Do they need to be spaced for keying? Is there a standard aux pin layout in the cnc arena?
There is no standard. I wouldnāt bother keying them, just a row of pins so that
people can plug in jumper wires instead of having to solder
1 Like
expanding on my earlier request for stepper motor support.
The 3D printer world has created a standard stepper driver module (2x8 pin module on 0.1" centers) that can take one of several modules
https://www.aliexpress.com/item/CHANGTA-50pcs-lot-A4988-stepper-motor-driver-Module-with-Heatsink-For-3d-printer-mainboard/32810150134.html
https://www.aliexpress.com/item/5pcs-lot-3D-Printer-Stepstick-Drv8825-Stepper-Motor-Driver-Reprap-4-PCB-Board-Free-shipping-replace/32514641632.html
https://www.aliexpress.com/item/5X-MKS-TMC2130-V1-1-For-SPI-Function-Stepstick-Stepper-Motor-Driver-with-Heat-Sink-Ultra/32850690868.html
all can be driven with 3 pins (enable, step, direction)
The last version also has a SPI interface so you can get feedback to know when it has failed to make a step (like when it has hit the limit of how far it can go)
If we could afford the board space to have a mount for one of these, it would make it very easy to use the aftermarket Z axis kits that people are buying instead of trying to motorize their router adjustment
1 Like
I donāt think itās worth the effort to have ground and 5v next to each aux pin.
There are times when itās handy, but I would not bother (especially not if they
have to be packed this tightly)
Most of the things that are going to use the aux pins are not going to use both,
and many will not use either.
I wonder whether there is enough compute time available to accurately drive a stepper. We are already running into situations where a calculation takes longer than one ātime slotā to complete, and things like motion planning will add to that overhead. Steppers need accurately timed and counted drive pulses. How much overhead does that introduce?
1 Like
remember that we are already doing calculations to count the encoder steps,
running a stepper is not likely to be that much worse.
also remember that we are talking about the Z axis, not the main X/Y axis.
A stepper that isnāt moving requires zero CPU, a DC/encoder/PID loop
needs continual attention, even to stay in the same place.
The software flags a stopped axis as inactive, so the PID detaches if the axis is idle. The encoder pulses are counted in an interrupt service routine written in inline assembler for speed. Very little time is spent there.
Looking at the servo boards, it looks like the reset/enable is usually tied high, so the board needs 5V, ground, step and direction. That could be driven off the present Maslow board using the pins of two AUX ports, with their 5V and ground. If the software could be shown to work and plays well with others, that would probably be the time to look at accommodating a servo board and making provision for the additional motor current and motor connectors. It could be that cabling the whole AUX group onto a separate board would be the way to handle all that.
4 Likes
I have been looking to modify the existing motor shield to include a breakout for a HC-05/06 module and to power the arduino from the 12V connection. This looks like a better board to base my work on.
Would this be of interest to anyone else?
what is a hc-05 module?
David Lang
k
I started down that road - but found it better to put the $$ in a Pi3.
PM me if you want my dev notes.
Thank you
Sorry I forgot to mention that, yes itās a Bluetooth module.
That would be great if you could send me that info. Doesnāt seem like I can pm you yet.
@gmihovics Welcome to the group BTW.
Here are some URLs:
I was going to make it a stackable shield.
https://arduino-info.wikispaces.com/BlueTooth-HC05-HC06-Modules-How-To
Brother Chris is a blast - The project he is talking about uses a Mega Arduino as a controller too.
Edit: Iām sure you could get these modules by China Post for less but amazon is $9 -$13 when I looked.
Thank you
1 Like
Hello @gmihovics, welcome!
Having access to all those XIO pins of the Mega opens up lots of possibilities. Here are some thoughts.
- There is a compile-time option in the Maslow firmware to run in emulator mode, without the need for a PowerControl board or motors. That way one can run tests on software changes quickly without having to worry a bout unexpected things like running the motors to destruction (voice of experience). With that and a low-cost Mega you can develop anywhere, even before youāve got a Maslow on hand. The GroundControl software behaves just as if you were making mountains of wood chips.
- The software is timing-sensitive, and the position tracking is interrupt-based, so software serial or interrupt-driven serial drivers might need some crafting to get along with it.
- There are ways to damage a Mega that is running on the Vin supply when plugging/unplugging the USB connection (again, v.o.e.), just be careful.
- The gcode stream over the serial connection has no error checking/correction, so reliability of that stream is crucial.
- Many routers are electrically noisy, and the chain-drive motors probably are as well.
Keep us posted as you look into Bluetooth, I know there are others here who are interested in getting that going as well.
2 Likes
Thanks @Bee, this was exactly what I was thinking. I had planned on making it a stack-able shield as well but found I needed a new power source since I was not using the usb connection anymore.
@blurfl, you raise many of the same concerns I have doing this too. The plan was to tap directly into the USB TX0 and RX0 which should not interfere with the timing in the software. The interference and reliability of the stream is something I am concerned about but would have to see when everything was done.
Might have to re-think this for a bit.
Thanks, everyone.
1 Like
I looked into this about 8-10 months ago regarding developing an iOS version of Ground Control.
In the end wireless really is ill-suited for the GC to Mega connection.
Errors in this critical link mean lost workpieces.
I prefer (but havenāt yet gotten to) the Raspberry Pi.
Placing the potential source of error that is wireless communications at ādisplaying the machineā vs āoperating the machineā is a better option.
3 Likes
This is why I use Ubuntu Mate on the Pi 3 connected locally to the Maslow and remote to that desktop. I maintain a reliable local connection. If the remote connection were to drop Iād loose vision. It has been rock solid for me so far. I have worked through the New Pi3B+ hardware and have Ubuntu Mate running and am doing testing to see itās as solid as the Pi3.
Thank you
2 Likes
I was hoping to avoid networking more machines in the garage but it might just be the most straight forward way to go.
The problem with radio based communication is it is susceptible to interference at anytime and the router itās self could be a source of that interference. It is an uncontrollable condition in most installations and therefore unreliable. My layout allows for radio failure and the Maslow will keep doing itās thing. At the moment the Maslow relies on a non interrupted line of communication or you are FUBAR. This is how I came to remote desktop as a solution. You can outfit a Pi3 for less than $60 as a headless solution. The you can use a 286 pc running a remote desktop client. This allows some pretty old hardware that predates the internet to play in the game. If we add some command line options you could telnet in. I only suggest any of this over a local net not from outside BTW. I have several internal local netās that are not part of the internet. I would not Telnet from the WWW as it is no longer safe.
Thank you