Well done Mark it looks fabulous. Something to aspire too.

Great !! Keep on placing on PCBs!

Looks awesome, can I know the micro-stepping value in x and y axis

@markmaker4345

3 жыл бұрын

It is 16 micro-steps. More info is here: makr.zone/choosing-a-motion-controller-the-panucatt-azteeg-x5-gt-32bit/455/, config files linked towards the end of the article.

Hello! What acceleration and speed did you achieve? And what is the maximum acc/speed with the stock tinyG? My liteplacer had bad vibrations above 12 m/min in stock configuration and did not go above 35 m/min by Y and 25 m/min by X. I replaced motors with 1.8 degree ones and got 40 - 45 m/min but still with vibrations. I placed some rubber between motors and body, and vibrations diminished but not disappeared. Is it inherent property of tinyG or it can be overcome in some other way? Thanks in advance.

@markmaker4345

5 жыл бұрын

Hi I never used the tinyG. I'm using the Smoothieware Azteeg X5 GT 32bit Motion Controller and the Bigfoot Trinamic TMC2660 drivers: www.panucatt.com/azteeg_X5_GT_reprap_3d_printer_controller_p/ax5gt.htm I've achieved feed-rate 40m/min on X and Y, acceleration 5ms-2 on X and 2.5ms-2 on Y. Accelerations are not really optimized yet, just doubled/halved values until it worked. I needed to heavily tune the drivers/choppers (TMC2660) to achieve that. Without tuning I didn't even reach half of that, it stalled at ~18m/min. It sounded like vibration but I think it's just the chopper's off-time getting in the way of itself. Once I find the time, I will write up what I did for tuning here: makr.zone/ But this is highly specific to the Trinamic TMC2660 drivers. These drivers can be tuned under Smoothieware using SPI. smoothieware.org/advancedmotordriver So unless you have the same or similar driver/controller, I can't really help you. Only one thing I learned that is generic: If you can, increase the voltage. It's the voltage, not the current that sets the speed limit in these high speed applications. Also note this is still only a test run. I'm currently stuck with designing the new PCB revision that I actually want to assemble. :-) _Mark

@wormball

5 жыл бұрын

Thanks for fast answer! But youtube did not notify me. :( Why don't you use the tinyG? My accelerations are 10000 (X) and 7000 (Y) *10^6 mm/min^3. It is said that tinyG uses cubic acceleration curve but i do not know what this curve exactly is and how to convert it to m/s2. I hope this is better that constant acceleration along all the way. Here is my setup running at 48 (X) and 40 (Y) m/min: kzread.info/dash/bejne/fahhw9KOg7jAns4.html . It moves slightly faster and do not use up looking camera but has to move longer distance by Z and has longer software pauses between moves, so overall speed is approximately the same. Maybe i will sink the nozzle holder in the table as you did, so i will able to set smaller Z distance. > I needed to heavily tune the drivers/choppers (TMC2660) to achieve that. It will be interesting to read! But why didn't you try another drivers? With your board it is much easier than with mine. I suspect that my robot will move better with finer microstepping (tinyG supports only 1/8) but to try that i have to buy different board for more than $100. > Only one thing I learned that is generic: If you can, increase the voltage. Thanks, this is useful suggestion. However with tinyG i have a little room to do so as it accepts no more than 30 V. But i have another idea - use motors with lower inductance! Now i use 42BYGHW609 of 3 mH, and 42BYGHW811 has 1.8 mH so theoretically it can rotate 1.66 times faster. Loose motor can move at 60 m/min (1500 rpm). Sometimes it can 70, but sometimes it stalls. Nema23 motor (i forgotten to look part number) can only 50. Stock 17HM19-2004S can 35. Also i noticed that current setting has some optimum value (about half maximum in my case), and both decreasing and increasing potentiometer setting results in decreasing maximum speed. I always thought that if the driver can not sustain desired current, it can drive motor at lower current (so if we increase maximum current, the maximum speed at least should not drop). But it seems that my driver can't. Is this a bug of tinyG/drv8818 or your driver behaves in the same way? And do you know the cause? By the way, what nema23 motor are you using and why? I use the same 42BYGHW609 on both axes and all is ok. > Also note this is still only a test run. At my video it even does not place parts! Cos it placed all the parts before with lower speed. However most likely it will place a lot of parts soon.

@markmaker4345

5 жыл бұрын

Impressive video! > Why don't you use the tinyG? There are several reasons why I didn't chose the TinyG. 1. I absolutely wanted to use OpenPNP (for Bottom Vision) and Jason recommends against TinyG here: github.com/openpnp/openpnp/wiki/Motion-Controllers#tinyg 2. I wanted to add a fifth axis for the PCB conveyor. See here: makr.zone/planning-the-work-area/117/ 3. I thought its better to have modular drivers. And I thought the TMC2660 are much superior to the ones in the TinyG. Your video seems to demonstrate the opposite. 4. Like you said, TinyG only supports 1/8 microstepping. That's ok for 0.9° steppers but not for the 1.8° NEMA23 steppers I wanted to use. I knew TinyG has a better motion planner than Smoothieware: github.com/synthetos/TinyG/wiki/Jerk-Controlled-Motion-Explained#what-is-controlled-jerk-motion and I knew it could accelerate faster because the jerk control prevents stalling at the lowest and highest speeds, where the motor is weak, but I did not realize it also affects the max speed. I did actually look at the TinyG successor gQuintic but it is not out yet. github.com/synthetos/g2/wiki/gQuintic-Specs Your video has certainly rekindled my interest. Thanks! > By the way, what nema23 motor are you using and why? See here ("Upgrading the Y motor"): makr.zone/the-non-kit-assembly/150/ > Also i noticed that current setting has some optimum value (about half maximum in my case), and both decreasing and increasing potentiometer setting results in decreasing maximum speed. I observed the same with the Liteplacer kit 0.9° steppers but not with 1.8° steppers (the TMC2660 can control current by SPI). This was an additional reason to replace the 0.9° steppers. I thought it's just some physical limit, after all they must do double the steps per meter. > But i have another idea - use motors with lower inductance! Good point, I didn't even know they are readily available. Will also look into that. _Mark