blt3dp's mostly metal delta build - AKA Metal Gear

Quote
DjDemonD
Looks okay but to be fair I'm not great at scanning these things and spotting the errors. More likely to fire it up and debug it that way.

Yeah, I'm gonna configure the probe and just print. I'll go from there.

Quote
FA-MAS
Got the new Igus bearings with adjustable preload. Got them installed and I was able to take out most of the play. If I took out any more, they would not move on the rail.
I am slightly disappointed that the adjustment screws are plastic, felt like I was damaging them just by turning them.

I bought the same years ago and didn't even one print with them. They are not suitable for a delta printer. Adjustable preload or not, doesn't matter.
Maybe you are luckier, but everyone in the german subforum demounted the TWE rails. There is always a little play, at least in one axis.
If you tighten the adjustment too much, the play is low but the friction to high.

Expect that you will not be satisfied with them for long time.

I learned my lesson, mine are still for sale.

---

Lars - Blog, Thingiverse, Git

_{After successfully ignoring Google, FAQ's, the board search and leaving a undecipherable post in the wrong sub-forum don't expect an intelligent reply.}

Quote
toolson
I bought the same years ago and didn't even one print with them. They are not suitable for a delta printer. Adjustable preload or not, doesn't matter.
Maybe you are luckier, but everyone in the german subforum demounted the TWE rails. There is always a little play, at least in one axis.
If you tighten the adjustment too much, the play is low but the friction to high.

Expect that you will not be satisfied with them for long time.

I learned my lesson, mine are still for sale.

I've gotten them adjusted to where I think they're best. You're right 2 of them are more than acceptable. There's one that took more adjustment and has a very slight amount of play. Of course I would like it to have absolutely no play. I'm sure that even regular linear rails in our price range (knock off hiwin, etc.) would have some play, unless I found a good deal on rails on ebay, which I am keeping an eye out for.

But, even though this machine seems to have passed the wife's noise test, I don't think it's going to pass the size test. I know my CAD drawing had gotten outta control when I had the frame assembled. I'll probably end up either printing parts for and assembling a Tantillus, or figuring out how to machine something of similar size. And it probably won't use linear rails.

I'm thinking the build volume of this printer could print an entire Tantillus, though I have an idea about using 8" square steel tube.

Edit: might anyone have suggestions for where to get good linear rails, providing I don't find any on ebay?

Edited 2 time(s). Last edit at 11/17/2017 06:08PM by FA-MAS.

Straight from Hiwin or from Robotdigg.com.

Said it before but the cheapy rails work too, if you give them a good clean and make sure the bearings are decent.

My kossel XL calibrates to 0.004 sometimes, can't be much play in anything including the rails.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I replaced them with hiwin clones.
These have no play and printing without any trouble. Last month my big delta passed the 24 kilometer of filament line. That are a LOT of running hours.

If you really want to have IGUS rails, what i cannot recommend, you had better gone for the N-Series rails.Take two sliders per carriage and put some preload to them, that is a delta proven system.

Years ago i made this picture of the TWE sliders:


If you look closely you can see some gray spots on slider parts.
These are the only point where they touch the rail, this cannot work.

I know you paid hard money for this and really want to believe that they will work sufficient...

---

Lars - Blog, Thingiverse, Git

_{After successfully ignoring Google, FAQ's, the board search and leaving a undecipherable post in the wrong sub-forum don't expect an intelligent reply.}

Having some issues with the dc42 ir sensor board. It blinks 4 times, but probe always reads 0 on the web interface and doesn't light up when I put anything near it.

Quote
FA-MAS
Having some issues with the dc42 ir sensor board. It blinks 4 times, but probe always reads 0 on the web interface and doesn't light up when I put anything near it.

If the led blinks at startup like it is supposed to but doesn't light up when you put your finger under the sensitive area, ask your supplier for a replacement. If it responds to your finger but not to the bed, then you probably have it mounted too high. The bottom edge should be about 1.5mm higher than the nozzle.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42

Quote
FA-MAS
Having some issues with the dc42 ir sensor board. It blinks 4 times, but probe always reads 0 on the web interface and doesn't light up when I put anything near it.

If the led blinks at startup like it is supposed to but doesn't light up when you put your finger under the sensitive area, ask your supplier for a replacement. If it responds to your finger but not to the bed, then you probably have it mounted too high. The bottom edge should be about 1.5mm higher than the nozzle.

It does not light up when I put my finger under it.

Edit: ran autotune and it seems fine now.

Edited 2 time(s). Last edit at 11/18/2017 03:54AM by FA-MAS.

Wait a second, does the heater have to be on?

I turned it on and now it triggers.

Edit: Also is there a way to increase the timeout before it determines a slow heat up is a heater fault? I know with this particular thermistor and heater combination, it needs like 10 more seconds.

Edit: Ran autotune and created an M307 for my config file, it's fine now.

Edited 2 time(s). Last edit at 11/18/2017 04:52AM by FA-MAS.

IT PRINTS!

Think the first layer is a little low. That'd be my probe trigger height to adjust?

Glad it's working.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
FA-MAS

Quote
dc42

Quote
FA-MAS
Having some issues with the dc42 ir sensor board. It blinks 4 times, but probe always reads 0 on the web interface and doesn't light up when I put anything near it.

If the led blinks at startup like it is supposed to but doesn't light up when you put your finger under the sensitive area, ask your supplier for a replacement. If it responds to your finger but not to the bed, then you probably have it mounted too high. The bottom edge should be about 1.5mm higher than the nozzle.

It does not light up when I put my finger under it.

Edit: ran autotune and it seems fine now.

It should respond to your finger whether the heater is on or not. If you hold your finger right up against the bottom of the sensor then it won't light up because you will block the phototransistor. But it should be easy to find a position just below the sensor that makes the LED light up. Similarly if you use a piece of white paper instead of your finger.

If it sometimes responds to your finger and sometimes doesn't, then you might have a poor soldered joint on either the phototransistor or one of the IR LEDs. The temperature of the hot end might make the difference between getting contact and not.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Getting alot of Ajax errors, timeout disconnects (even though I had just sent commands with the web interface), and inability for web interface to connect to firmware. Also sometimes after print has started, it immediately disconnects with an Ajax error and won't reconnect to the firmware.

Running 1.19 on both firmware and web interface.

Also, say my probe triggers at 3.366 mm and when I go to print it's just a little too far away, like 0.1, which direction should I take it? Add 0.1 or remove 0.1 from that?

I'd say add 0.1mm to probe offset in G31.

But to be fair it won't take long to try both ways. If you use babystepping during first layer for example to go up 0.15 to get to a good height as I did recently then I changed G31 Z0 to G31 z-0.15 and now it prints exactly 0.3mm first layers. So in your example it's reversed as your probe triggers above the bed and you are too far away, my probe triggers below the bed and I was too close.

Edited 1 time(s). Last edit at 11/20/2017 03:47PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Yeah, I did baby stepping to arrive at the 0.1 number. I wasn't sure which way to affect the trigger height.

It always confuses me as probe trigger height and nozzle offset are inverse. I have been asking dc to include a save offset button next to babystep buttons,but it's "gcode everywhere" which works well so they're sticking to that.

As for ajax see duet forum, long thread on this topic. I've never had an issue as I setup a dedicated router next to the printers so signal strength is max and no interference with any other devices as it's on a WiFi channel which is relatively clear.

Probably should have bought an ethernet there is a cable running to the router, but they weren't available when I bought my wifi.

Edited 2 time(s). Last edit at 11/20/2017 03:53PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Yep, adding increasing the trigger height to 4.766 got me an absolutely perfect first layer.

I also upgraded DuetWiFiFirmware (19.2), DuetWiFiServer (19.2) and DuetWebControl (19.3). Also increased the AJAX retries to 3 and decreased the non travel print speeds to 90 mm/sec, it seems to have taken care of the disconnects. Just doing some Slic3r tweaking now.

Edit: Spoke too soon, AJAX disconnect about 45 minutes in. Says due to timeout, I was able to reconnect this time though.

Edited 3 time(s). Last edit at 11/21/2017 02:43AM by FA-MAS.

I'd consider moving to the 1.20beta series especially the wifiserver (you can use this wifiserver version with 1.19.1 FW as I've been experimenting to see i there is a bug in the probing G30 code of late), a lot of work has been going into solving these problems.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I went to DuetWiFiFirmware-1.20b7, DuetWiFiServer-1.20-b8-nosleep and DuetWebControl-1.19.3

Also inspected the wireless in my area, some of my neighbors are using interfering channels. I switched to an unaffected channel, there are a couple more access points on the channel, but the receive signal strength is better.

Still getting disconnects, but I believe it's because my Duet WiFi is under the bed, the antenna isn't exterior. The receive signal strength is really bad compared to my laptop which is right next to the printer. I wanted to get a mount printed for it so i can place it correct, I'm hoping that'd resolve the disconnects.

Edited 1 time(s). Last edit at 11/22/2017 02:54PM by FA-MAS.

This is how I mounted the Duet WiFi under the heated bed while allowing the antenna to protrude outside the frame.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I will eventually be doing it that way, I installed enough t-nuts for your face plate and the aluminum bottom. But I don't have the bottom yet. I'm thinking of a 2 piece solution that would utilize the nuts on the front as well as the bottom.

Having a weird issue with home position
When I home the printer, M114 reports Z position at 271.399
But that's not what's set in my config.g
And If I command it down to Z0, it stops about 5mm above the bed.

Edit: just added to homed height until Z0 came correct.

Edited 1 time(s). Last edit at 11/24/2017 12:21PM by FA-MAS.

How do I deal with this?



Edit: Run G32 before G29?

Edited 1 time(s). Last edit at 11/24/2017 06:55PM by FA-MAS.

Yes, I'd certainly do that. You can also set z level by going to be centre then issuing G30. I'd move this sort of thing to the duet forum.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Quote
FA-MAS
Having a weird issue with home position
When I home the printer, M114 reports Z position at 271.399
But that's not what's set in my config.g
And If I command it down to Z0, it stops about 5mm above the bed.

Edit: just added to homed height until Z0 came correct.

Homed height (M665 H parameter) is how high the nozzle is when all three carriages trigger the endstop switches. The default homedelta.g file lowers the effector 5mm after homing so that it can be centred, so the actual nozzle height after the homing script finishes will be 5mm less than that.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

So here’s a question, my build is fairly accurate. My towers are at 90 degrees to the bed, my rod lengths are as close to equal as you can make them, but I’m noticing that my first layer is lower in areas. I’ve noticed that if I clear any bed transforms with M561, if I take the nozzle to Z0 in the center, if I move it around, in some spots, it is about 1 mm off the bed. Any ideas?

See [duet3d.com] for common geometrical issues.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

If you've check all the usual stuff I. E. [www.duet3d.com]

Then I'd be surprised if it's giving you problems.

Belt tension often overlooked, need to be even.
Carriages mounted slightly rotated or off centre.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions