Closed Loop Stepper E-Stop Wiring with CL57T and ESS with C25 BOB

Hey all!

I'm pretty new to the CNC world. and am currently running a small Taig mill with a Geckodrive and 2NM steppers. The system itself works pretty well, but I'm running some long toolpaths for injection molds, and I'm running into the occasional missed steps and axis binding. I'm also tired of babysitting the damn thing so I decided to upgrade to a closed loop system using motors kits from Stepperonline that I'm pairing with an Ethernet SmoothStepper and a C25 BOB for the ESS.

https://www.omc-stepperonline.com/ts...-2m-cable.html

Wiring everything together is pretty straightforward, but I'm having a hard time understanding how to wire the alarm state into the e-stop so an alarm state stops the machine.

I've attached diagrams from the documentation for both the closed loop driver and the BOB, but I'm not sure how to make this work together. I see that I come off a 5V terminal on the BOB and wire it to the e-stop, do I just daisy chain those to the Alarm+ terminal on the drivers and daisy chain all the grounds together? Sorry if this seems really obvious, just want to make sure I'm on the right path.

Thank you for your time

I took a stab at making a diagram for the alarm e-stop, does this look correct?

Hi,
whether your diagram is correct rather depends on the stepper drive.

If, for instance, the opto-transistor is actively conducting when there is NO stepper alarm, ie NC (normally closed contact) then yes your circuit is logically correct,
there are some practical considerations that would need to be considered, but it would work.

If, however, the opto-transistor is non-conducting when there is NO stepper alarm, ie NO (normally open contact) then your circuit will not work.

The second of the two pics you posted suggest that the opto-transistor is non conducting WITHOUT alarm condition and conducting WITH alarm condition. Can you confirm?

Craig

Hi,
lets address some of the practical considerations that I mentioned which assumes the phot-transistor is normally conducting. Perhaps this is how the manufacturer has
arranged it or you can program the drive to work that way.

With NC limit switches using snap action micro switches or relay contacts its common to wire them in series. With all of the contacts closed, ie NO fault condition, then current will pass through
the series combination. The only limit is the (hopefully) very small contact resistance of each switch, lets say 100mOhm. Three switches in series would have a combined resistance
of 300mOhm, which is still very low and the circuit would work as intended.

The switch in the drives are not relay contacts....they are phototransistors. Even when a transistor is saturated, ie conducting, there is still a small residual voltage between the emitter and
collector. A small signal transistor in switching service will have a V_CEsat of 0.1V to 0.5V. Photo transistors tend to have V_CEsat near the upper end, lets say 0.5V.

If you connect three saturated (conducting) phototransistors in series you could end up with a voltage drop of 1.5V. If your controller is a 5V device then 1.5V is a significant fraction
of the voltage swing and means a much reduced noise margin.

If however your photo-transistors have a V_CEsat of only 0.1V then the series combination will have a voltage drop of 0.3V, which is perfectly respectable
and your proposed circuit will work fine.

Craig

Hi Craig,

Thanks for your response! The following is from the manual (https://www.omc-stepperonline.com/download/CL57T.pdf)

ALM+ : Configurable Digital Output Signal: A configurable OC output signal. It takes a sinking or sourcing 20mA current at 5-24V. It can to be configured as one of the 3 types, ALM ALARM (default), IN POSITION, or BRAKE CONTROL through PC software.

Can I assume that the low sinking/sourcing current means the saturation voltage is low enough for my proposed diagram?

Also I can configure the output to be NO or NC through software.

Hi,
the way you have drawn your circuit diagram you must program the drive such the the opto-transistor is conducting IN ABSENCE of an alarm condition.
You will need to measure the voltage drop across the transistor when conducting 20mA say to determine what the V_CEsat is low enough for your scheme to work.


The bigger question is why you are doing this?. Remember the ESS has a total of 51 inputs and outputs. There is no need to combine all the ALRM outputs into one circuit,
you have enough inputs to have one input per ALRM.

My new built mill has three servos, but my BoB/ESS has five servo outputs should I expand. Each of the five (potential) servos has a dedicated ALRM input to the BoB/ESS combo so that I know which
servo faulted, it shows an LED on screen. If any of the servos fault then any one or more will cause an Estop.

This means I can program each one individually to be NC or NO as best suits that particular servo. I don't have to worry about trying to wire each servo in series so the V_CEsat
problem just disappears.

In the early days of Mach when everyone used a parallel port that had only five inputs you had to be creative to combine as many limits, alarms, homes etc as possible, but why do we still try to do it
now?

For instance my new mill has six limit switches (++ and a -- limit per axis), three homes switches (one per axis), and each has its own BoB/ESS input.......I mean why not I've got plenty to spare.

Craig

Hi JonP - What about the axis binding? Your new motors are not going to solve that? have you addressed the mechanics? Peter

Quote:

---

Originally Posted by joeavaerage

Hi,

The bigger question is why you are doing this?. Remember the ESS has a total of 51 inputs and outputs. There is no need to combine all the ALRM outputs into one circuit,
you have enough inputs to have one input per ALRM.

My new built mill has three servos, but my BoB/ESS has five servo outputs should I expand. Each of the five (potential) servos has a dedicated ALRM input to the BoB/ESS combo so that I know which
servo faulted, it shows an LED on screen. If any of the servos fault then any one or more will cause an Estop.

This means I can program each one individually to be NC or NO as best suits that particular servo. I don't have to worry about trying to wire each servo in series so the V_CEsat
problem just disappears.
Craig

---

Trouble iis, aren't you relying on software only to trigger the stop this way?.
I've always been told to incorporate a hardware stop where possible in a circuit.
The photocell voltage drop is an issue on 5v. Isn't that one reason why many controllers now use 24v?. (as well as noise suppression advantages).

Personally in this whole instance I would rather:
Have my estop running at 24v through an input (5v is too sensitive to any emi noise for my liking). Then use a seperate 24v/5v opto module to connect the estop and the alarms in such a way that they can shut down each other and activate the controller.
Or something like that.

Simplify it: with the drive alarm outputs connected to a single 5v relay. Attach relay to stop circuit so any drive will switch it. Have estop trigger controller and cut power to drivers.

Hi,

Quote:

---

Have my estop running at 24v through an input (5v is too sensitive to any emi noise for my liking).

---

This is fundementally and theoretically flawed. When electrical noise approaches the level of the signal that when your system will have issues. When the signal and the noise are equal
that would be a signal-to-noise (SNR) ratio of 0dB. If the noise is half the signal level SNR=3dB, when it is 1/4 of the signal SNR=6dB and when its 1/10th SNR=10dB. You should aim for 10dB
or better.

The transfer of noise energy into a circuit is a POWER transfer. If the circuit into which the noise energy is being transmitted is high impedance then the noise will be high voltage but low current,
conversely if the circuit impedance is low then the noise will be low voltage but high current, in either case the noise power is the same.

Lets say your 24V control circuit conducts 5mA when active, thus the circuit impedance = 24/0.005=4800 Ohm. The signal power would be 24 x 0.005= 0.12W If a 0.12W noise energy pulse
were conducted into this circuit it would have a 0dB SNR.

Compare that with a 5V control circuit. A 5V signal at 24mA, for a circuit impedance of 208 Ohms, and a signal power of 0.12W. In identical fashion if a noise energy pulse of 0.12W were conducted into this
circuit the result would be SNR=0dB. Its just the noise voltage would be approx 1/5th of the 24V case and the noise current would be approx 5 times the 24V case.

The noise immunity of both circuits is identical. It the circuit impedance that determines the noise immunity not the voltage level.

Having said that many industrial controls are 24V and thus there is a wider selection (cheaper??) of relays, switches etc to choose from and is preferred, but not because of noise, 5V can be just as
good if the circuit impedance is chosen correctly.

Craig

Quote:

---

Originally Posted by peteeng

Hi JonP - What about the axis binding? Your new motors are not going to solve that? have you addressed the mechanics? Peter

---

I've been trying to figure this out. I oil the leadscrews often, but the binding seemingly is pretty random. It mostly happens in the x and y axis, but has happened once on the z. It seems to usually happen during rapids (maybe I need to turn down my rapid speeds?), I'll hear one of the steppers whine and slam the e-stop. I'll then tell it to go to XYZ zero to see which axis caused the issue. I did a bit of troubleshooting in this thread:

https://www.cnczone.com/forums/stepp...20706-cnc.html

My decision to upgrade the steppers and controller is to ditch the parallel port so I can use a more modern version of Windows on my mill computer, and for the alarm output to give me a little more peace of mind during really long toolpaths.

Hi JonP - Maybe more an acceleration issue then a velocity. Dial back the accel. and check out the mechanics. Plus

Quote:

---

Originally Posted by joeavaerage

Hi,
the way you have drawn your circuit diagram you must program the drive such the the opto-transistor is conducting IN ABSENCE of an alarm condition.
You will need to measure the voltage drop across the transistor when conducting 20mA say to determine what the V_CEsat is low enough for your scheme to work.


The bigger question is why you are doing this?. Remember the ESS has a total of 51 inputs and outputs. There is no need to combine all the ALRM outputs into one circuit,
you have enough inputs to have one input per ALRM.

My new built mill has three servos, but my BoB/ESS has five servo outputs should I expand. Each of the five (potential) servos has a dedicated ALRM input to the BoB/ESS combo so that I know which
servo faulted, it shows an LED on screen. If any of the servos fault then any one or more will cause an Estop.

This means I can program each one individually to be NC or NO as best suits that particular servo. I don't have to worry about trying to wire each servo in series so the V_CEsat
problem just disappears.

In the early days of Mach when everyone used a parallel port that had only five inputs you had to be creative to combine as many limits, alarms, homes etc as possible, but why do we still try to do it
now?

For instance my new mill has six limit switches (++ and a -- limit per axis), three homes switches (one per axis), and each has its own BoB/ESS input.......I mean why not I've got plenty to spare.

Craig

---

The reason I was doing it the way above is because I'm not the brightest bulb, I just assumed that was the easiest way to get it done. I'll go ahead and do one alarm per input.

I’m building an almost identical system with Nema34 motors, CL86T drivers, ESS, and a C25S Bob. I’m stuck on the alarm signal too. I was going to run it through the estop originally like you. I also thought about running the alarms to unused limit switch inputs. How did you ultimately solve this problem?