trident Chris' 250 Trident Journey

[claudermilk]

It appears I have some serious tuning (re-tuning) to do. The printer is printing like poo right now. Even after running input shaper, I'm seeing ringing. Overhangs are awful as is bridging. I also have pretty bad divots on layer seams on most of the layer. I look back at some calibration cubes from months ago and 

[mvdveer]

4 hours ago, claudermilk said:

I look back at some calibration cubes from months ago and 

Awwww crap! Hate it when that happens. Lot in the words: "If it ain't broke then don't fix it" But then why else would we build a Voron if we cannot upgrade and improve it? I feel your pain though. Good luck

[Buurman]

I am kinda at the same point, that why I went of TAP again and back to Klicky, since my prints werent consistent.

Dont think it was TAP though, still working on stability, feel like the CW2 suddenly had extrusion stress or something, first layers and infill layers are just sometimes thinner and thicker.. makes me crazy.. 

First it was this nozzle, and replacing it with a smooth one clearly helped, and turning the extruder push strength a little also helped.. maybe it was slipping a bit.. dont seem to trust my CW2 now also..

All in al.. know how you feel... had a perfectly good printing printer and.. well.. F***ed it up...

Good luck in your troubleshooting!

 

I have printed a Sherpa and impressed me big time, maybe im going that way.. mini stealth with Sherpa and Revo.. lets see! going light!

[claudermilk]

Time for an update.

Quote

Inigo: I am waiting for you, Vizzini. You told me to go back to the beginning. So I have. This is where I am, and this is where I’ll stay. I will not be moved.
Brute: But the Prince gave orders —
Inigo: — So did Vizzini — when a job went wrong, you went back to the beginning. And this is where we got the job. So it’s the beginning, and I’m staying till Vizzini comes.

Tap is working reliably and I'm loving it overall. But, I managed to completely bork my calibrations, so as Inigo said, I went back to the beginning. I grabbed Andrew Ellis' latest SuperSlicer settings and used that as a baseline & ran through his calibration process from the start. Now the printer is back to printing high quality parts again! If anything, I may have it printing better than ever and possibly slightly faster. My input shaping results are showing the highest acceleration limits I've seen on this printer--I can run it right up to where my testing shows the motors starting to lose steps.

I now have the full range of Revo nozzles from 0.8mm all the way down to the "experimental" 0.15mm. I posted an album of my last batch of tests using the smallest nozzles. The tank is 0.25mm nozzle @ 0.1mm layer, the figures are 0.15mm nozzle @ 0.07mm & 0.08mm layers. For the tabletop wargaming they are intended for, they are more than good enough. I'll have to try priming & painting them to see what a finished product looks like. My friend who fed me the files (so I could print up some of this for him) told me they are better than some pieces he had previously purchased & many of the guys in his local group produce.

[Buurman]

What is your max Volumetric speed with the Revo? (or didnt you set this?)

[claudermilk]

I've set it to 11 according to E3D's information. For the 0.15 nozzle I reduced it to 7. Revo is not a high-flow hot end. I'm ok with that, for now I subscribe to Nero's philosophy: I'd rather have reliable, high-quality parts than #speedboatrace.

[claudermilk]

Time for another mod! Because why have a well-running printer without messing with it and try to break it?

I'm about to install the WobbleX mod for lead screw mounts. This is adapted from the HevORT printer and the mod is also by MirageC79. The idea is to try and eliminate Z banding from any wobble in the lead screws. Granted, my LDO ones are pretty darn good, but why not give it a try? The Mellow parts just arrived (about two WEEKS early  ), so now I have to catch up and print the parts.

[mvdveer]

9 hours ago, claudermilk said:

Time for another mod! Because why have a well-running printer without messing with it and try to break it?

Because it is running well, that’s why. Will be interested to see how you go. I got a very well running Trident as you know 

[claudermilk]

I didn't get the WobbleX installed last night for two reasons:

1. Turns out the printer wasn't as well-running as I thought. There is a lot of belt dust building up. Upon further inspection my pin idlers were crooked, so the belt was riding up on the edge of the bearings and rubbing the printer carrier. PLUS, the M3x30 tension adjustment bolts are too long. So one bottomed out and tore the heat set insert loose. I ended up having to de-belt the left side and chop the carrier piece apart to get it removed. In the process I broke the front plate locator pins. Good thing I printed up spares. Now it's all good with M3x25 tension bolts, and some very careful adjustment. The belts are remaining centered, so this issue should not crop up again. Since I was that deep in, I went ahead and finished up my release version of the hot end carrier and SB front plate with Rainbow Barf logo LED. That works nicely and I'm playing with LED definitions again. It then became too late in the evening to get into tearing the bed apart, so WobbleX will come later.

2. It turns out the read bed mount part has a design flaw with the pin retainer section. There needs to be a recess in the middle so the ball bearing can ride on the pins. I'm not the first one to run across this, there's an open issue on GitHub. No files uploaded, so I'll be impatient and jump into Fusion360 myself to add those & reprint.

[polygonprint]

I have been following along slowly (when I have time to read through) with the build progress claudermilk. Great read and I have learned a lot. Much of it has helped me with my own Trident build. I'm at the point where mine is 90% finished and printing.... although not terribly well. Anyway, just wanted to say thanks really! 

[claudermilk]

14 hours ago, polygonprint said:

I have been following along slowly (when I have time to read through) with the build progress claudermilk. Great read and I have learned a lot. Much of it has helped me with my own Trident build. I'm at the point where mine is 90% finished and printing.... although not terribly well. Anyway, just wanted to say thanks really! 

I'm glad to hear this thread helped out.

To help with getting your Trident printing well, go through the Ellis Tuning Guide. I use his configs as my baseline.

[claudermilk]

I got the WobbleX mod installed and on first impressions I think it will help. 

  

The point is to decouple the leadscrew from the bed with that machined metal middle bit. There is a steel ball that rides on the set of four metal pins. This allows the leadscrews to wobble all they want, but the bed isn't forced to follow.

Just from running the bed up and down the length of travel my ear tells me it's working. My Trident had developed a horrible howl when rapidly running the bed down--that's gone now. It happily runs the entire travel silently now. I'll do a Z banding test print later today, but the silence makes this mod worth it to me.

Installation isn't too hard. I just dismounted the bed at the spherical bearing mounts and set it on the bottom. Then it was fairly simple to dismount the parts from the carriages and swap over to the updated parts. The rear mount took a little more thought. To make it easy, I pulled the mount link off the drag chain and got that bolted up before re-mounting to the carriage. I totally thought of that before twisting into a pretzel trying to feed those tiny flatheads through the cable in-place.  I inserted the balls after setting everything in place, which was a bit of a pain. I'm not sure if sticking them to the magnetized center plates first would have been easier--but I was worried about them getting knocked loose. And I didn't want to lose my balls...

I do lose 10mm of Z travel, but that doesn't bother me--I have yet to print anything even close to the max range. I now have 220mm of Z travel available (printer definition updated accordingly).

[claudermilk]

I printed up a Z wobble test print and it looks good. I don't see any wobble at all now. Seems like this mod works. 

[Gimmi]

Since I am currently building a Trident, I will of course also include the WobbleX mod in the project. Ordered it from Mellow three times and in the same breath a vz-hextrudort extruder. I still have one question at this point, is there a tank ball mod for the Trident? Can't find one.

[claudermilk]

Not that I am aware of.

[claudermilk]

Since my wiring wore out (sad story here), it's time to rewire the printer. I scored a set of hartk 2-piece toolhead PCBs from West 3D during the holidays and have put off this project. Well, now is the time. I also sourced and LDO harness rather than build my own and and LDO breakout board from their kits for the electronics bay end. Lastly, I got the new v2.4.1 5-24V optical Tap PCB while I'm in there changing everything.

So far I've gotten the toolhead disassembled and the wiring pulled from the gantry down to the deck panel. I re-did the wiring for the toolhead and got the PCBs all installed. The Omnifixo third hands made some of the soldering prep really easy.

All soldered up and installed.

  

One really nice thing is the stock Revo wires just plugged in, no mods needed. The thermistor is the perfect length here, and the heater is only a bit long, the slack easily handled.

I have to build the aux 2-wire cable since that isn't included in the LDO part--and that's where the LED signal is carried--but that's easier than building the entire 14-pin main harness.

[Penatr8tor]

@claudermilk That's the same PCB I used in my Switchwire. One thing I really liked is that it eliminated a good number of duplicate/redundant ground and power wires. Less wires in the drag chain = less chance of wires getting pinched or broken.

[claudermilk]

That and it looks to be much easier to remove and replace the SB front. Messing around with all those plugs and cramming them in gets to be a pain.

I spent a while studying the LDO and hartk docs and now I'm pretty sure I have a good idea of the jumpers to make between the breakout and the Octopus. I also have some smaller cable raceways coming in; I've always been unhappy with the massive ones I used and now I've found smaller ones more to the size LDO provides. So I'll do some more cleanup under the printer while I've got it torn apart.

[claudermilk]

 

I have the entire above-the-deck wiring redone now; all wires that move are new. I'll tweak with the PCB alignment a bit before finalizing.

I'm well into tearing apart the electronics bay. I've removed the oversize cable raceways and replaced with smaller ones. That allowed shifting the Pi/Octopus DIN rail to allow a raceway on both sides now, though it's still tight. Some minor rerouting of wiring not involved in the gantry to take advantage of that and clean things up. The Octopus, Pi, and cooling fans breakout bracket had room to shift and make space for their new helper.

Because I originally wired the old school way, it was a single run from the toolhead straight to the Octopus with one hugh bundle of wires. I terminated in placed, which meant that individual runs were intermixed. The upshot is that to remove these I had to cut them and fortunately I left lots of slack under the deck. I realized that I don't have to build all-new jumpers between the breakout and Octopus--just trim the existing old wiring to size and re-terminate! Win! It cuts the mental effort of making sure my wiring is right almost in half. The main difficult ones now are just the extruder motor and probe. I also have to update the Y Endstop and build the 5V CT jumper per hartk's instructions from scratch.

Hopefully I'll have this wrapped up tonight.

 

[claudermilk]

It's alive! Well, mostly.

I was able to simplify a bunch of the wiring by reusing what's already on the Octopus, so the crimping was cut almost in half, and wire routing was already halfway done. Fans and endstops were pretty simple. I ended up with two screens (phone and tablet) from LDO's documentation showing both ends of the extruder harness so I could make absolutely certain i got that one right--and I re-pinned it twice.

It's a bit cleaner than before but not as super-sanitary as some folks. It's good enough and better than before.

The LDO breakout board, fully loaded.

I triple-checked all my wiring connections on the breakout, then YOLO'ed it and fired up the printer. No smoke or sparks! It booted! Woo! X and Y endstops trigger correctly, so I can home XY! The extruder moves! In the correct direction! The heater heats!

Now for the issues.

The toolhead LEDs are having a seizure, so I need to fix that. I recall seeing that others had a similar issue with the hartk PCB, so I will have to dig around to find the solution there.

The more annoying one is the Tap optical sensor PCB is not communicating with the Octopus correctly. I am getting LEDs on the OptoPCB. Blue with nothing in the sensor, red with something in the path. Cool. Also, it didn't smoke. However, Klipper is not responding at all. I've shifted to the PB7 port and set it 24V because of the hartk PCB. I've tried with the pin definition as "PB7" and "^PB7" that only determines if it's stuck at triggered or open. I used the included wire with the OptoTap and made sure it's plugged the way the bag label instructs (black plug on OptoTap, white plug on hartk PCB). I included the ABL-J40 wire in my triple-checking and I thought I had it good. Clearly, I have the 24V correct since the magic pixies remain captive. I'll have to check that again when I have a chance. But I'm also not seeing the LED on the OptoTap board change state when mounted.  So I have a mechanical issue as well to dig into. 

[claudermilk]

Reprise this one.

I found the solution for the wacky LEDs on the Voron Forum. Simple: add a 220ohm resistor inline in the NeoPixel signal cable and it clears up the interference introduced by the 2-piece toolhead PCB and breakout PCB.

Tap is now working again with the simple process of moving the signal pin back to PG15. I also epoxied a bit of filament to the top of the Tap carriage part so it triggers the OptoTap sensor.

Now to reprint all of Tap to the r8 versions and rebuild my toolhead. Again.

[mvdveer]

2 minutes ago, claudermilk said:

Now to reprint all of Tap to the r8 versions and rebuild my toolhead. Again.

Oh the joys  . Sure it will be worth it in the end.

[claudermilk]

It will. Just like this week's pain of rewiring and reconfiguring have been. The printer is fully up and running and is now printing the simple plate that it blew up on. Printing beautifully so far; I even nailed the Z offset right off the bat--no live adjust needed.

It looks like in all the faffing about I managed to break the webcam mount. I think I can acetone weld that back into place. Until then, it looks like it's mounted on the end of a Harley handlebar. 

[Penatr8tor]

3 hours ago, claudermilk said:

Printing beautifully so far; I even nailed the Z offset right off the bat--no live adjust needed.

Congrats Chris!!! Hope it stays that way for a while. 

[mvdveer]

9 hours ago, claudermilk said:

The printer is fully up and running and is now printing the simple plate that it blew up on. Printing beautifully so far; I even nailed the Z offset right off the bat--no live adjust needed.

Well done - back in bussiness.