oe8bck

Hi,

I just want to mention that I improved the robustness of the communication a lot by adding an Amidon Ferrite Core to the USB cable:

http://www.amidon.de/contents/de/d382.html

http://www.amidon.de/contents/media/mansper.jpg

I used this type of core as it was available in my shack:

http://www.amidoncorp.com/ft-100-43/

KR,

Christof

Well, I just used a 12V battery to power the stepper brick but it's still the same. With AC motor, the stepper is stuttering and this way it is not working as it has no torque.

KR,

Christof

Hi, thanks for the quick and helpfull answerts!

Well, as I got the AC snubber filter for the AC-line of the inverter and also attached some ferrites on the signal and supply lines. I did not remove the Raspberry, I just did some separation and re-routing of the cables, so that there is only the ethernet cable "closing the loop". I also replaced the 12V supply of the stepper with another one, both are switching power supplies. Now the Stepper Brick does not reset any more. 

So now the system is working in principal, but there is still flaw in the stepper function. If the AC motor is running, the stepping is not smooth, it's "stuttering". If I unplug the AC motor and leave everything else as it was, the stepper works smoothly. I think I will look at VPP with an Oszi to see what's going on there. I would also like to have a look at the current through VPP, but I cannot find R17...

BTW:

I used a simple "Powered USB Hub" without any special isolation.

KR,

Christof

Hi,

Well, thanks to confirm what I thought about the forum posts I read. My question for RS485 is a little bit different.

Regarding the EMC, I already tried what you suggested, the situation improved, but is still not acceptable. I inserted with a Raspberry with a brickd which is connected via Ethernet to the Laptop. Did not help. Maybe I'll try it with WLAN, but this is not an option for the demonstration. I also seperated it physically without the AC motor in the loop. This is the option that really helped, but there is still a reset if the motor is powering on orf off. Also a powered USB bridge was not chenaging the reset behaviour. I will also insert some shielding, the cases as well as the lines with proper shielding, not conenctied on both sides but only on one side. I think this will also improve the situation a little bit.

I did not try to disconnect the stepper supply, that might give a hint. One of the things I will try next as you suggested is a battery to de-couple supplies. But the very first thing I'll do is to apply an AC snubber filter, I got the delivery today.

I am looking forward that the AC-line snubber filter solves the problem, but if not my very last idea is the RS485 line. I orderer yesterday a Master Brick with RS485 Extension and it's on delivery today, so I think I'll get it on Monday, as it has not to cross the Atlantic :-)

KR,

Christof

Hi and thanks for the quick answer!

I know that it's not officially supported, but I thought that someone already patched the brickd in a way that it is using the serial line interface which is RS485.

The connection is quite straight-forward I think.

AC Mains -> inverter ------------------------------> AC motor (1 phase)

              |-----------------<-> RS485 interface

                                          |

    |-> laptop power supply -> laptop <-> USB

                                    |<-> USB <-> Stepper Brick <-> Stepper

    |-> switching power supply --------------------------->|

So there is a loop with the stepper switching power supply an in this loop (really physically in between) is the 1-phase AC motor.

As I said I would like to improve the power supply connection and think that RS485 is a better way to connect the Stepper Brick than USB.

KR,

Christof

Hi,

I found some topics regarding connection to external MODBUS devices from the RED brick, but I have a little bit different setup.

Currently I have a frequency inverter controlled via MODBUS and a Python module. I also have a Master Brick with a stepper motor and its Python module. There is a Python program accessing both modules and it works well as long as I do not connect the AC motor to the system. So stepper control wors, inverter control without motor works, but starting the motor resets the Master Brick.

So there is somewhere a coupling from the motor to the Tinkerboard system. I did a lot of tries to lower this coupling from the power supply side and also with some ferrite beads on the cables. It is improved now a little bit, but still not functional. 

The last thing that comes into my mind in order to reduce coupling is the USB interface. I know from other projects that the ICs used there are very susceptible with respect to EMC. The MODBUS interface with its physical layer RS485 is much more robust. 

So I would like to replace the USB connection with a MODBUS interface. Is there a way to do it?

BRICKD<->PC<->MODBUS<->RS485 Extension<->Master Brick

To make it a little bit mor "thrilling", there is a timing constraint given by the customer. He wants a demonstration by mid of next week, so there is not time to order additional hardware. Any ideas are highly welcome!

KR,

Christof