Breakout Boards

[phil-barrett]

I started a breakout board discussion in Issues prior to the creation of github Discussions. It clearly belongs in this section. There is a lot of good info there and hopefully, it will move over here. I will close out the Issues thread and point people here.

[phil-barrett]

A question for people about breakout board features. I am designing a follow on board to the Teensy 4.1 BOB and would like to hear your feedback. The goal is to have higher end features. It will have a higher price tag, I am not sure yet how much I am hoping to keep it less than 20% higher.

Here are features that are under consideration. I would love to hear your comments.

• pluggable screw terminals
• 4 Layer board for better EMI rejection
• FRAM for Odometer logging
• 12V Isolated limit and control inputs for better EMI rejection
• Host USB support
• Isolated 0-10V output section
• support for iMXRT1062 quadrature decoding
• support for spindle sync for lathes
• I2C daughterboard mount points
• Serial IO daughterboard mount points.

In addition I am looking at several daughterboards - isolated ADC, Modbus, pendant/MPG

[mlewus]

Hi Phil, these are features I would be interested in.

From your list:

1. pluggable screw terminals - I would not pay extra for it. Perhaps you can lay out the board such that the user can select the terminal type @ assembly?
2. 4 Layer board for better EMI rejection. This will add substantial cost, and the board still must be shielded (metal box or whatever) in a high noise environment. 2 layers is OK unless you are seeing signal integrity problems in a shielded environment or need 4 layers for routing.
3. FRAM for Odometer logging. Can we mount a Cypress FM25Vxx in the existing U3 flash chip pads on the back of the Teensy? These are available from 4K to 1M bits and claim to be drop-in replacements for the SOIC-8 W25Qxx flash chips. Full disclosure, I have never used these.
4. 12V Isolated limit and control inputs for better EMI rejection. Yes, but... Adding isolated off-board power is not going to buy you much, if anything, in the way of noise immunity. It's a safety thing rather than a noise thing. Shielded and properly grounded sensor cables are still required. You may want to add a section to the manual about how to properly shield & ground cables and equipment, as it is a common source of user error. However, isolating the supplies will prevent the potential for ground loops and high voltage getting to the board and/or user, which is reason enough to do it. 12V/24V are also more common values for industrial sensors and will allow the use of standard relays, lit switches, inductive sensors, and the like. The question is cost. I'm not sure it's worth it.
5. Host USB support - Not important to me.
6. Isolated 0-10V output section. I would like to see this. Some motor drives (ie, KB Electronics KBIC DC PM drives) can be speed controlled with a DC voltage. But that voltage is referenced to earth ground. Connection to a USB-connected GRBL board will allow a ground loop which can cause signal integrity issues, and potentially damage the cable, board or computer. You don't need Hi-Pot levels of isolation here. 2x line voltage will do just fine, call it 500V. The IPC2221A spec is 2.5 mil spacing for an uncoated PCB @ 500V. Running isolated traces under the Teensy is OK as long as you can maintain spacing. The opamp should be protected from a situation where the far end of the cable is potentially disconnected from the motor drive. You might add a Shottky or ideally a TVS diode clamp to protect the opamp's output from exceeding VDD+0.5V or VSS -0.5V. Otherwise the open cable makes a great antenna :)
7. support for iMXRT1062 quadrature decoding. Not important to me.
8. support for spindle sync for lathes. If the cost is low, say <$2, then yes.
9. I2C daughterboard& Serial I/O mount points. No.

A few things not on your original list:

• The LM358 can't source closer than ~2V to the + rail with a 5k load. Even the good 12V supplies regulate no better than +-3% and most are more like +-5% (11.4V), and are typically not adjustable. A couple of solutions: (a) use the LM358 but change the voltage spec for an external voltage of 13.5V-30V rather than 12V, or (b), use an opamp that can drive closer to the positive rail, within the choices allowed by LCSC.
• Add a dedicated normally closed (open to stop) E-Stop terminal block, with the pre-built firmware configured so the board will not start without the switch installed and closed. That way users will have to go out of their way to avoid installing an E-Stop rather than the other way around.
• As mentioned by others, a pendant/MPG interface would be very welcome. Also, the ability to support a small touchscreen LCD. I suspect these are just headers that could be added by the user, so cost is minimal, assuming you have adequate free I/O pins. If I had to choose, I would take the pendant first.

Thanks for giving us the opportunity to add our voices to the next gen of this great board.

Mark

[HuubBuis]

I tried to contact you using tindle but that didn't work.

Regarding your boards who are really OK:

• Shipment to Europe is very expensive. If you have to order other parts like the teensy, than shipping cost add-up.
• Wouldn't it be possible/feasible to offer a fully assembled programmed and tested board?
• I prefer pluggable screw terminal and/or JST-XL headers
• most commercial boards use 24V DC
• Isolated inputs is safe and reduces noise
• Isolated 0-10V output section is an absolute "want to have" but also expensive. I think 8 bit will do, to reduce cost. I have plans to replace my pot for spindle speed control by a small processor bord (ESP32) to keep manual control and add a slow start feature.
• without support for spindle sync for lathes, you miss a whole market but I am not sure if it is a big one.
• Modbus and pendant are on my todo list. I see possibilities for Modbus and Mach3 communication.
• quadrature decoding for rigid tapping on a lathe also requires a suitable spindle motor. I have seen quadrature decoding used for "manual" control of the lathe and the same method could be used for a mill
• pendant/MPG is wanted by most users but I prefer them connected to the GUI.
• How about SD card and diagnostic display?

[phil-barrett]

Huub, Tindie doesn't notify me of messages to me. Sorry for the slow response. I have it turned on so am not sure what is going. I get notifications from github.

I have looked into being a Teensy distributor. The volume they want is higher than I am willing to sign up for. Though, I never thought I would sell more than 100 BOBs yet I crossed that mark a while ago.

I am looking at providing fully assembled boards. The Unkit approach appeals to a narrow segment. The minimum build numbers are a little higher than I would like but I think it makes a lot of sense. The risk is that I might have to price it too high to make it work financially.

You are not the first person to mention isolated 0-10V and I agree that it has value. So, I am looking at adding it. The idea is that it would use an optoisolator on the PWM input to the filter/amp section. The filter/amp section and the Limit+Control inputs would share a single 12V source input that is isolated. I am trying to determine if I need to change the board layout to move the two sections physically adjacent. The shared isolated 12V rail would have to run under the teensy, negating the isolation, I think. The other complicating factor is I use 12V for the relay drivers so would have to look at isolating them as well or maybe have 2 sets of 12V input. It gets messy very quickly.

I hear you on the GUI+Pendant point. That pushes the PC next to the CNC machine so it has some downsides.

SD Card is on the teensy and grblHAL has the basic elements for it. I agree it is worth doing though it appears a fair amount of code needs to be written to make it easy to use. There are file systems for it that run on the Teensy 4.1 and present a USB memory drive interface. Integration would be needed. It would be really fantastic to have a network drive that you just drop a gcode file on.

Display device. I think that's a great idea. Probably through the i2c interface as it doesn't seem to need a huge amount of bandwidth. Or, It might be possible to add an SPI header if higher throughput was needed. Perhaps SPI is better left for a future design. I could see the display providing UI for running GCode from the SD Card.

[HuubBuis]

The risk is that I might have to price it too high to make it work financially.

If you look at your competitors, you know how high you can go.

You are not the first person to mention isolated 0-10V and I agree that it has value.

That was reasoned coming from my profession where reliability goes way beyond costs. I have no idea what the extra costs are today. The main purpose is to eliminate ground loops and protect the GUI computer and connected hardware for a disaster. I currently have a "similar" protection level by using low cost WiFi or Bluetooth. Maybe you should reconsider this option regarding the costs!
For my lathe I planned to make a PWM to Analog converter, ESP32 based. An optocoupler on the input side will provide the isolation. A switch and a pot meter will provide manual settings. WiFi and/or Bluetooth for configuration, updates and maybe remote settings.

I am trying to determine if I need to change the board layout to move the two sections physically adjacent.

If you don't, you could suffer from capacitive coupling.

I hear you on the GUI+Pendant point. That pushes the PC next to the CNC machine so it has some downsides.

I prefer the GUI+Pendant. Once implemented, it works for all supported controllers. If you run without GUI, it is another story.
I use 7" and 10" tablets or a 7" touch screen for the GUI. It takes very little space and for me it works very convenient in my crowded shop.

I could see the display providing UI for running GCode from the SD Card.

There are grbl forks that do this already (3d printing) so the coding work for this is partially done. Never the less, not something you can realize in an afternoon.

There is a lot that can be done. I don't think every thing has to be ready at once. I you provide basic functionality and a grow path, that could be enough to be attractive. In the end, maintaining a lot of different hardware versions will take you more time than it took to develop.

Even if you technically make the best board, It doesn't mean that customers see the need for it!

[phil-barrett]

On isolated 0-10V. It isn't that expensive - basically another opto and some board space. I might want to use optos for the full set of spindle outputs so a bit more cost. Maybe a day or two of reorganizing the board layout. I've switched to Kicad - it seems more friendly for that sort of thing than Eagle.

If you look at your competitors, you know how high you can go.

True, to an extent. Though often times the competition is ignorance. I do have a sense of the upper price limit so it just becomes a matter of threading "the needle".

Even if you technically make the best board, It doesn't mean that customers see the need for it!

TruDat!

[terjeio]

External supply for the 0-10V is readily available? From the Huanyang VFD manual:

Two other VFDs I have manuals for and a BLDC driver also has drive voltage available so my guess is that most have? I am using that to supply a transistor based low-pass filter driven via an opto-coupler to achieve a low-cost solution. A plus is that the circuit "adjust itself" to the provided supply voltage, max. allowed beeing dependent on the filter capacitor rating.

[phil-barrett]

Interesting. Would have to swap the 358 for a rail to rail op amp.

[mlewus]

Another issue with using VFD-provided supplies is that you don't know which VFD a user will want to use. So the characteristics of the supply, how it changes with line and load, etc. are unknown. Many opportunities for problems.

I ran into this when I designed the PWM to DC converter for Sherline (KB Electronics). The KB board has an internal 12 volt supply with maybe 10mA of headroom at 120V, and I was originally going to use it. But as the line voltage drops into a low but allowable range the current headroom on that supply drops to zero. Turns out that the 12v supply is the liming factor for line regulation on that board.

[phil-barrett]

Been playing with the numbers (timing and $$). It looks like the 6N137 is reasonably cheap in China but for some reason the VOM452T is 6X the cost. Is there a significant difference between the 6N137A and 6N137? The data sheet on the non-A shows pretty decent timings - 75 nS tr, 40 nS tf, 40ish nS TPLH and 30ish nS TPHL for R_L = 1K at around a 10 mA I_f. Propagation delays all track closely together with temperature. Of course, these number come from eyeballing the DS charts. Plugging into my spreadsheet, I see 0.02% distortion at 5 kHz, 0.04 % at 10 kHz and 0.22% at 50 kHz, 50% DC. Good enough, especially since the grbl default is 5K. About $0.20 in volume.

On the issue of load, I love the idea of using the VFD's 10V output. However, Mark brings up a good point about relying on an unknown and potentially unreliable source. I think I will move to a rail to rail opamp (couple of ok and cheap jelly beans out there). and document what VFDs it does (and doesn't, eventually) work with. I will crank the numbers for the opamp circuit and simulate it as well just to make sure but I'm guessing that it will come in at less than a 10 mA total load. The user can go with a separate supply if their VFD doesn't cut it.

edit: fixed 6N137 typo.

[mlewus]

The R-R is a great solution. If the on-board VFD supply doesn't work, it's easy enough to use an external wall wart or whatever.

I tried to look up the 6N127/6N127A datasheets and came up with nothing. Can you post a link?

[phil-barrett]

sorry, that was a typo. 6n137.

[mlewus]

The A version has a few tighter specs, nothing serious. Input capacitance 28p vs 60p which will affect how fast the driver can switch it. Withstand voltage is 2500V compared to 5000 V. A few other minor things. It’s not recommended for new designs which may be why it’s so cheap. Looks ok

[frdfsnlght]

Seeing as how microcontrollers never have enough I/O pins, has anyone considered adding I2C shift registers to handle "low frequency" outputs? I imagine one or more chained shift registers at a single I2C address to handle things like motor enables, motor directions, relay outputs, that sort of thing. The higher frequency stuff like motor step and PWM could still be directly tied to controller pins. Imagine all the low frequency outputs on shift registers and how many pins that could free up?

I don't know if anything like that is possible for the inputs.

[terjeio]

Some drivers already has optional code support for an I2C based expander (for outputs only), SAMD21 (MKRZERO) is one. I am not too keen to use any for inputs though due to additional overhead for interrupt handling. Shift registers may also be used, bit banged or via I2S if an I2S peripheral is available.

[frdfsnlght]

I just found this chip: MCP23017. 16 pin I/O expander addressable over I2C. Seems like it could open a lot of things up.

[mlewus]

Microchip makes a number of these expanders in 8 and 16 bit widths, for I2C and SPI. I’ve used them for many years. They are fast (1.7MHz I2C, 10MHz SPI), robust, and work well. Complexities creep in because you probably want every I/O to be optically isolated and outputs should be able to drive relatively high voltage, high current loads like relays, so you need that target application information at design time.

[mlewus]

I think you should look at the overall topic of accessories like pendants, displays, keyboards, and now extra IO. It is difficult to predict what an individual user will want. One solution would be to add 2.54mm header(s) for accessory boards to be mounted either on or off-board. The RTX1062 allows comm devices to be more or less freely mapped to I/O pins. If you bring out 4 or 5 I/O pins plus power, it could support I2C or SPI Interfaces plus address selects. Alternatively if you have the I/O available (12 I/O + 3.3/5V), you could bring out a Mikroe Bus header. That would still allow I2C or SPI but adds the range of over 100 Mikroe Click boards. These support everything from relays to expanded IO to industrial comm interfaces to motor control.

Either way, a wide array of accessories could be supported at the added mfg cost of only a header.

[mlewus]

Of course...

[mjscosta]

Hi @phil-barrett, I'm not a typical user, but at the moment I have a couple of interests in grblHal, one "cheap" alike Eleksmaker boards with support for onboard two sockets for laser engraving being able to have max possible speed for using on laser engraving and with support for jobs from sdcard, to avoid bottleneck with timing on the sender.

Another, I'd want to have 6 stepper platform to play with, nothing specific, maybe built a DIY 3D printed robot arm or any other stuff alike, I'm going to check if I can get grblHal into Bigtreetech SKR Pro v1.1 ...

On my opinion, most users would prefer a board with preinstalled software with no need for compiling and going through the code ...

[mjscosta]

Also for sure there's a marked for the plasma cutters, consider using an MPCNC, it's just incredible what we can do nowadays on a DIY CNC thanks to the OSS community!!

[terjeio]

I'm going to check if I can get grblHal into Bigtreetech SKR Pro v1.1

The STM32F4xx driver should be extended for that - a board map file and setting up the clock tree is what is needed for basic functionality.

I am currently working on a CNC breakout board for ST Nucleo dev. boards using this driver, adding support for Trinamic drivers to it is part of that. Ethernet support is also planned.

[mjscosta]

Thanks for the hint.

[frdfsnlght]

How about more limit switch inputs? I'd like to see separate min/max inputs for each axis, so Xmin, Xmax, Ymin, Ymax, Zmin, Zmax, Amin, Amax, Bmin, Bmax. After another discussion elsewhere about autosquaring, some things work better with more of these inputs.

Just noticed you also have a discussion about input pins (#140). Maybe this suggestion belongs there.

[mlewus]

If you do another version of the board have you considered using Kicad instead of Eagle? While Eagle has a huge user base in the hobbyist world, I’m concerned it will go the way of Fusion 360 and get even more restrictive over time. I use a program called DipTrace for most of my stuff, but that one is not free. I also use Kicad and it has improved quite a bit over the past couple years. And it is open source (gpl v3).

[phil-barrett]

way ahead of you. The Pro version is done with Kicad. Made the switch about 6 months ago.

[mlewus]

I watched an interesting video on YouTube last night where one of the major contributors of kicad was discussing its future. They really seem to be going after the big guys. It's here if you want to take a look: https://m.youtube.com/watch?v=6qcsJfoqbL4&t=2263s

…

On Sun, Dec 20, 2020, 22:24 phil-barrett ***@***.***> wrote: way ahead of you. The Pro version is done with Kicad. Made the switch about 6 months ago. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#130 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABKP5UIFIPHN5OZC4IENORTSV25YRANCNFSM4UUANSKA> .

[raenji-sk]

• 12V Isolated limit and control inputs for better EMI rejection
  I would welcome this, pretty much only thing holding me back is that most of theese boards do not support the powered inductive limit sensors, I think most can be powered by 10-30V. I do use some now with my grbl Arduino setup on 12V, plain ordinary wires and not one hickup in 2 years of usage. Had a ton of problems with the microswitches...

• pendant/MPG support - definitely yes if possible, with continous and step modes

[mlewus]

You can already use inductive sensors with Teensy 4.1 BOB as long as you use the ones with NPN open collector outputs. They would be powered from an external 12-24V source whose ground is common with ground on the BOB. The open collector pin connects to the resistor side of the input (TR1 - TR11). Just be sure you are using sensors with open collector outputs. Totem pole outputs may damage the sensor or breakout board.

[phil-barrett]

Yes, if you have a schematic or manual it should say. You can also test your sensor by hooking up +12 and Gnd. If the Signal to Gnd measures 12V then it is a Totem Pole (or Pulled High internally) and can't be used with the Teensy 4.1 BOB. I have several inexpensive Chinese Inductive Sensors and they are all totem pole.

[mlewus]

The ones you want are LJ12A3-4-Z/BX. These are 6-36V, open collector NPN output, 3 wires. You can get them from China for a couple dollars each. These are normally open but you can also get them normally closed. Avoid PNP!

[avizienis]

A bit off topic question. All good breakout boards seem to include some sort of output buffers for step/dir/en pins. I wonder how much of a necessity is this. I mean for external drivers each output is about 10mA of current to drive the opto. So 3 outputs per axis, times three for xyz and we are already in the 90mA peak output territory, which should strain most microcontrollers, and should kill the likes of Atmega328. However there are plenty of machines with drivers hooked up directly to arduino and they seem to be doing fine. What gives?

[phil-barrett]

From the 328P datasheet:

In addition, The above numbers are for steady state. When you factor in duty cycle, the numbers get a lot less daunting.

[avizienis]

Ok, makes sense. So why did you include 74HCT541 in your board? Lev shift to 5V?

[phil-barrett]

Yes, mainly for level shifting in a really low cost and compact package.

[mlewus]

Another advantage of the HCT series is that it’s TTL level compatible on the input, but it will drive its outputs to a minimum of 3.8V compared to 2.4V for bipolar (LS, ALS). This improves the noise margin when driving a TTL level compatible step controller.

[mlewus]

As I recall the atmega328 is rated for something like 40 milliamps per pin. There is also a package (thermal) limit of maybe 200mA. The bigger issue is that microcontrollers are not well protected against surges, spikes, and the like. These can cause damage but more likely will cause the controller to behave erratically and lose program control, what we used to call "falling off the track", until the watchdog timer kicks in or you reset the chip. Non-isolated systems also suffer from ground loops and other noise problems. But they do work well enough for some hobby uses, and the price is right.

…

On Fri, Dec 25, 2020, 17:34 avizienis ***@***.***> wrote: A bit off topic question. All good breakout boards seem to include some sort of output buffers for step/dir/en pins. I wonder how much of a necessity is this. I mean for external drivers each output is about 10mA of current to drive the opto. So 3 outputs per axis, times three for xyz and we are already in the 90mA peak output territory, which should strain most microcontrollers, and should kill the likes of Atmega328. However there are plenty of machines with drivers hooked up directly to arduino and they seem to be doing fine. What gives? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#130 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABKP5UL6L2XSLKXBGUCPZSLSWUHQHANCNFSM4UUANSKA> .

[timgow]

Hi Phil,
I'm building T41E5XBB v2.05 for my lathe.
Is spindle synchronisation available now or is further coding required?
Tim

[timgow]

Hi Terjeio,
Thank you for clarifying with regard to the index pulse, any aux pin.
Please let me understand why ST3 is not suitable for the multiple pulse per rev signal as standard, it's to ensure that it's an interrupt capable pin perhaps? I'm not understanding why the 1pulse per rev index is treated differently to the (in my case) 60 pulse per rev signal.

[terjeio]

I'm not understanding why the 1pulse per rev index is treated differently to the (in my case) 60 pulse per rev signal.

The latter is fed to a counter clock input so that it can be prescaled for high PPR rates in order to reduce interrupt load. The number of inputs that can be routed to a counter clock input is limited and is the reason for why the board has to be modified.

[timgow]

ok, i get the plot now, much appreciated. Kind of looking forward to the pcb surgery.

[phil-barrett]

Tim, I assume you have a V2.07 board. I will post a separate entry in Discussions.

[timgow]

Hi Phil, yes v2.07, much appreciated.

[mundsen]

This is great work :-) I have the v2.09 board and plan to get a pro board - better screw terminals..
I`m going to order a new Teensy 4.1 for the pro board - should I order a FRAM module to go with it? if so, what kind/model?

In the driver forlder there is a i2c keypad folder - is there a supported keypad/pendant?

[terjeio]

I`m going to order a new Teensy 4.1 for the pro board - should I order a FRAM module to go with it? if so, what kind/model?

I believe the Pro board has FRAM mounted as standard.

In the driver forlder there is a i2c keypad folder - is there a supported keypad/pendant?

AFAIK only my keypad code which is used in the MPG&DRO I made for my lathe.

I have HALified the MPG&DRO code and added a driver for the RP2040 - but it still uses a MSP430 for the keypad handling due to lack of pins. I want to expand the MPG&DRO code so that the main MCU can program the MSP430 from a binary blob...

[mundsen]

Do you have a spare pcb for the MPG&DRO project? possible to buy?

[terjeio]

I do not have any spares but are considering making a few PCBs for the Pi Pico.

[mundsen]

I use 12V for my relays - is it possible to "trigger" the Aux1 or Aux2 when Feed/Hold is active?

I plan to use buttons for Cycle/Start and Feed/Hold with integrated led ring (Green LED and Red Led 12V)

If so I can let the Green led/Cycle/Start button stay on in normal operation and the Red Feed/Hold stay off

If in Feed/Hold mode swap, so green is off and red on

Not important/critical - only an idea if it is simple to set in config..?

[terjeio]

I use 12V for my relays - is it possible to "trigger" the Aux1 or Aux2 when Feed/Hold is active?

Yes, with a small plugin that subscribes to the grbl.on_state_change event and sets the output via call to hal.port.digital_out.

grbl.on_state_change example can be found in this plugin and hal.port.digital_out in this.

[Klippert]

Just out of curiosity, would such a plug-in live in the new Garbo-Hal repository or in its own repo?

[terjeio]

I don't know - maybe in an odds & ends subrepo in the new grblHAL repo (as a link to your repo?), or just a snippet you keep for yourself? It would depend on how it could be beneficial to other users.

I have added a way to add private code without making any changes to the master code for use-cases like this.

[terjeio]

I'll add this snippet to the templates repo later:

/*

  my_plugin.c - plugin template for setting auxillary output on feed hold

  Part of grblHAL

  Public domain.

*/

#include <string.h>

#include "driver.h"
#include "grbl/protocol.h"

static uint8_t port;

static on_state_change_ptr on_state_change;
static on_report_options_ptr on_report_options;

static void onStateChanged (sys_state_t state)
{
    static sys_state_t last_state = STATE_IDLE;

    if(state != last_state) {
        last_state = state;
        hal.port.digital_out(port, state == STATE_HOLD);
    }

    if(on_state_change)         // Call previous function in the chain.
        on_state_change(state);
}

static void onReportOptions (bool newopt)
{
    on_report_options(newopt);  // Call previous function in the chain.

    if(!newopt)                 // Add info about us to the $I report.
        hal.stream.write("[PLUGIN:MY PLUGIN Template 2]" ASCII_EOL);
}

static void output_warning (uint_fast16_t state)
{
    report_message("An output port is required for my_plugin!", Message_Warning);
}

// Tell the user about which port is used for the output
static void output_port (uint_fast16_t state)
{
    char msg[30];

    strcpy(msg, "My plugin port: ");
    strcat(msg, uitoa(port));

    report_message(msg, Message_Info);
}

void my_plugin_init()
{
    if(hal.port.num_digital_out == 0)                   // This plugin requires one digital output port,
        protocol_enqueue_rt_command(output_warning);    // complain if not available.

    else {
        port = hal.port.num_digital_out - 1;            // Claim the
        hal.port.num_digital_out--;                     // last free port.

        on_state_change = grbl.on_state_change;         // Subscribe to the state changed event by saving away the original
        grbl.on_state_change = onStateChanged;          // function pointer and adding ours to the chain.

        on_report_options = grbl.on_report_options;     // Add our plugin to to the options report chain
        grbl.on_report_options = onReportOptions;       // to tell the user we are active.

        protocol_enqueue_rt_command(output_port);       // Tell the user about which port is used for the output after startup is complete.
    }
}

Create my_template.c in the same folder as driver.c and copy the code above, recompile and upload.

[Klippert]

Hello, I do the same Thing. That is use buttons With led rings. I found some 5v buttons That Can actually be wired so they go on and off With the input power. So when it’s “High” the led is on and the Burton is pressed. Through That only works With latching buttons and for things like door and vac override. The rest I’m in the same situation as you :) Ie. It’s be nice to have the green ring blinking when the program is running or something :) Best Kristoffer lør. 10. apr. 2021 kl. 13.30 skrev Oyvind Amundsen ***@***.***

…

: I use 12V for my relays - is it possible to "trigger" the Aux1 or Aux2 when Feed/Hold is active? I plan to use buttons for Cycle/Start and Feed/Hold with integrated led ring (Green LED and Red Led 12V) If so I can let the Green led/Cycle/Start button stay on in normal operation and the Red Feed/Hold stay off If in Feed/Hold mode swap, so green is off and red on Not important/critical - only an idea if it is simple to set in config..? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#130 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AM3JUEAPTU4VTD4WACQ6LXTTIAZGPANCNFSM4UUANSKA> .

[phil-barrett]

I don't think it is currently possible to selectively light the button LEDs from grblHAL. However, it is a very interesting idea that dovetails with my thinking about general indicators. I just posted a discussion on that point.

[mundsen]

Testing my controller setup - Love the grblHAL + the Teensy 4.1 breakout board
Controller: grblHAL Teensy 4.1
Stepper drivers: MKS TMC2160
Laser Opt Lasers 6W + AirAssist
12V relay module
Power Supply: 24V (stepper drivers + valve ), 12V (relays) and Opt Laser power
24V solenoide valve for AirAssist control
Automatic exhaust fan (the fan in the video is for test only)
In this test I have 2 layers in LightBurn (The LED display show the laser power)
1 - 50% power and AirAssist Off
2 - 80% power and AirAssist On (red led in valve goes on)
The exhaust fan goes on when the laser is on + I have a switch I can use to manually set it On
All working fine 🙂
Im working on a wireing diagram + Ill share my firmware/LightBurn settings when I`m done

video-1618065958.mp4

[phil-barrett]

No, it is probably ok. Your relay board uses optos for input so it will work with OC outputs with no pull ups. That is a nice relay board, I may get a couple.

[mundsen]

V1 ready

[mlewus]

I really like the air assist setup!

[mundsen]

Hope this can help/inspire other users :-)

[mundsen]

I did not include it in the diagram, but I recommend using a water separator/filter (if you use a air compressor)

[ajtalbot1]

Looks amazing!

…

On Sun, Apr 11, 2021 at 1:55 AM Oyvind Amundsen ***@***.***> wrote: [image: HAL_Setup] <https://user-images.githubusercontent.com/3058577/114297995-52ddf200-9ab4-11eb-977a-05830d5346e8.png> V1 ready — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#130 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AG64EGXIMYSQPQ4ZC37ZD2LTIFPZHANCNFSM4UUANSKA> .

[phil-barrett]

I agree. Really well done.

[Klippert]

I see it as something That there might be quite a few People who Would like to implement. But i might be completely wrong. Through i Would prefer to ADD it to the same repo, as it’s pretty useless on its own ;) man. 12. apr. 2021 kl. 10.00 skrev Terje Io ***@***.***>:

…

I don't know - maybe in an odds & ends subrepo in the new grblHAL repo (as a link to your repo?), or just a snippet you keep for yourself? It would depend on how it could be beneficial to other users. I have added a way to add private code <https://github.com/grblHAL/Templates/blob/master/my_plugin.c> without making any changes to the master code for use-cases like this. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#130 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AM3JUEDHKP4GFMPL3MG2RK3TIKSCNANCNFSM4UUANSKA> .

[terjeio]

Through i Would prefer to ADD it to the same repo, as it’s pretty useless on its own ;)

I am not sure I want to host all kinds of plugins in the grblHAL repo as it will lead to the grblHAL team taking over maintenance and support? Better to add a short description and link to the authors repo instead?

IMO your idea could be a good candidate to add as a template for others to build upon rather as adding it as a standard plugin.

[raenji-sk]

Hi, any guides on connecting the relay outputs on the Teensy 4.1 BOB? I mean which g/m code activates which output from the ones on top of the board (spindle/mist/vac/aux...)
Tried wiring up a few today, mist is working with the m7/m9 command, but I can't figure out where to wire the spindle. Tried both spindle pins and none reacts to m3/m5 command.
I'd like to wire the following, compressor, spindle, mist solenoid, vacuum, simple on/off state for now for all. But I am struggling to find any documentation which would explain how to do it, specifically the commands to control individual outputs to either add a start/stop macro to the gcode sender (i use bCNC at the moment) or to modify the postprocessor.
Furthermore, is it possible to wire the control for spindle rpm (vfd type, probably the 0-10V output would do) and laser at the same time and switch between them?

[phil-barrett]

No that firmware is fairly old now. Are you set up to build grblHAL yourself?

[raenji-sk]

Alright, I will try to build a new version then.
I should have everything needed as I do make some boards of my own, mostly atmega based thoug, but have used the older teensy models, 3.1\LC, in some projects too.
I've seen some guides, or links to the guides in your repo Phil, are they still OK and usable?
If you have some tips and tricks on how to make it as fast and painnless I'd certainly appreciate it - which files to get, what to configure etc. to get a base firmware with stock settings linke the older ones in your repository for example.

[mundsen]

I use it this way:

1. Install arduino software + teensyduino + github Desktop
2. Open github Desktop and add this url: https://github.com/grblHAL/iMXRT1062
3. Download repository to local disk (modify settings..)
4. open the ino file in folder examples/grblHAL_Teensy4_Upload/ (in arduino)

[raenji-sk]

Thank you @mundsen. It does compile correctly with the default settings. Will try to adjust some of the settings, recompile and flash it to Teensy.

[raenji-sk]

Thank you all for your help. Latest version flashed on Teensy and all works good. Mist relay is good, spindle is now working too.
I tested it with M3 and no S command before, so my mistake as the speed parameter needs to be set for it to work.

[mundsen]

A stupid question (I guess) - I have a 4 axis setup. X,Y,Z and A (rotary) - where/how do I enable the A axis?
XYZ tested ok using io GCode sender and LightBurn

[mundsen]

I set N_AXIS 4
in my_machine.h #define BOARD_T41U5XBB

But I get obly XYZ in IO sender

In Lightburn I get an error if I try to test the rotary (A-axis)
error:20
Unsupported or invalid g-code command found in block.
On or near line 2:
Stream completed in 0:00

XYZ axis run ok

[terjeio]

You have to set in in grbl/config.h as shown above, line 35. Remove the comment characters // when setting it to 4.

[mundsen]

Have done that - in config it says

#define N_AXIS 4 // Number of axes

[terjeio]

Strange, what is the $I response?

[mundsen]

error:20
Unsupported or invalid g-code command found in block.
On or near line 2:
Stream completed in 0:00
Rotary setting update successful
Starting stream
error:20
Unsupported or invalid g-code command found in block.
On or near line 0:
error:20
Unsupported or invalid g-code command found in block.
On or near line 0:
error:20
Unsupported or invalid g-code command found in block.
On or near line 0:
error:20
Unsupported or invalid g-code command found in block.
On or near line 0:
Stream completed in 0:00
Rotary setting update successful
$0=5.0
$1=25
$2=0
$3=0
$4=1
$5=0
$6=0
$7=0
$10=511
$11=0.010
$12=0.002
$13=0
$14=70
$15=0
$16=0
$17=0
$18=0
$19=0
$20=0
$21=0
$22=0
$23=0
$24=25.0
$25=500.0
$26=250
$27=1.000
$28=0.100
$29=0.0
$30=1000.000
$31=0.000
$32=1
$33=5000.0
$34=0.0
$35=0.0
$36=100.0
$37=0
$39=1
$40=0
$43=1
$44=4
$45=3
$46=0
$62=0
$63=3
$64=0
$65=0
$80=1.000
$81=0.010
$82=0.000
$84=0.000
$85=10.000
$90=0.000
$91=0.000
$92=0.000
$95=0.000
$100=5000.000
$101=500.000
$102=250.000
$110=500.000
$111=500.000
$112=500.000
$120=10.000
$121=10.000
$122=10.000
$130=200.000
$131=200.000
$132=200.000
$341=0
$342=30.0
$343=25.0
$344=200.0
$345=100.0
$347=5.0
$348=2.500
$349=25.000
$370=0
$372=0
ok
$I
[VER:1.1f.20210313:]
[OPT:VNMSL,35,1024,3,0]
Target buffer size found
[NEWOPT:ENUMS,RT+,ES,TC]
[FIRMWARE:grblHAL]
[NVS STORAGE:*FLASH]
[DRIVER:iMXRT1062]
[DRIVER VERSION:210313]
[DRIVER OPTIONS:USB.2]
[BOARD:T41U5XBB]
ok

[mundsen]

If I open Machine Settings in LightBurn I get A, but only $2,$3,$23 parameters, not $102, $112, $122 and $132 as I get for XYZ

[terjeio]

Are you sure you have changed the file that is used by the compiler?

Here is a_Help > About_ dialog from ioSender when 4 axes is compiled in:

The red circle highlights number of axes - extracted from the $I report.

If I open Machine Settings in LightBurn I get A...

Do you manually set up Lightburn for 4 axes or does it configure itself from the $I repsonse?

[mundsen]

Mine is 3 not 4, so I guess I do something wrong

I run the ino file in the D:\grblHAL\iMXRT1062\grblHAL_Teensy4\examples\grblHAL_Teensy4_Upload folder

I guess that does not compile and include altered settings?

[mundsen]

I was the stupid one!!
Did edit files in the folder downlaoded by git desktop, not the files in the my documents\arduino.. folder

Lesson learned :-)

[mundsen]

All 4 axis ok 🥰

[raenji-sk]

Tried to connect some more stuff to the Teensy BOB and everything works great, updated the Fusion 360 postprocessor to controll the coolant options and also Aux0-2 ports. Here's a quick overview of the setup if someone want's a refference usecase of the Teensy and grblHAL :) https://www.youtube.com/watch?v=YX9hMMTAbhs

[mlewus]

Very nice setup you have. Is there a reason that you're using isopropyl alcohol rather than a water-based coolant for aluminum? I've been using wd40 applied manually, but it is labor intensive and means I can't leave the machine running.

[raenji-sk]

Thank you. I have quite a lot of isopropyl from my other hobby, resin 3D printing. So that's why. Also companies like datron use this with their high end super fast aluminum milling machines, so if it works for them I gave it a go.
It leaves no residue on part or tool and evaporates quickly taking the heat with it.
I would not use it with steel or stainless though as an occasional spark might be pretty devastating...
Did not have time to reaserch other cutting fluids, and also the mdf spoilboard would probably not take any moisture well. So for now I use isopropyl for some cuts, but even air works pretty good most of the time.

[phil-barrett]

Update on the Teensy 4.1 Pro. I have received my first shipment of assembled "unkit" board. Several issues:

• Assembly kit components are in short supply
• Because of that, the cost of components is higher than I had expected.
• FRAM is sold out world wide except for some very questionable sellers so no FRAM.
• I am heading out the door for a 4 week holiday and decided to hold first shipment until I return. I want to give good support but can't while traveling. sorry.
• There is a problem with 0-10V. It works in isolation but hooked up to a VFD runs higher than normal - 70% and above runs at full speed. I have some ideas but no time to debug.

[mlewus]

Hi Phil, regrettably you chose quite the crappy year to try to launch a new product. What is the price of the new board?

…

On Tue, Aug 10, 2021, 11:31 PM phil-barrett ***@***.***> wrote: Update on the Teensy 4.1 Pro. I have received my first shipment of assembled "unkit" board. Several issues: - Assembly kit components are in short supply - Because of that, the cost of components is higher than I had expected. - FRAM is sold out world wide except for some very questionable sellers so no FRAM. - I am heading out the door for a 4 week holiday and decided to hold first shipment until I return. I want to give good support but can't while traveling. sorry. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#130 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABKP5UJIY2FYJLUJ45LELUDT4HVIRANCNFSM4UUANSKA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

[phil-barrett]

Yeah, the supply chains are a lot more fragile than I thought. Good thing I'm not doing this so I can eat...

The board itself without FRAM is 35.99.
With FRAM (assuming a return to pre-covid pricing), 39.99. When FRAM actually becomes available, I will probably retire the no-FRAM version.

The USB Assembly kit is $28
USB+Ethernet Assembly kit is $40.

For what it's worth, I will probably end-of-life the original board and offer a 4 Axis, no digital input BOB for maybe $24. Assembly kits would be similar to what is offered now.

[mlewus]

Thanks for the reply. Have a great vacation

…

On Wed, Aug 11, 2021, 12:47 PM phil-barrett ***@***.***> wrote: Yeah, the supply chains are a lot more fragile than I thought. Good thing I'm not doing this so I can eat... The board itself without FRAM is 35.99. With FRAM (assuming a return to pre-covid pricing), 39.99. When FRAM actually becomes available, I will probably retire the no-FRAM version. The USB Assembly kit is $28 USB+Ethernet Assembly kit is $40. For what it's worth, I will probably end-of-life the original board and offer a 4 Axis, no digital input BOB for maybe $24. Assembly kits would be similar to what is offered now. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#130 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABKP5UJVF6GJOH6XUDEEZKLT4KSSLANCNFSM4UUANSKA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

[frdfsnlght]

So, is vacation over? How was it?

It doesn't look like the new board is on Tindie yet? Where/when I can get hold of one of these? I have a spare FRAM chip (assuming it's the same one used by your previous board). I'm interested in the USB/ethernet unkit version.