Mini Stealthburner, Sequins, and Picobilical

[claudermilk]

I have a set of the RGB sequins for the Mini SB on my new LDO V0.2 and am thinking about how to hook them up. I've found the community how-to for hooking them to the X endstop port and may end up doing that. But, it would be nice to be able to control them as RGB. So I'm looking at the frame PCB schematics from LDO's Github and it looks like the unused filament sensor port might be usable. It shows 3.3V, ground, and two signal pins. Since the X endstop route works and it has 3.3V that should be ok. The catch is the circuit shows a 10 ohm resistor instead of the 100 on the RGB ports. I'm thinking I might need to insert a 100 ohm resistor in the signal wire to make the colors come out right. Am I on the right path here, or is this a dumb idea? I'm still a total noob with electronics, so I really don't know if this would work or if I'd release any magic smoke and/or angry pixies trying this.

As for "losing" the use of the filament sensor there, I have the spec foot-mounted one installed and working.

 

[brystro]

Hi, I was thinking the same thing (that way I don't have to run longer wire from the toolhead to the back of the printer. Have you tried this?

 

[claudermilk]

I haven't done any modding of my printers since I posted. Now that the holidays are over and my gift manufacturing is done, I can start thinking about breaking things again.

 

[Yuuan]

I haven't done any modding of my printers since I posted. Now that the holidays are over and my gift manufacturing is done, I can start thinking about breaking things again.

Did you ever figure anything out? I also just gotten the LDO V0.2 and I have some RGB sequins but I am also sort of lost on how to hook them up.

 

[claudermilk]

I i still haven't touched that. Other projects have back-burnered that one for now. One of those projects is telling me I "need' to ditch the min SB and swap the tool head to a Dragon Burner and Galileo2 extruder.

 

[majic79]

I've gone for an EBB36 on my mini-SB, that has a few spare parts and could probably drive those - might need to get some sequins and do a guide

 

[NoGuru]

I i still haven't touched that. Other projects have back-burnered that one for now. One of those projects is telling me I "need' to ditch the min SB and swap the tool head to a Dragon Burner and Galileo2 extruder.

Yeah I did the change to Dragon Burner and I love it. Simple design that just works.

 

[claudermilk]

Yeah I did the change to Dragon Burner and I love it. Simple design that just works.

I just put parts on order. Fabreeko finally got their Honeybadger 4010 blowers in stock again. West3D has the Galileo2 kit in stock, so now all I have to do is wait on that hardware to show up, and print parts.

 

[NoGuru]

I just put parts on order. Fabreeko finally got their Honeybadger 4010 blowers in stock again. West3D has the Galileo2 kit in stock, so now all I have to do is wait on that hardware to show up, and print parts.

Yeah I ordered the Galileo2 from Ali and 2 days later West3D had them in stock, oh well, gotta wait on the slow boat.

 

[claudermilk]

Mine's already shipped. I should see it in a few days. Of course, I've been waiting a while to see both the fans and kit in stock at the 3D vendors, but I've been in no big hurry; other projects are keeping me busy, and work went insane last month.

 

[ChrisA]

I have a set of the RGB sequins for the Mini SB on my new LDO V0.2 and am thinking about how to hook them up. I've found the community how-to for hooking them to the X endstop port and may end up doing that. But, it would be nice to be able to control them as RGB. So I'm looking at the frame PCB schematics from LDO's Github and it looks like the unused filament sensor port might be usable. It shows 3.3V, ground, and two signal pins. Since the X endstop route works and it has 3.3V that should be ok. The catch is the circuit shows a 10 ohm resistor instead of the 100 on the RGB ports. I'm thinking I might need to insert a 100 ohm resistor in the signal wire to make the colors come out right. Am I on the right path here, or is this a dumb idea? I'm still a total noob with electronics, so I really don't know if this would work or if I'd release any magic smoke and/or angry pixies trying this.

As for "losing" the use of the filament sensor there, I have the spec foot-mounted one installed and working.

I had the same problem, I used the X-Stop connector to get the wires to the back of the printer. Then from there, I am driving the lights with a GPIO pin in the Raspberry Pi. Be carful Pi3 pins can't source much current, but they can sink a few milliamps, enough for a sequin LED. So the sequin is powered by 3.3 volts from the Pi and the GPIO is inverted. Works well.

Eventually, I might change this and have the GPIO trigger a small MOSFET (2n7000) and then run 5 volts but the LED is bright enough on 3V3.

The best solution is planned for some months down the line when I replace the tool head with one that is based on Orbiter2 and then use either CAN or USB. Then I will use neopixels in place of the LEDs.

But for now the Sequins are wired via the umbilical to the Pi3 GPIO and it just works at a cost of well under $1.

 

[Yaaj]

Just received white sequins (NOT the RGB ones), and did some testing the lazy way with the lab PSU :
3.3 V : 5mA ; maybe the EBB36 I²C pinheader could do it (PB3 PB4 ; 8 mA max) but pin 1 is 5V, not 3.3V ; additional resistor needed
5.0V : 20mA
Are PB3 PB4 5V tolerant ? Was unable to find the information in the datasheet. (I²C pins are 5V tolerant on the Bluepill)

 

[claudermilk]

I now have the Dragon Burner and G2SA built and working. I ended up using extra NeoPixels I had laying around from the initial SB build for he Trident. That means just grabbing 3.3V and gnd won't work since I need a signal too. I looked again at schematics and I don't see anything that looks like it will work for another 3-wire NeoPixel hookup, so I think I just need to run the wires back to the Pi and attach to the GPIO for signal and piggyback off the 5V supply to the Pi. I don't feel like tearing the printer down for that right now, so I'll tackle that later.

 

[Yaaj]

I've gone for an EBB36 on my mini-SB, that has a few spare parts and could probably drive those - might need to get some sequins and do a guide

At first, didn't notice the RGB control pin on the EBB36 is buffered by a SN74LVC1G17 : +- 50 mA 5V available on the RGB pinheader (pin #2)
I hooked the two (white) sequins in parallel. The RGB output has an onboard 100R resistor. Each sequin has a 100R resistor. Equivalent circuit is 300R in series whith each LED. Direct voltage 3.4V : each LED will draw 5mA (total 10mA)

Not super bright but acceptable...

(please correct me if I'm wrong !)

printer.cfg :

[output_pin nozzle_LED]
pin: EBB36:PD3
pwm: True
cycle_time: 0.02 ; 50 Hz
value: 0

 

[claudermilk]

An update to this. Since my part cooling fan died, I had to tear the tool head apart to replace it. I decided while I was in there I may as well finish the wiring to the LEDs. I ended up jumpering the 5V in to the Pi power. I took a couple of 3-position Wagos and spliced the LED power in that way. I then took the signal wire and stuck it on the Pi GPIO12 pin since that's available (the preferred GPIO18 is occupied by the LDO hat). When I powered up, the Pi still powered up and no magic smoke or angry pixies were released. After getting the umbilical properly seated (oops), the printer boots up.

That done, I tried defining a led_effects setup for the tool head LEDs. No errors there, but it's also not running any effects. I see in klippy.log a lot of "Neopixel update did not succeed" errors. I did some more digging and turned up some issue thread on the Klipper GitHub project; apparently Klipper does not support driving NeoPixels on a Linux mcu build. But here's the weird thing: I'm looking at the printer and the nozzle LEDs are lit even though no effects function and the logo is inert. So something is driving them.

I guess the next step is figure out how to drive them entirely via the Pi. Hopefully there's a how to on that, because that's now well out of my small area of knowledge.

I guess worst case is I have always-on nozzle LEDs which is the primary goal (and happily they are not too bright).

 

[claudermilk]

Did a bit of play. I looked at the Adafruit and Pihut NeoPixel tutorial pages. I can get the LED to turn on and off and give it colors via the libraries running a short Python script. So I know that I have the basics hooked up right. However, I only get one of the nozzle LEDs to turn on. The other nozzle and the logo don't respond at all--I likely have a wiring issue past the first one.

For now, I've kludged together a couple of on/off Python and shell scripts I can call from a macro. So I can at least have *a* nozzle LED for now. I just don't have the patience to tear that tool head back apart at the moment to troubleshoot my wiring soldering (I suspect it's a bad solder joint in the little bridges between the nozzle LEDs).

 

[ChrisA]

Did a bit of play. I looked at the Adafruit and Pihut NeoPixel tutorial pages. I can get the LED to turn on and off and give it colors via the libraries running a short Python script. So I know that I have the basics hooked up right. However, I only get one of the nozzle LEDs to turn on. The other nozzle and the logo don't respond at all--I likely have a wiring issue past the first one.

For now, I've kludged together a couple of on/off Python and shell scripts I can call from a macro. So I can at least have *a* nozzle LED for now. I just don't have the patience to tear that tool head back apart at the moment to troubleshoot my wiring soldering (I suspect it's a bad solder joint in the little bridges between the nozzle LEDs).

I bet the issue is the timing. The library you use has to "bit bang" the GPIO pin. this works on a microcontroller that is not running a multitasking OS. But software delay loops and such don't work well on Linux because at any unpredictable time, the CPU might stop running your Python script and do something else and then get back to running your script.

 

[claudermilk]

Yes, in their issue thread the response is that it's a timing issue. I'm sure for the fancy led_effects display it's going to break the functionality; for my current kludge, I can have it pause for however many milliseconds to flip the on or off switch.

The other option I can think of for now is to wire these into one of the other strips, but that's now a wiring mess (either run out to the front of one of the daylight PCBs--bleh, or to the bed logo--double bleh). That or find some kind of expander board to get a 4th NeoPixel port for Klipper to address--not sure how much extra bits I really want to cram in the back of a V0 for this.

For now, I have to tear the tool head back apart (again, bleh) to fix my poor soldering. Then make sure all three LEDs tun on when I tell it to. From there, probably tweak my scripts to take some parameters--learn how to do that in Perl--and call it a partial success.

 

[claudermilk]

Quick update here. I had some time (and motivation) to take the toolhead apart again and re-do the bad nozzle LED jumpers. Swapped out the oversized, badly-soldered wires for smaller 30AWG with nice, clean joints. Now all three LEDs light up as expected. I had to tweak my little Python script, but now I can drive each LED to show the color I want. Next is to learn a bit more Python and make these follow the led-effects colors for the rest of the light show on the printer.

LOL, I said "Perl" before, no, it's Python.