Hello all,

I am new here on the forum and new to SS, having previously struggled with parallel ports and lost step all round. The machine in question is a Chester super lux that was butchered and modded. Currently actuated on 3 separate chinese stepper drives ant a 50V supply. Software is the latest plugin directly from Greg and the latest drivers from the website. Signal cables are all UTP twisted pair cables for the time being but will be replaced with audio shielded cables today.

A few questions to the learned ones on here:


The problem I am having is that Mach returns an external e-stop whenever an axis hits a limit, and not an axis limit indication.

I have switched off the power to the motors and jogged the axis virtually without the axis actually moving on the machine and pressed a limit. This also gives me an e-stop condition. If I only press the switch without an axis being jogged, mach tells me that a limit has been trip. To make sure it is not noise, I have disconnected the e-stop and done the same thing with no change, with and without an axis physically moving . in the filtering boxes I have 100 in each one. Debounce in the mach settings is 0.

The chinese breakout board I am using seems ok as I have tried the same using the parallel port and it works fine.

Could the driver/plugin be playing up? I do not know enough about the logic the USS follows in order to troubleshoot fully.

I am also slightly perplexed as to the backlash compensation. My axes have 0.02mm and 0.05mm real backlash. These values have been set up in mach and work with the LPT as normal. quick acceleration to take up the lash and then proceed as normal. Does the SS work differently? Asking because I don't get that super quick jolt in the new direction.

Been reading on here that the watchdog time increasing may help solve some issues so will give this a try.

Looking forward to your response.

George

I have noticed today that when the limits are tripped, mach briefly flashes the limit trip info before going into external e-stop condition


watchdog at 100 did nothing either.

I see that the capacitor and resistor are still on my board. Will get rid of those and report back.

Nope...no joy even ground to the case and with the capacitor and resistor removed...

Noticed that Mach does return a limit triggered message a split second before going into external e-stop.

Can you attach your XML? Also can you draw a diagram or attach a photo of how you have your switches wired up?

Hi Jeff,

Please see xml attached. The best way to describe my limit switch connections is the gecko G540 recommendation which you will see in the G540 manual on page 8 of ten.  All the limit switches are normally closed. When the limit is reached, the circuit is broken. see attached photo, (hand drawn)
The only difference is that the pin numbers are governed by my new breakout board.

The breakout board I am using is: http://www.savebase.com/infobase/downloads/Breakout_Board_Stepper_Driver/User_Manual_5_axis_Breakout_board.pdf

I had tried the exact same thing I was doing with the machine, but using the parrallel port (no SS) this morning.

What I would expect, and what does happen when using the lpt:
- you hit a limit
-mach tells you you hit a limit
-you press reset
-mach allows you to slow jog off the limit
-once off the limit mach goes into e-stop and you have to press the reset button

What I get with the SS:

-if I just press a limit switch mach says that a limit has been triggered, which is fine


- if I am moving and hit a limit (even with the motor power switched off so SS is sending signals but not physically moving the machine)
-mach tells me for a split second that the limit has tripped and then says external e-stop requested.
-when you press reset you can jog off the limit and keep going even after the limit switch has been released.
-in the SS diagnostic, the external e-stop does not show as active when the limit is tripped.

Would these phenomena be the reason I do not get the super quick jerk that takes up the backlash? when you jog off a limit in LPT, mach does not do the backlash comp jerk.

I will make a video of the above to make it even clearer.

Why are you using a break out board with the G540? The G540 has its own BOB built right in. You need to take the extra BOB out of the mix and wire your switches directly to the G540.

Jeff,

didn't make it clear sorry. I am not using the G540. I recently replaced it with a breakout board and higher amperage drivers and thought I could explain the wiring I have for the limits by referring to the G540 since you are probably very very familiar with it.

As you say, the g540 does have a BOB but I was not happy with the restriction in the rapid speed imposed by the 3.5 amp dirvers.

You have any ideas on the behavior of my SS?

OK, I was just confused by the reference to the G540. Anyhow, I've never seen the symptom you are describing. One thing I would do it not power the BOB by the USB port connection, use a small 5V power supply instead. If the BOB is putting out any noise on the 5V line and you are using the UBS power it could feed the noise back into your PC. Also many PCs cannot provide much current on the UBS ports so using an external power supply or powered USB hub for the SmoothStepper.

Second wire up only one switch and configure it as a single input (say X home) in Mach. How does the machine work with it set up as a home switch? If that works OK, change Mach so that switch is a single limit switch (say X++) and disable it as the home switch. How does this work?

The users manual for your Chinese BOB is so poorly written there is not telling how it actually works. Maybe the BOB asserts an EStop when a limit is seen. Since the BOB expects a lot of enable signals from Mach (to enable each stepper drive) maybe it disables all of the drives at once when a limit is hit which could cause a huge power supply spike.

Jeff,

makes sense. I had disconnected the enable pins and the cables as the SS would flake out every time the drives were enabled. The power to the switches is coming in through the 5V screw terminals and is taken from the computers power supply. Considering that the BOB works even without the 5V being attached, it is taking probably taking voltage through the LPT port pins.

I can have a look at connecting the limits to the auxilary IO's of the SS and see if that helps. Ideally I would be looking at a simpler BOB.

Really just goes to show how flaky the USB is so I am this close to getting rid of the board and going ESS.

Tonight I spent a good hour and a half reconnecting my limit switches in series. Now I only activate pin 10 for everything (++. --. home)

To be on the safe side, I rewired the e-stop mushroom button to normally open.

So when the limit switches get tripped, all the input pins should read low. This is what they do.

Tested the limit with machine stationary: mach returned limit tripped
Tested the limit with machine pulsing movements: limit trip for a split second followed by the external e-stop

So basically no change, so it is not the huge current draw, since there is no current passing.

Going to have some food and then see what happens if I connect the switches to Port 3 on the SS directly.