Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: simpson36 on September 20, 2009, 03:32:05 AM

Title: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: simpson36 on September 20, 2009, 03:32:05 AM
Well, wonder of wonders, here is my first circuit design and prototype electronic thingy  ;D

With guidance from tech support at CNCdrive, I was able to figure out this circuit. This was created to take care of the weak error line signal from the CDCdrives products, but it will take even a very weak signal anywhere from about 3V to 15V and put out multiple strong 5V signals to drive a remote LED and even a stubborn BOB with hungry optoisolators.
This circuit solves, for example, a specific problem discussed here not long ago. A solution to noise causing false e-stops on limit switch or other e-stop lines is to increase the voltage to 12V, however, that creates the problem of interfacing that back to the 5V BOB. This circuit solves that problem. The 12V line from the limit switch goes into the Error Line pin and the 5V out goes to the BOB.

First I made a prototype with a Radio Shack proto board. Schematic has been added here - scroll down to chapter 5.5:
http://www.cncdrive.com/datasheets/Dugong_manual.pdf (http://www.cncdrive.com/datasheets/Dugong_manual.pdf)

This is the top and bottom of the first prototype:

(http://www.thecubestudio.com/CNCdrive/Interace4050BackWEB.jpg)
(http://www.thecubestudio.com/CNCdrive/Interace4050FrontWEB.jpg)

After the circuit was tested and working, I went on to design and rout a custom PC board and put it all together. I have purchased the SMD version of the chip in case a wave of motivation overtakes me (unlikely any time soon).

Here is the routed PC board:  1/32" end mill 28K rpm and 1/16" endmill at 7K RPM.
(http://www.thecubestudio.com/CNCdrive/ErrorLineInterfaceRoutedBoard.png)
And top and bottom views of the completed part. . . less IC chip. . and prior to cleaning up the rosin:
(http://www.thecubestudio.com/CNCdrive/ErrorLineInterfaceBottom.png)
(http://www.thecubestudio.com/CNCdrive/ErrorLineInterfaceTopTrans.png)







Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: Chaoticone on September 20, 2009, 11:36:40 AM
Loks good Simpson.  :)

Brett
Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: simpson36 on September 26, 2009, 03:40:38 AM
Thanks!

Here is the G-code to route the board for anyone so inclined. Ignore the speed on the first two files. I used 28k RPM die grinder mounted on the mill head for the 1/32" end mill. The 1/16" end mill for the second two did run at 7K in the machine spindle (max speed), but would be better run at routing speed if you have a way to hold it.  

If you want to rout and drill each size with the same endmill, you can combine the first two and the last two files.

http://thecubestudio.com/CNCdrive/ErrorLinePCBtrace.tap (http://thecubestudio.com/CNCdrive/ErrorLinePCBtrace.tap)
http://thecubestudio.com/CNCdrive/ErrorLinePCBdrillSmall.tap (http://thecubestudio.com/CNCdrive/ErrorLinePCBdrillSmall.tap)
http://thecubestudio.com/CNCdrive/ErrorLinePCBdrillBig.tap (http://thecubestudio.com/CNCdrive/ErrorLinePCBdrillBig.tap)
http://thecubestudio.com/CNCdrive/ErrorLinePCBCutOUT.tap (http://thecubestudio.com/CNCdrive/ErrorLinePCBCutOUT.tap)


Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: Jeff_Birt on September 27, 2009, 10:20:24 AM
When I see multiple outputs from a circuit board labeled E-Stop warning sirens go off in my head. The purpose of setting Mach up with an E-Stop input is to let it know when an E-Stop occurs, but this is an auxiliary function. The E-Stop circuit however is its own animal and should not depend on Mach know an E-Stop occurred for all motion to stop (power removed from drives) A proper E-Stop circuit should no depend on any single IC on any circuit board for proper operation.
Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: simpson36 on September 28, 2009, 02:19:26 AM
When I see multiple outputs from a circuit board labeled E-Stop warning sirens go off in my head. The purpose of setting Mach up with an E-Stop input is to let it know when an E-Stop occurs, but this is an auxiliary function. The E-Stop circuit however is its own animal and should not depend on Mach know an E-Stop occurred for all motion to stop (power removed from drives) A proper E-Stop circuit should no depend on any single IC on any circuit board for proper operation.

This is equivalent to the argument that one should not own a gun because guns kill people. The circuit is just a tool. It can be used in any way the user chooses. As I mentioned, the specific purpose is simply to amplify a signal and/or to  take a variety of different signal levels and voltages as input.

Servo drives have error lines that communicate operational problems. This little circuit simply takes that error signal and amplifies it for reliability and as a bonus, allows the user to have remote LEDs to easily monitor what's going on. The chip has enough pins to handle two drives. Those drives would be in the system anyway and would have error lines, so the circuit board does not 'add' anything new nor eliminate any options relative to the preferred e-stop scheme.

It also would provide a convenient means to run a higher voltage on limit switches. Presumably your e-stop scheme uses limit switches? There are undoubtedly a dozen other possible uses for this little guy having nothint to do with e-stop.

Setting up an e-stop scheme outside of MACH makes no sense to me for many reasons.  Occasionally I read posts about this concept, invariably presented only as a vague reference (as you have also done), but so far I have never seen anyone provide specifics on the how and why of this theory. I would be most interested in an explanation.
Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: Jeff_Birt on September 28, 2009, 09:17:48 AM
Quote
Setting up an e-stop scheme outside of MACH makes no sense to me for many reasons.


You are completely backwards on this (I am only speaking towards an E-Stop circuit here). Your E-Stop circuit should be able to function independently of Mach and any other control/circuit board. The E-Stop HAS TO function even when the control fail (including Mach and/or PC).

For example: on my STDR-4C drive box the E-Stop kills AC to the power supply and consequently the G540 drive. So the motors stop, there is no power present to move them and the G540 stops thereby signaling Mach that something is wrong (external E-Stop). The E-Stop in Mach is only a signal to let Mach know that an E-Stop occurred Mach is not or should not be depended upon for E-Stop purposes.
Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: simpson36 on September 28, 2009, 10:32:57 AM

You are completely backwards on this (I am only speaking towards an E-Stop circuit here). Your E-Stop circuit should be able to function independently of Mach and any other control/circuit board. The E-Stop HAS TO function even when the control fail (including Mach and/or PC).

For example: on my STDR-4C drive box the E-Stop kills AC to the power supply and consequently the G540 drive. So the motors stop, there is no power present to move them and the G540 stops thereby signaling Mach that something is wrong (external E-Stop). The E-Stop in Mach is only a signal to let Mach know that an E-Stop occurred Mach is not or should not be depended upon for E-Stop purposes.

I don't know what you mean by 'backward'. Clearly I am talking about servos  . . but you respond using a stepper controller for your argument, so you are not addressing the topic, or you do not understand it. Steppers are open loop so there is no equivalence to my comments on the usefulness of the circuit board at issue.

Perhaps you have a narrow definition of e-stop, but  you have not provided any more useful info than in your first post, so it is not possible to debate the issue. Blanket statements without any explanation, facts or details is not useful. You are just stating your unsupported opinion as if it was fact. To make a useful conversation, you would need to provide, as a minimum, the following:

1) define what you mean by e-stop.
2) what triggeres your e-stop? If a servo drive faults, what mechanism do you prefer to stop all movement and shut off the spindle?
3) if your DC stops as soon as you kill AC power, it would seem your PS is not designed properly as it should have large capacitors.
4) You state that killing the AC power to the controller is the correct way to stop the stepping action, but you do not explain why a catastophic problem with MACH, like a software crash, which would accomplish exactly the same thing, is unacceptable.
5) presumably you are aware that even if killing the AC would instantly remove power from a DC servo motor, that could actually CAUSE destruction because the servo drives would then have no way to stop the motors and they could easly coast into mechanical stops.
6) in your scenario, what is the purpose of 'informing' MACH of your external e-stop. To what end?



 
Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: Jeff_Birt on September 28, 2009, 10:57:56 AM
I have designed/installed countless automated industrial systems: I do know what I am talking about. I have limited this discussion to E-Stop as it was the point I first brought up. I am not picking a fight with you, just stating that a circuit like this is not proper to be placed in an emergency stop chain.

Depending on the machinery involved many different things may need to happen upon an E-Stop condition, therefore it is impossible to discuss EVERY possible scenario in a post such as this. On servo systems it is common to have a braking resistor switch in at the same time as the mains is cut to the DC to bring the axis to a controlled stop. On a vertical axis it is common to have a brake on the drive motor that comes on automatically when power to the drive motor is cut. Pneumatic/hydraulic systems will have their own requirements.

The example I gave previously was just that an example: it described a simple emergency stop circuit that did not relay on the control (Mach/PC) to work. Other equipment I have designed say an automated welding cell, something that is more complicated with light curtains/pressure mast would have all of these safety devices wired in series to a MCR (Master Control Relay, a safety relay that will always work) that might cut power to a rotary table while the same time signaling the robot controller that an E-Stop occurred. The robot controller than initiates its own internal E-stop scenario (remove servo power, lock brakes).

The point is that each device responds independently to an E-Stop condition, none of them are dependent on some master controller (Mach/PC) to tell them that an E-Stop has occurred.
Title: Re: Circuit for remote LED indicator and reliable e-stop signals - photo & link
Post by: simpson36 on September 28, 2009, 12:06:11 PM
I have designed/installed countless automated industrial systems: I do know what I am talking about. I have limited this discussion to E-Stop as it was the point I first brought up. I am not picking a fight with you, just stating that a circuit like this is not proper to be placed in an emergency stop chain.

First FYI, I enjoy a good fight, so long as it is a debate with facts and not an argument with emotions. Second, if I though you didn't know what you were talking about, I would not waste my time discussing anything with you at all, so please consider that. I'm hoping there is something for me to learn here . . . if I can drag it out of you . . ;)

There are two talking points here. One is the methodology of your e-stop scheme. I do not agree with the little bit you have detailed, but I agree that it is a large subject not suitable for a hijack of a much simpler thread topic.

So, to confine my response only to the portion of you statement that I have quoted, you still have not spoken at all to your continued contention that the simple circuit is 'not proper' in an e-stop scheme. The circuit does nothing but take a weak signal, one not designed to drive an optoisolator for example, and boosts it to a full solid 5V signal. It does not create a signal, nor block the signal, nor interpret or change the signal in any way. There is simply no way this can have any negative effect on any e-stop scheme you devise.

You have talked around this, but not address it specifically. Can you specify how an amplifier on an error line can in any way effect whatever you, I or Joe down the street considers to be the optimal e-stop arrangement?