-b6ad-c0ddf8aec5b4n%40googlegroups.com.

[Jeremy Lizakowski]

Also, depending on what you are trying to teach, it might be helpful for students to take a field trip somewhere that has good examples, and let them drive the questions if they are comfortable (valuable skill for later).

This could be in construction, self-driving cars/trucks, industrial robots, University labs, etc.

[Michael Wimble]

I use a RoboClaw which has an input which can be assigned as an e-stop. I use this for software control emergency stop. That is, when something has penetrated my third perimeter. By my E-STOP switch is a big, red button that can be easily whacked and cuts off all power from the battery. It doesn’t just stop the motors. I have a deadman switch on a game controller that I can enable to turn off the motors when I’m trying experiments with live driving. But the E-STOP turns off ALL power and doesn’t rely on relays which can fail. Of course the e-stop could fail as well as arcing contacts and broken springs can fail. I haven’t addressed that yet.

[Marco Walther]

On 10/26/21 12:56 PM, Michael Wimble wrote:
> I use a RoboClaw which has an input which can be assigned as an e-stop.
> I use this for software control emergency stop. That is, when something
> has penetrated my third perimeter. By my E-STOP switch is a big, red
> button that can be easily whacked and cuts off all power from the
> battery. It doesn’t just stop the motors.

Did you actually find one which is spec-ed for 100+A DC for those heavy
systems? I use a 'big red button' for my lawn mower, but that's running
on 7S LiPO (25.9V) and normally << 5A.

Otherwise, you would end up with a normally-open relay as Jim found and
your 'big red button' cuts the power to the coil.

> I have a deadman switch on a
> game controller that I can enable to turn off the motors when I’m trying
> experiments with live driving. But the E-STOP turns off ALL power and
> doesn’t rely on relays which can fail. Of course the e-stop could fail
> as well as arcing contacts and broken springs can fail. I haven’t
> addressed that yet.

For my RC-car based systems, I use a RF dead man switch which 'triggers'
on lost heart beat/RF signal or button release. I tried to find a RF
link which is short enough to be useful. It will break around 50 -
100ft. Those are Arduino-based and try to actually 'break' via the
RC-ESC. When you just cut of the power in an RC car in motion, it tends
to just keep coasting:-( Not exactly, what you want.

>
>> On Oct 26, 2021, at 10:44 AM, jim The STEAM Clown - Jim Burnham

>> <jim.the.s...@gmail.com <mailto:jim.the.s...@gmail.com>>

>> <mailto:hbrobotics+...@googlegroups.com>.

>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/hbrobotics/f76887de-ee0b-4570-b6ad-c0ddf8aec5b4n%40googlegroups.com

>> <https://groups.google.com/d/msgid/hbrobotics/f76887de-ee0b-4570-b6ad-c0ddf8aec5b4n%40googlegroups.com?utm_medium=email&utm_source=footer>.

>
> --
> You received this message because you are subscribed to the Google
> Groups "HomeBrew Robotics Club" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to hbrobotics+...@googlegroups.com

> <mailto:hbrobotics+...@googlegroups.com>.

> To view this discussion on the web visit

> https://groups.google.com/d/msgid/hbrobotics/03344361-8186-4477-BE38-2BF233F6A810%40gmail.com
> <https://groups.google.com/d/msgid/hbrobotics/03344361-8186-4477-BE38-2BF233F6A810%40gmail.com?utm_medium=email&utm_source=footer>.

[Sergei Grichine]

For my ride-on lawnmower I had to hack its dead man's switch, located in the driver's seat. I put a plastic insert into its connector (effectively engaging the switch permanently, as if driver is in). The insert can be pulled out by a small gearmotor - which results in "driver fell out" condition, cutting ignition of the gasoline engine.

The gearmotor is activated by a cheapo 433MHz long range remote switch, with the receiver having its own 12V LiPo battery. So, whole system is absolutely independent of the rest of the robot.

It works great, but on my wishlist is a "test mode" operation, so that I could test the switch without pulling the plastic insert. Check the battery, radio, etc.

BTW, for a 100+ Amp interrupter you might need to do better than an automotive relay...

Best Regards, 
-- Sergei Grichine
   

To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/6c4edd38-4f7f-1096-ebaf-c16257a4d63e%40gmail.com.

[Sergei Grichine]

Btw, search Amazon for "Battery Disconnect Switch" - I use one on my tractor's negative wire, and it withstands starter current just fine. The one with little red "keys", about $10. The key engages with a push in, but flips back real easy, can be pulled by a rope or something.

[jsam...@pobox.com]

My belief is that if I press an eStop button that I expect my vehicle to stop. Period... Maybe even with dynamic braking...

I don't believe that Michael Wimble's vehicle draws 100A or more, so that is irrelevant. But he has a point that if your eStop button kills all power then that protects you if your vehicle is smoking or on fire. (But you lose data logging...)

I have the belief that if I press an eStop that any movement will stop. (either a manual local eStop or a remote eStop switch). I have a mobility scooter drive motor and a giant servo for steering.

I use automotive 40A relays in series with both motors. They are controlled by my low power eStop circuit.

I have a low power series circuit that controls my eStop. It has power from the main battery (fused) . The is in series with my main eStop switch (a quality switch and not an industrial crappy switch). That is in series with my eStop relay (from a Lora radio receiver). And the opposite side my of automotive relays have a MOSFETs to ground so the software gets  a "vote" from my microcontroller. So everybody has to agree that "safe" movement is allowed.

My main motor draws over 30A with full acceleration with a 24V supply. It draws about 20A at 12V at full acceleration. It may draw 80A if stalled. But I haven't measured it. The 40A relay my be under spec'd. Two relays in series may be a more "reliable" solution.

I attempted to make my remote control system "fail safe" under all conditions. (out of range. low battery, timeouts, radio interference, bad software...) It is not adequate yet...

So I keep trying to make it reliable under all circumstances...

Maybe I will be on Camp's HBRobotics ROS meeting tonight...

Jeff Sampson

On Tuesday, October 26, 2021, 03:49:01 PM CDT, Marco Walther <marc...@gmail.com> wrote

[jsam...@pobox.com]

Just out of curiosity, what is the reasoning behind you statement "BTW, for a 100+ Amp interrupter you might need to do better than an automotive relay..."

Jeff Sampson

On Tuesday, October 26, 2021, 05:45:19 PM CDT, Sergei Grichine <vital...@gmail.com> wrote:

[snip]

[jsam...@pobox.com]

Sorry, I screwed up this attribution.

Michael Wimble said he has a low powered system. Marco Walther asked he could account for 100+A system....

Jeff Sampson

[Wayne C. Gramlich]

All:

There is a great deal of good content on this thread. As the robots get
larger, the power involved becomes more significant. The high end robot safety
standard, insists on removing all power and having a passive brake that is
has a spring to engage it to slow a robot in motion to a stop.

In general, have plenty of E-stops. There is nothing I hate more than an E-stop
you can't get to (i.e. the robot it coming towards you, it is hard to reach the
E-stop at the back bottom of the robot.)

My preference is to separate the power for the motors from the computers.
Turning off motor power still lets the computer(s) log what is going on.
Be careful, just turning off power does not necessary stop a robot, it could
coast for a while. Most H-bridges are NFET based, which will not clamp the
motor windings when power is removed. So the motor can free wheel for a while.

Regards,

-Wayne

[Chris Albertson]

If this is a wheelchair and the wheels are run through a gear reduction, it will not freewheel very much when the power is removed.  But if you really wanted to apply brakes use a double throw relay that can switch a low-value power resister across the motor when the power is removed.  This will also prevent the robot from rolling down hills with the battery is disconnected.    But check if you need this, I'd guess not f there is gearing.

The e-stop on my milling machine is connected to the enable pin on the stepper motor drivers and they stop nearly instantly.  But again there is gearing, in this case, lead or ball screws.

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/870b5bd1-7264-2eed-6457-70955bf3688c%40Gramlich.Net.

--

Chris Albertson
Redondo Beach, California

[Jeremy Lizakowski]

Safe stop is an interesting topic if the vehicle doesn't stop easily. Airplanes have it the worst, in this sense.  Once airborne, you have to land.

  

Self driving cars have a simpler problem, but a real one:  you can't safely do a sudden-stop on a freeway, or an icy road in the mountains. Pulling to the side in high-speed urban rush hour might be more complicated than driving straight.

/digress

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.

To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/2012508064.898726.1635290440925%40mail.yahoo.com.

[Chris Albertson]

"E-Stop" is not the same as "safely parking".  It is usually a large red button that is about 30mm in diameter and clearly labeled and it's purpose is to instantly freeze the machine.  Usually, it is within arm's reach of a person's workstation.   They look like the image below.    If you robot is powerful enough to cause injury, you need at least one of these.   They should freeze the machine, not cause it to pull over and stop or land.   That is some other function, not e-stop.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/CAKmxcbwmVTBM1A6zSsYGZeD_xnScPydwFmZnE4p25zTgYwqo3g%40mail.gmail.com.

[Chris]

Hey Jim,

Lots of good advice here.  I specially like the feature of shorting the motor for braking.

We like to separate out a power path to the computer.  Since it is logging what is going on, it is handy to have it continue to run to capture what went wrong.

-Chris

[Michael Wimble]

Part of the problem might be a different perception of what an estop is for. If you can walk up to the device and calmly push a switch to stop the robot, that to me is not the purpose of an estop. In my mind, the robot is about to or is doing something dangerous, smoke is pouring out, etc. you want everything to stop, and now, not just the motors to stop. That won’t stop the smoke. 

On Oct 27, 2021, at 7:15 AM, Chris <chris.t...@gmail.com> wrote:

Hey Jim,

Lots of good advice here.  I specially like the feature of shorting the motor for braking.

We like to separate out a power path to the computer.  Since it is logging what is going on, it is handy to have it continue to run to capture what went wrong.

-Chris

--
You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/037c1aa3-0339-4b3c-86e3-c12137b77e5cn%40googlegroups.com.

[Wayne C. Gramlich]

All:

I use a robot moving on a ramp to illustrate some of the issues.
A robot is going up a ramp and E-stop is triggered. What happens?

A. The wheels a slammed into 100% lock up mode (short the motor windings)
and the top-heavy robot pitches forward.
B. Power is removed from the motor drivers, the existing momentum causes
the robot to move up the a ramp a little, comes to a stop, and starts
to accelerate in the opposite direction down the ramp because the motor
windings are not shorted.
C. The power is removed from the motor driver's, but the the motor windings
are pulse width modulated in braking mode to bring the robot to safe stop
leaving the motor windings shorted together at the end for full braking.
D. The excess power from the emergency braking in the previous option surges
into the motor power circuit which has no dissipation resistors to remove
the excess power. The magic smoke is released from the electronic components
and the robot catches fire.
E. There is power dissipation circuit that radiates the power from the emergency
braking.
F. The all power is removed, a the solenoid that was holding a spring actuated
brake engages and brings the robot to a safe stop.

Food for thought.

Regards,

-Wayne

[Carl Sutter]

In a first round prototype, I was looking for e-stop switches that can handle high current.  I used these high amp e-stop switches: https://amzn.to/3CkUXkL

They were fine, but arcing on the contacts was a concern and we had some failures.

But, most e-stop switches are not high current - assume you are using a relay.  So, in the second phase, I added a golf cart contactor.  Many of these have a lower resistance coil and waste a lot of heat.  These were the best:

https://amzn.to/3nCmszY

https://www.alliedelec.com/product/white-rodgers/586-117111/70101935/

I was shutting all of the robot power off with these, but one could argue that the motor power only is sufficient.  In our case, we were also concerned about battery shorts etc. outside the motor control.

There are also solid state contactors, which I have not tested thoroughly.  They would presumably not dissipate as much heat.  Cost is higher.

With the contactor, you can use any kind of smaller e-stop button.  I used these since they were waterproof (outside application):  https://amzn.to/3jKa4gh

One advantage of that e-stop switch was that it had NO and NC contacts.  I had Roboteq controllers with a disable line that you just tie low to shut down, so I used the NC contacts to energize the contactor, and the NO contacts to shut down the motor controllers.  So, some redundancy there.

One high-end remote e-stop vendor: https://www.fortrobotics.com/wireless-emergency-stop

We used heartbeat type shutdowns too, and that is a good idea, but separate from the big red button when the robot is out of control.  I would not mix the two, and do not trust a microcontroller for an e-stop function since the main reason for needing the e-stop is when a microcontroller fails or gets a software update with a bug.  That happens all the time! :-)

I am not an expert on FIRST FRC gear, but they presumably have some standard approaches for this.  I do know that their controllers have a centralized remote shutoff feature for competitions.

[Chris Albertson]

Best practice is to make E-stop as independent as possible.  The worst design would have the E-stop signal go into the controlling microcontroller.  The best design is where E-stop removes power to the motors

Also the E-stop circuit is designed so that the motors stop if the circuit is open.  This way if a cable is cut or falls lose the motors are stopped.   A simple way is that a relay is powered through the E-stop switch(es) so that if any E-Stop switch is opened, then relay opens and power is cut.

But relays that are held "on" with continuous power waste power and you might not want to do that.  So next best system is that e-stop pulls up the "enable" line on the motor driver and if the wire breaks of the e-sttop switch is open the enables are pulled to ground

Do not run the signal through software.  The reason for having an e-stop is if the software has a bug.  However, sometimes you must.  a software e-stop is the only option if you are teleoperating the machine

These are really common with CNC machine tools, most of the designs are found in that community more so than with robots

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/f76887de-ee0b-4570-b6ad-c0ddf8aec5b4n%40googlegroups.com.

[Jeremy]

On Tue, Oct 26, 2021 at 9:32 pm, Chris Albertson <alberts...@gmail.com> wrote:

Maybe there's a better name for it, to not confuse concepts.  It has a lot more in common with an e-stop than it sounds.  

What's a better name for a safety control that is the closest you can safely get to an e-stop?  Maybe "Shutdown" or "Halt"?

I think we will see this concept more as robotics and humans are closer together.    e.g.  A humanoid robot that halts on a staircase could crush someone as it falls.  Instead, it could have a short time window (xyz milliseconds, enforced by a hardware watchdog) to slump or otherwise minimize it's inertia before impact.

[Chris Albertson]

I think what is needed is a multi-layed "stop".   The robot has at least a few layers of control. 

At the lowest are motors and motor drivers and that is the layer on which e-stop operates.  It freezes motion. It is what you use of you hand get caught in some gears.

The next layer of control is usually some kind of goal seeking where the robt tries to get to a waypoint.  This is true for both arms, legs and wheels.  This layers in in the middle.   One could add a feature where the waypoint is moved toequal the closets stable point.   It would not freeze motion and would be no good if you hand is caught in a gear.   In the spacecraft industry where I worked for some years, we'd call this "Save Mode" and the spacecraft would place itself into this mode when it detected a problem with some system.  Perhaps a computer failed or some other hardware broke or contact to controllers was lost..

Finally there is always a higher level of control where the robot is given a goal like "fetch me a beer".  I think a "stop" at this level means "do some some activity to become more safe"   Mmaybe this is for a car, itwill pull overt other shoulder and turn on the flashing lights

The details can be different but I think you can define a safe this that each control layer can do.   As you move up the action gets smarter but the reaction time gets longer

--

You received this message because you are subscribed to the Google Groups "HomeBrew Robotics Club" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hbrobotics+...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/hbrobotics/NBCO1R.SBRT1YZ37M813%40gmail.com.

[Mark Johnston]

Late to the 'party' here.   I found and use these RF handheld switches in ESTOP applications.

These are low current and thus only control some big-o relay or MosFet power switch on the bot but these are a great little remote switch that latches on or off.  It offers a relay that exposes to screw terminals   common,  NC and NO  so use whatever 'flavor' you need.

Go down 4/5 on this page to 'Making a remote RF Estop Switch.   MANY other uses besides robots by the way.

https://learn.ubiquityrobotics.com/mcb_pinouts_leds_userpower

Mark