The Sensor project - Adding Telemetry to a TAZ

Most of my 3d printing failures or glitches have been caused by 2 things. Overextrusion, or a filament stoppage. I see it time and again across all 3d printer forums, people who have their filament set wrong, or tales of 20 hour prints failing on hour 21. and it’s something that I believe we have the technology to work around. I’ve been looking to add sensors to my 3d printer ever since I first started playing with the old AO-100. Only recently have some sensor pieces made their way to the market that can be adapted for use with a TAZ.


My goal is to measure the width of the incoming filament, determine if it is still present and moving, eventually determine how much plastic is still remaining on the filament spool, and allow the printer to account for ambient temperature and humidity. I also may want to do auto bed leveling stuff. At the moment, I don’t have a good solution for the temperature, humidity and filament weight / used requirements, but filament width and movement monitors do exist.

I plan to build a block that will replace the upper right hand filament feed tube guide, that will incorporate the Tunell labs Filament movement monitor, and the Optical filament width sensor created by flipper. I’m planning on monitoring a single filament at this point, mainly due to space and board header constraints. Filament will feed in from the primary spool, past the width sensor and into the motion sensor. The width sensor might be slightly more accurate closer to the print head itself, but that adds weight and complexity there, and the circuit board is fairly big. For my purposes, the filament I use is usually fairly uniform. I just forget to change the overall diameter, so I am ok with relying on a measurement several feet from the nozzle.

I also have bed leveling sensors, though both of them I have on hand at the moment appear to be the wrong type. I’m not sure I want to go with auto leveling yet, mainly because I am really pretty good at manually leveling the bed, and the way it is now I don’t have to wait for it to auto level before my print starts.

I figure at the same time I’m going to finally install the new hexagon hotend and the nozzle cleaning retrofit. I’m not going to use mini style leveling, but it would be nice to have a bit less gunk baked onto the nozzle.

So, that’s the plan so far. I don’t know how well, or even if it is going to work. But we’ll see how it goes!

Filament movement monitor

Fillament width Sensor


Excellent projects piercet. I look forward to following this thread to see how it all turns out

Thanks!

I’ve started design work on the mounting plate for the two fillament sensors. My current thought is to keep a guide tube on the bottom, possibly wtih a straightening bearing of some sort. I really don’t like either of the stock case designs. The fillament width sensor light placement is too tall and in the way. i’m thinking I may use a couple of mirrors to lay that light sideways. The fillament motion sensor case is just a bad design. I’ll probably borrow from the Wade extruder Idler design, but on a smaller scale for that piece. The whole assembly looks like it will occupy around 70mm-ish of the back corner rail, but it shouldn’t stick out any more than the existing fillament guide.

I also put together the new toolhead. it still needs a few things wired up but it’s getting there.

What ever happened with this?

The parts are still on my desk. I got sidetracked by the anti wobble project and some of my ROV wiring projects. Its next on the project list though.

Alright, starting to work on this again. I’ve ordered a series of Mill-max spring loaded contact pins, both soldier cup and crimp fitting (https://www.mill-max.com/new_products/detail/138). My goal is to make 2 things. A quick release mounting contact block for the Taz X extrruder carriage, and a quick connect block for the sensor filliament block, which will probably need to be taken off for cleaning occasionally.

My goal is to come up with an arrangement that will work with existing Taz tool heads, but will use the space to the right and left of the screw mount to attach a pair of 14-16 pin spring terminal blocks, or possibly a single 22 terminal pair in a location and a manner that will allow someone to attach at least 2 extruders, 2 sets of fans, and 2 +/- 2 wire auxiliary devices. If it goes well, works easily and is cost effective, I’m going to propose it as a mounting standard. We’ll see how it goes.

I have the wire for the sensors in place across the upper gantry now, using wire clips I’m using a standard black Cat6 cable, which will leave me several conductor pairs for other additions later.

I have 2 more parts to repair a large christmas display merry go round yet to print with the red fillament, and then I’ll be back on designing and printing the housings for the sensors. They have been giving me more trouble than they should, mainly because I’ve been getting destracted by shiny things. I’m thinking they will basically be on a flat plate, that will screw to a 90 degree mounting bracket of some sort. I have a good 3 inches I can play with to space the filliament off the frame if necessary without swapping the stock guide tube out for something longer, but I want to keep everything as close to the frame as possible so it looks nice and doesn’t become a snag hazard. I also need to accomodate a post of some sort for the guide arm for the second filliament run, which will not be passing through the sensors, but will be needed for those times I run the dual head extruder.

The mount for the Z probe is done and tested to my X openbuilds extrusion mod. There are other mounts available for use with a conventional Taz arrangement, so i’m probably not going to work up anything too custom for that. This project is probably going to bleed over into the X openbuilds carriage design elements somewhat, along with the cable chains, just because everything comes together about that same point, but quite a bit of it will be usable by a stock Taz anyways.

I always thought about how one could do a quick-release for the various hardware and electronics you can put on the carriage. I like the idea of using those pins. Excited to see how it works out

I use astroprint to do my printing - through a raspberry pi.

I have often thought about building a filament filter that also includes a feed sensor driven by a atmel chip.

The idea is this -
It’s just a simple encoder driven by the filament passing through a passive pull system. if the encoder doesn’t change state within a set amount of time it triggers a GPO - which goes to my astroprint pi - which triggers a pause command and an alarm.

Let me know if you want to collaborate and share some ideas!

I think the next step will be some sort of secure extruder carriage latch. I don’t want to add a bunch of mass to the carriage, but at the same time, I think a working, solid quick release would be worth the penalty. I’m thinking that a V shaped dovetail running the entire length of the carriage with a spring clip at the top or a Cam ratchet clip might do the trick. If I make the dovetail integral to the carriage as part of its backbone I think I can keep the mass about the same overall, maybe a bit higher due to the latch. I figure the components of the pin contacts are going to add a bit of weight as well, but it should still be manageable. Definitly need to incorporate an upper extruder cold end mounting screw post as well still. That retrofit extruder that Discojohn came up with or modified for the hexagon that has the post extending up to the existing mount hole works really really well for oscillation damping so I don’t want to lose that, but at the same time that could be a bit more compact and still functional too. Some of the latch designs from that NASA “print a space clampy thing” contest a while back would probably work well.

Whatever I come up with will need to work with both the Openbuilds extrusions and the Stock Taz carriage footprint to gain widespread acceptance though. I don’t think anyone will have a problem swapping out the extruder mount carriage plate itself for something different, so long as the bearings line up with it either way, and you still have the option of using a screw secured toolhead if you want to.

That sounds like it should work, You’ll need to account for retraction, but the theory is definitly sound, and thats basically what the one I bought does, just at the Rambo board level rather than via a Pi board. You should still be able to connect to it via I2C if nothing else.

The problem with using just the rambo board is that Astroprint will not know that the print has paused.

The entire thing would work off pin change interrupts, so direction shouldn’t matter as long as it’s spinning. You could also use it to calculate filament usage.

I’ve been toying with the design of the extruder pin layout and latching mechanisms, and I’m starting to come up with a plan. The best flat surface cam latching mechanism that I’ve run across is the Sony VCT-u14 Large video camera tripod plate. It’s basically a large wedge with a notch in the middle to allow for a retaining pin.
VCT1_large.jpg
For a 3d printer, a printable cam lever, either top mounted or side mounted would probably be ideal. Slide the toolhead into the V notch, which would supplement or replace the existing one, pull the lever, and locked in place extruder with a minimal amount of extra material. That’s the theory anyways. The difficulty there is that the pins would basically need to be on the bottom to align properly.

Here are the model numbers of the crimp variants of the pins I am planning on using, if anyone wants to grab a set.
Pins:
https://www.mill-max.com/products/socket/826-XX-XXX-00-001000 - PN 826-22-004-00-001000
Contact point:
https://www.mill-max.com/products/socket/430-XX-XXX-00-240000 - PN 430-10-216-00-240000

On the actual sensor portion of things, I’ve been debating placement of the filament path. Right now it feeds up the rear right hand corner of the Taz gantry. The filliament feed sensor itself does however have an “arm sensor” button on it that you will want to push before longer prints to turn the sensor interaction on. There is a pin header to relocate that button, but it might make sense to move the whole filliament path to the front right corner of the frame instead to minimize extra cable and assemblies. It shouldn’t impact printing at all, as the geometry to the extruder carriage would be essentially the same.

The next question is the feed into the sensor block. I want to line the internal path with plastic tube for friction relief and ease of feeding as much as possible, but I also kind of like the idea of a couple of roller bearing wheels below to block to straighten any incoming filliament before it hits the sensors. The simplest pathing would be a couple of zip ties and a section of tubing. Then there is the whole dust filter / filliament oiler issue. I never use a dust filter, but some people swear by them. Should I include space at the top of the sensor block for a dust filter? Internally, the path is going to go from the width sensor, to the feed sensor, then out the top with the new anchor point for the filliament guide tube. that will be anchored to the top sensor block and will no longer pivot since the filliament angle entering that tube will be constrained by the sensor block itself. Does the incoming feed need that pivoting capability to make a better feed path though? I dunno yet.

I’m thinking of changing the filliament spool arm at the same time. The stock one works pretty well, but a ball bearing one with adjustable drag might also be nice. That might end up being its own project though. This one is getting convoluted enough as it is.

Got the spring pin samples in. I ordered the wrong number of pins for the pin blocks, but the size looks like it should work. The larger block is the contact block, with a U.S. Quarter in the background for scale. These are teh Crimp Barrel variant, though they look alot like the soldier variant as well. Current plan calls for crimping each pin and socket individually, then recessing in the block, then mounting the block to the appropriate location. I may end up going with a larger size pin to ease assembly. The weight of the unit is minimal, and there is some additional space there, and the cost isn’t that much different so going larger may make for easier assembly by humans with fingers.

The bracket to add the nozzle wiper to the Taz. http://www.thingiverse.com/thing:1261303

You basically print it, remove the built in support material, add 2 M5 heat set inserts, then use it to drill the mounting holes into the bed, either for an M3 bolt tapped, or a through hole with a nut on the bottom if you have enough room. If you are using the Openbuilds Y axis you may have to go with the tapped hole approach as there isn’t alot of clearance there.
Retrofit_nozzle_Wiper_1_0_a.stl (94.2 KB)

The nozzle wiper, which isn’t technically a sensor, but was part of this project anyways is now installed. I also finished the probably final X axis cable chain brackets.

The housing for the sensors is still in progress. I think i’m on revision 20 of ones that I don’t like so far. Its possible to get writers block when designing apperently. So bear with me, I’ll get the rest of things done here eventually.






one more


Thanks to everyone in the mini forum that gave me the right heights

So, mini project update - This project isn’t dead, I just keep getting distracted by shiny things. Presently that’s the open source “uses leftover Taz parts” laser engraver frame I’m building which will be released at some point after it is done being designed and tinkered with. The sensors themselves I have tested with my Taz, and they do seem to work with a bit of firmware modification (taz 5 firmware may have the values present already)

Aside from the sensor project (which includes the retrofit carriage connector and the spring pin thingy, the following 3d printer modification projects are also in progress to some degree:

  1. Bondtech core “standard” filament Greg Wade style extruder with duall hobbed thingies. - Parts ordered, base frame design finished, idler and other components not started
  2. Taz 6 Openbuilds X axis retrofit - pending existence of a stabilized RTM Taz 6 design
  3. Maybe some variant of a dual extruder - not started
  4. PC-less 3d printer jog pendant maybe? - possibly not even feasible
  5. project "Pup tent style removable stowable enclosure of doom - Pending sewing machine
  6. Maybe project “will a ballscrew fit on a Taz” - I kinda don’t want to do this one because my Taz is printing wonderfully and I know exactly how much of a pain in the ass it is to fit a ballscrew to a relitivly small 3d printer leadscrew space.

And that’s about it. If anyone has anything in particular they would like to see done to a 3d printer, let me know!

Hey Piercet, have you looked into either using a Raspberry Pi or better yet, a WiFi enabled Arduino like a Particle Photon to give you text/email updates? I currently have a home security/monitoring system in place using the Particle Photon that sends me text messages based on installed sensors I have… door magnetic reed switch/smoke sensor/motion sensor etc…

I have one of the Wink Hub units here in my house for security and door sensor monitoring, etc. One of the Octoprint projects would work great for most of that notification piece. I like having the ability to control my printer directly with repetier though, so I don’t want to give up that USB port on the Rambo board to enable that. I’ve been thinking of swapping in one of the boards with the faster processors, like a Duet board or a smoothieboard, at least one of which has a built in Ethernet port. There are “make my sensor into a wireless communication sensor” pods you can attach as well, so somethign like that could be worked up. Once I get the laser cutter project put to bed I should be back on this one though.

Have you looked into using a hall sensor for the filament sensor? Like this one: https://www.youmagine.com/designs/filament-diameter-sensor . Maybe the hall sensor could be incorporated into the tunell monitor as one mechanism that can monitor movement and filament diameter. Also, I read that there may be some issues with using a light sensor to measure clear/semi-transparent filament width.

I vote for keeping the filament sensor without a filament filter, since filament filters are easy to make and add anywhere.

I like the simplified connector design. I had issues swapping extruders before. But since my openbuilds update, I modified my TAZ to use the 16 pin connector on the newer TAZ’s and used 3 of the unused pins for my proximity sensor. So it’s just one connector to unplug and I added a thumb screw for the mounting screw. So I can swap tool heads in about 20 seconds: http://i.imgur.com/hFwDzvY.jpg

In case I need to flash the firmware for different toolheads, I have a small custom LCD with rotary dial connected to my raspberry pi that allows me to quickly flash a custom firmware. Although, now this could be easily accomplished with a very inexpensive tablet ($50 Amazon fire) connecting to octoprint via web interface.

Keep up the good work!

Hmmm, that hall sensor is a new one. You could rig that up as a multi axis one as well, and it would fix the transperant filament issue. A pressure mount won’t work as well on squishy filliaments though, so maybe a combination sensor or a toggle switch for type?

I am curious about your LCD mod you mentioned. Do you have any pictures or details?