TB6560 timings EMC2

[y2kiah]

I cannot seem to find any information about proper timings for the cheapo 3-axis TB6560 driver board that I'm running. Started out default timings of 5000ns, 5000, 20000, 20000 (Step Time, Step Space, Direction Hold, Direction Setup).

The motors ran very choppy, so I upped the step time to 6000 and that got the motors running more smoothly. Currently I've tweaked the timings to 6000, 2000, 10000, 10000. Looking at the values for the built-in list of drivers, I see such a huge variation that I'm not at all satisfied with what I have, and further tweaking would take a long time to even notice a difference. I'm not even sure if lower values are necessarily "better" as I have been assuming.

Also, there is a lot of vibration in the motors currently, some times more than others, but occasionally it gets really bad, and I just know these motors can be run much smoother. Should I try micro stepping to resolve the vibration problem, or could this be a matter of fixing the timings?

Anyone with this driver chip found good settings that work well? Any help would be appreciated.

[pfred2]

I don't have my TB6560AHQs yet but I have been studying this a lot lately:
http://www.toshiba.com/taec/componen.../382/27885.pdf

I don't know what your board allows you to adjust, but just the ICs have a lot of different ways of setting them. M1, M2, and DCY1, DCY2 seem to be the critical adjustments to me looking at the data sheet. It may be best not to try to switch those while the device is enabled.

I also found this interesting as well:
http://www.glyn.de/data/glyn/media/d..._en_090324.pdf

Reading many horror stories about people trying to use this hardware leads me to believe that lots of people are doing things wrong. Popular opinion is that one should not apply any more than 24VDC to these ICs or try to drive a motor any more than 1.75 Amps per phase either. No matter what claims have been made about the product! I know all about the 36V 3 amps but I don't believe it for a minute. Maybe with extreme care the amps per phase can be increased slightly, but I doubt it as of now. I bought a few to blow up though so I'll find out.

In a few weeks I will know much more definitively about this component myself. I've more resources than most do to figure out how to use these devices.

First draft of my test circuit:
http://img519.imageshack.us/img519/237/tb65601.png

If you're running EMC2 come hang out on the IRC channel I usually make it in there after 6PM EST myself. #emc on irc.freenode.net More information here:
http://www.linuxcnc.org/content/view/4/8/lang,en/

[y2kiah]

thanks pfred2, yeah I've already set up the decay and current with the dip switches on the board itself, running it on 24V, lowest current.
Here is my first video
YouTube- Home built CNC machine first test
I'm really mostly concerned with the timings for the driver to be entered into the EMC2 stepconf wizard. I browsed through the TB6560 datasheet and did not find anything relevant there (maybe I missed it though...) and if it's not there, I'm not sure where it would be. Someone who has tweaked the timings of this board might come along and help us both out.

I would love to reduce the vibration of these motors to as low as possible. I might have to try microstepping, I read in another post that it can smooth out the motors, but I would still like proper timings.

[pfred2]

In the datasheet under "Electrical Characteristics" it says:
Minimum clock pulse width C = 330 pF 30 uS (the C in question here is the C hooked up to pin #7 OSC on the TB6560AHQ)

Now I think EMC2 uses ns so you're going to have to convert X 1,000 but I'm bad at math and my memory isn't always super accurate either so you'd better double check that! You can also use a C there in a range of 100-1000pF. See section 8 of the datasheet.

My TB6065AHQs haven't arrived yet but I am excited to try some microstepping with them myself. With microstepping the value of C can play a factor in top motor speeds.

Forgetting about stepper motors for a moment the smaller you slice up time on your PC the less time you have for other threads on your system. So, your best timing is the slowest you can get away with and still get the speed out of your motors you need for your machine. That 30 us example was a "minimum" timing. Everything is a trade off. The only real question that remains is just how fast do you actually want to spin your lead screw? Longer timings are better, unless it doesn't allow you to run fast enough.

I think if you switch to microstepping to reduce resonance you should run smoother than with full step sequencing.

Other possible causes for rough stepper operation would include, but are not limited to, noisy power supplies (some switching supplies aren't the best), noise in a stepper drive circuit (drive crosstalk), weak power supplies, weak step or direction control signals (some ports aren't too swift either), sunspots ...

Anyhow good luck.

[ashfaqjuna]

Hello

I am new to cncs and this forum too.

I have a TB6560 based controller too with 3 axis.

the spindle is manually controlled

I will try out the settings and let you know.

BTW, I have tried this on MACH3 demo running on windows and the steppers run a lot smoother.

The settings given to me for mach 3 are as follows:

Steps per: 320

velocity 2000mm/min => 33.33 mm /s

acceleration : 300mm/s/s
Direction pulse: 5 us
step pulse 10us



does this give any ideas?

thanks

AShfaq

[geekinesis]

I have tried the settings you suggested with some success.

the power supply is 12v

I have 1.2 a steppers. I am setting the current on the board to .8 amp and - have set the decay dip switches to slow and - I have set the step dip switches to 1/16.

So far this is the quietest way of running the motors and the only noise is the motor (hiss) that others have reported.

I get not vibration at all with these settings and smooth quiet operation.

I can set the current to the motors rated 1.2 a but then the vibrations are more audible again. I am using the board with a laser so holding torque isnt an issue.

My only thought is if I am harming the motors or potentially the board by under powering the motors?

[hazardouschurch]

Try this link and set up similar to his machine. This got my motors turning his machine is a 7' by7'. You can change the size to your machine and be able to have moderate speeds with his setup. Its a good staring point to set up emc2 for the driver. My motors had some vibrations but not enough to hurt anything once in the machine.

Morgan's EMC2 Config for TB6560 driver - Zen_Toolworks_Wiki

[Tom_M]

I bought this board on ebay and am having troubles. Wondering if anyone can help.

Tried both 12V 20A and 24 volt 10A supplies.
All LEDS on board light up, and stepper motors lock.
No commands from Mach3 Jog will move any axis.
Motors are NEMA 23 5V 1,2 A. All seem good from reading the coil resistance.
I can spin the motors by using the manual pins on the 15 pin D connector.
I tried for 2 days playing with all the Mach3 settings. Timing, pin assignment, safety limit pin settings..... nothing.
The seller from whom I got the board from included a xml file of specific settings for Mach3, and a video of MY BOARD working. So as of now I'm still of thinking the board is operational, but I'm overlooking some computer/software setting.

I can devote another couple days to this before I get fed up and toss it into the bin-lol.

Thanks,
Tom
www. thelasercutter.com

[hazardouschurch]

The board is also known as HY-TB3DV-M. Try the link it is the manuel and has detailed instructions on how to set up in Mach 3. If this doesn't work i would try ubuntu with EMC2 but that is just me. Use the link in the previous post for EMC2. I found it Easier to just go with linux but then agin my drafting soft ware is on a different computer a mac with parallels and XP.

Good luck i hope you find your problem
Hazardouschurch

http://fozztexx.com/CNC/HY-TB3DV-M_3Axis_Driver.pdf

[Tom_M]

I feel like a fool, but at least a happy one.
Change the parallel port address on Mach3 and everything spins!!

[sport537]

I have my emc2 config for the tb6560 board set up like the Morgans EMC2 Config on zentoolworks. When testing each axis i receive the following error:

Stepconf encountered an error. The following information may be useful in troubleshooting:

Traceback (most recent call last):
File "/usr/bin/stepconf", line 1914, in update_axis_params
halrun.flush()
IOError: [Errno 32] Broken pipe

Any ideas on how to fix this?

thanks-KM

[hazardouschurch]

how were you testing it were you jogging the axis or running gcode?