Experiences with NEWKER / SZGH CNC1000 MDb/c, view into performance and precision

Hi all,

as I was able to gather massive experience about the pros and (mostly) cons of a SZGH CNC1000 controller, I want to share them, to help one or the other with information before buying and having bad surprises.

Short facts: the controller including operation panel, MPG and all cabling, was bought via Alibaba for 1111USD incl. shipping, w/o tax/customs.

Delivery was fast via FedEx. The hardware is really making a good impression.

The documentation is unfortunately in typical chinglisch gibberish, and not at all sufficient for a complex controller.
Don't want to brag about the service more then I need - let me say just this: the missing information needs to be requested via a whatsapp channel on a daily basis, which allows 1-3 messages at max. And after 1 day, Mike, the counterpart, seems to have forgotten all of the history, so restart from scratch.

Additionally I could identify the original manufacturer of the controller called NEWKER, who does the main HW as well as SW. They were more friendly and helpful. And the controller would have been cheaper from them ex works. Oh well...




I don't want to share all experiences now - who wants to know more, get in touch directly. I am open to share any.


So, let's get to the main point of my experiences: elementary for me is precision and performance.
And my result after testing this controller is: in categories precision and performance it is not at all up to the expectation for modern fast machines (high speed spindle, linear rails).

How do I get to this conclusion - I did run 2 test scenarios:

1) In the first scenario I run a program with 5 circles in 5 different speeds (F500,1000,2400,3600,6000). This program is run for many different settings between no smoothing to maximum. During the test I measure the generated pulses for the axis drive, and calculate the difference from the ideal path. Please find the table and graph about the parameter sets and the deviation attached.

2) In the second scenario I run a small part of a adaptive clearing of a pocket - operation, which consists of many small fast X-Y-segments. I measure the vibrations of the machine - in this case not by sensor, but by feel, subjectively. There were hefty operations, which stopped first at parameter settings 28_25_25_512 - which leads to the problem right away. You can either choose smooth processing of small fragments and accept horribly deviations or accept very jerky movements (which definitely damage the machine mechanics) during small fast segments.

This scales the faster you go - so who is fine to finish machine with F500 or less might be satisfied with this controller.
For oll others, who are using high feed and high spindle rpm - you might become happy with another controller.


Hope this helps one or the other...

Regards
Daniel