Hello OpenBuilders, I need help. I have an ACRO 2020 with the BlackBox X32, a z-axis gantry and brushless Genmitsu spindle motor and controller. I have already used it for a couple of projects. Only hick-up was that sometimes I was not able to connect to the BB but after pressing the reset button it always worked. I was planning to add a custom module to the gantry that operates at 5V (Arduino Nano plus 5V stepper motor) and reads the analog 10V tool signal from the BlackBox (with proper downscaling to 5V for the Arduino). When the ACRO motors are powered, the Arduino module goes crazy. I disconnected all but one stepper motor and hooked up an oscilloscope to analyze the 24V stepper signal (downscaled to 12V using 20kOhm voltage divider) and also measure the EMI in the (now disconnected) wires intended for the 5V module. The attached screenshot shows the two signals (red = stepper motor, blue = EMI in other wire). The BB stepper voltage seems to step-wise switch from -24V to 0V to +24V to 0V while idling (holding current, $1=255) at a frequency of around 12.5kHz. Is this a normal PWM type of holding voltage? All cables are unshielded and bundled together and have a total length of around 5m. I know this is asking for EMI trouble. However, the EMI voltage in the free wire shows peaks of about 16V! This is not normal, is it? This behaviour is the same no matter which of the steppers and stepper cables I connect and which other cable I measure the EMI voltage. And I have checked every wire for resistance (all close to zero) and any faulty leak to ground (none found). The PSU outputs a stable +24V. Polarity of all wires has been double checked. And now the BlackBox seems to have stopped working at all. Even when completely disconnected except with a USB cable, I get the connection error. I have tried different USB cables and computers, UGS and CONTROL and reset the BlackBox multiple times, to no avail. ```` [13:19:38] [ connect ] PORT INFO: Port is now open: /dev/ttyUSB0 - Attempting to detect Controller... [13:19:38] [ connect ] Attempting to detect Controller (1): (Autoreset) [13:19:38] [ connect ] Attempting to detect Controller (2): (Ctrl+X) [13:19:39] [ connect ] Attempting to detect Controller (3): (others) [13:19:39] [ connect ] Attempting to detect Controller (4): (DTR Enable) [13:19:40] [ connect ] ERROR!: No Response from Controller - See docs:blackbox:faq-usb-connection-failed [OpenBuilds Documentation] for troubleshooting information. Closing port /dev/ttyUSB0 [13:19:40] [ disconnect ] PORT INFO: Port closed ```` Can someone help me understand the problem? Even if I have to get a new BlackBox, I don't want this to happen again. Many thank in advance! Eric
Yes, reasonably, the drivers do what they have to to lock the motor in place, maintain current setpoint, etc Stepper Motor Operating Principle And Microstepping Explained - the scope trace looks typical, just the driver doing what it should. Shows why separating signal wirings from current carrying wires is important - parallel for 5m - quite a coupling! (1:1 transformer essentially, and as you see, stepper = high freq AC) See section 2.1 of https://docs.openbuilds.com/doku.php?id=docs:blackbox-x32:faq-emi That one sounds like section 6.3.1 of https://docs.openbuilds.com/doku.php?id=docs:blackbox-x32:faq-usb-connection-failed - with all the odd connections for measurements and other arduinos, resistive dividers etc - could be that something went wrong there? Short / overvoltage / backfeed / etc, could be from the coupling 16v above too, tricky to know for sure
Thanks a lot for your reply, Peter. Yeah, I had already read all the documentation about EMI etc you linked. I just didn't expect the coupling to be so extreme (some back of the envelope calculations predicted EMI in the range of 10mV). Looks like I fried the BlackBox by feeding the 16V EMI back into the 10V analogue output. Is that a possible explanation? Looking ahead (rebuilding with a new BlackBox), I consider the following changes: - use shielded cables for the steppers, earthing the shield - putting ferrite cores around all cables. does this also make sense for the EMI sources (stepper cables) or would that negatively impact the stepper current signal? - rerouting the wires (then shielded and with ferrite core) for the 5V module, as far as possible from the stepper wires Any other suggestion? Thanks!
Hard to know for sure, would sort of expect that to only take out the 0-10v section's drive Opamp, but when components fails, who knows - opto is connected to MCU's PWM pin, 24v Supply and then its 0-10v output so could happen I guess I'd think more during the testing/measuring: slip of a probe, wired up wrong wire, shorted something, etc Normally not done, but guess you could Only suppresses High freq interference, still not going to fix it all Thats always the better option, not as complicated, gets the job done - a little space / wire routing to avoid parallel paths etc, much better Didnt maybe try powering the 5V module from BlackBox's servo 5V or similar?
thanks for your feedback on the intended improvements! also the PWM itself could have been hit by the EMI through the 3V-to-5V PWM amp I built which shares the 5V source with the power wires of the 5V module at the gantry (if the reverse EMI voltage exceeded the blocking voltage of the transistors) that can never be ruled out unfortunately nope, I didn't touch the servo output of the blackbox
Or not EMI, just a malfunction in the circuit, before reconnecting it, do measure for backfed voltage on its inputs (just generic advice, I am sure its solid, but just a due dilligence check to prevent damage) Solid yes, just figured I'd mention (Servo output can give up to about 500mA, if one draws too much that could cause the onboard switchmode to suffer too)
