G_T's build log - trials of an insulated hotter Trident

[G_T]

I have replaced the breakout board in the electronics bay, and found a connector problem there. Now the temperature sensor problem seems to be gone.

I have also replaced the rear Z motor with integrated leadscrew, that had a failing bearing. Now no more thumping as it rotates, and the lateral shiftability at the motor is gone. I'm thinking that might possibly get rid of the occasional unpredictable rather slight Z artifacts I get when printing. It wasn't bad, but the thumping was certainly getting worse fast.

I did a test print, actually a reprint of the part I was printing when the temperature sensor problem reared up and trashed the print. I killed it pretty quickly. It is clear I'll have to do some print recalibration to get prints back to being good. Oh well, just more filament and time!

I also took a bed mesh and did show one or two points in the large mesh were off. In my experience that means I should spend an hour cycling the bed up and down to work in the new motor and shaft to bearing fit. I've seen that make notable improvement in Z consistency in the past, for instance when I replaced the POM nuts with the PEEK ones I machined.

That's early in this thread. The POM nuts were the first thing I had to replace as I started ramping up the temperature. They're only good to about 85C before they'll deform, and the tiny bit of friction they encounter is enough to up the temperature at the threads a few degrees over chamber temp. Then throw on a little grease and that 85C-ish HDT is probably more  like 75C. The erosion of the nuts was obvious by the black grease and plastic gunk coming off the nuts. (I've seen numbers as high as 120C for HDT also, so it probably depends on the manufacturer) The nuts started to fail as soon as I started going above 65C chamber temp. So if any of you want to follow in my footsteps and make your printer be able to do a good job with hotter printing for materials like polycarbonates (not the EZ or adulterated ones) then don't use the POM nuts. Just replacing those nuts made perhaps 70C chamber as a max possible for occasional printing with probably ASA parts. IMHO hotter than that in the chamber and one should ditch all ABS or ASA, and use more metal and less plastic in general.

There are plenty of parts available for that sort of thing now, like the metal gantry I'm using and loving. I did stick with plastic for the rest of the parts but have reprinted nearly everything but the print head in PC-CF which can take a whole lot more heat than ABS or ASA and is pretty tough stuff. It isn't brittle, it doesn't lave layer adhesion problems when you print it hot, and under force it bends some but doesn't break.

In general that's the biggest thing I've found ramping up the chamber temperature, and the hot end temp. Layer adhesion gets a serious boost. Just have to print a little slower, but these sorts of filaments are not for fast printing anyway.

Sorry I rambled a bit. It's late.

Gerald

Edited June 17 by G_T
Correcting some numbers

[G_T]

I've done a little printing now, and definitely had to change the extrusion factor. I had to reduce it by 7% compared to what it was before! I might have gone a hair farther down than I should have, but a hair overextruded is suicide for this filament. Blobs form rock solid impact points for the nozzle to strike over and over. A hair underextruded eliminates the blobs. I do mean a hair.

I actually tune extrusion factor based on nozzle picking up any plastic when filling flat surfaces. I tend to tune to 3 digits and slowly change the number as the filament changes through a roll. I'm not saying anyone else should do this; it is just what I do.

The problem IMHO is PC at these elevated temperatures really really wants to coat the nozzle. So given any excuse, it will do so. But at some point the resulting blob will transfer to the print, and harden into a CF reinforced rock in the path of the nozzle. It can ruin a print.

BTW, at higher temps, no silicone sock, and no nozzle paint.

I tweaked a few settings but mostly it is like it was before I had the temperature sensor issues. I may un-tweak some of the settings. Not sure yet. The prints are almost as good - very close. Some more tweaking required but the prints are quite usable. Tests were PC-CF @ 120C bed 305C nozzle 78C chamber. I'm not going to get quite perfect prints since I'm running higher temp than what the extruder is happy with being ABS and not a high temp motor, with a tough and slightly irregular filament. Heck, the filament is sometimes even slightly oval (when I changed rolls, I mic'd the filaments)

One visible artifact of the extruder struggling is occasional 2-4mm long stretches with under or no extrusion visible in the side walls. I'm thinking that's the extruder missing steps. It's not long enough to be filament slippage. With lower temperature materials, such as PC-ABS, I don't recall seeing this happen. At higher temps, or longer prints at this temp, I've had the extruder motor simply shut down and go into air printing.

PC-ABS makes a much prettier print BTW. I expect straight PC would do so as well. This PC-CF I'm using for the printer parts is just industrial, if that makes sense.

I should probably do a few of the calibration prints to dial in some things a touch better. But will I? Probably not. I'll tweak on the fly. This filament will never dial in pretty like PLA.

Gerald

[mvdveer]

You are truly pushing the boundaries - good on you. Not something I will ever attempt.

[G_T]

It's interesting doing the higher temp printing. Some things just seem a bit different. At least the printer seems quite reliable at the moment, so I'm trying to print the rest of the replacement parts. Admittedly I've printed some parts several times as a tuning exercise, and to test the printer.

I'm loving that the printer no longer stinks when hot!

I just took a look at the 0.6 GammaMaster nozzle I've been using. I'm not sure how many hours, but I think it might have had some sort of toughened coating on it... which I wore through. The nozzle looks about 1 or 2mm shorter than it did a couple days ago! I wonder if the bore has opened up as well?

Nah, it was an illusion! The bottom of the nozzle appeared perfectly flat due to plastic buildup. Really it looked like a ground down nozzle! I pulled the hot filament roll from the printer and dropped it into vacuum, and let the nozzle sort of drip dry while hot. You sort of get used to handling hot rolls of filament. Pulling a roll from a printer when it is over 70C seems normal now. I'll have to be careful when I really crank up the temps!

So I scrubbed the nozzle clean and it looks like a nozzle again. I'm not certain it isn't a little shorter but it looks quite usable. Perhaps sometime I'll pull it off and check it under my microscope to see what sort of damage it is taking.

In case anyone is interested, Magigoo PC works very well on a textured PEI sheet for this sort of PC printing. I've had zero adhesion issues in a while now.

So I'm currently printing the parts for a MGN12H modified BOOP but I removed the X microswitch mount to make a cleaner part. I'll assemble and play with it when done. It might end up on the gantry with a mini-Stealthburner of some flavor as the new print head.

Gerald

[G_T]

I switched to printing PA-GF replacement RamaLama-II parts. Those are still ABS in the printer.

All seemed to be going fairly well (had to lower extrusion about 2.5%), then went into air printing after about half an hour. The print showed extrusion slowed down a lot for a couple layers. As soon as I killed the print, the extruder went back to working. Anything in the 70's for chamber temp has some odds of the current extruder motor fading or quitting. It only has to last long enough for me to print the rest of the replacement parts! But it is wasting time and filament.

If you haven't seen it, if you cook nylon a little too much you'll slightly burn it. It discolors some. Also if extruded a little too hot it comes out of the nozzle more kinky. But drop the temperature 5 degrees and it comes out a little cleaner. This time around I had upped the bed temp and lowered the extruder temp from my previous settings. Adhesion with Magigoo PA on textured PEI seemed good enough for a change.

I may add another +5C to the bed and -5C to the extruder next attempt. That would put me at 125C bed and 255C extruder for the Fiberlogy PA12+GF15 filament. Chamber was about 71C so this change would boost the chamber temp to 75C. I'm thinking the hotter the chamber, the less nylon wants to warp. I hope. It also seems to improve bed adhesion which is seriously important with nylon.

Gerald

[G_T]

Since the extruder is now giving me so much trouble, I decided to just reprint the parts and build a new CW2 with a high temp motor. I used 3DXTech's PC-CF, the higher temp stuff, using a 0.6 nozzle. It's what I've been mostly printing for a while now.

Amazingly the existing extruder had no significant hickups during printing. Lately that's unusual. I printed in 3 batches, and hovered over it a bit. The chamber temp wanted to go on the high side of 80C but that would have been the death of the current extruder (about 85C HDT!). So I ended up dropping the bed temp progressively a couple degrees to keep it down to 78C or a touch over. Even then I expected printing to fail, since it often does, but this time it kept working all afternoon.

Before lowering the bed temp, I tried using the exhaust fan as a temperature control. But it turns out I've sealed the printer up enough that it can't pull in enough air to move the chamber temp more than a couple tenths of a degree. So much for that idea. The exhaust fan really only works now if the door is open, so what's the point?

I haven't cleaned up the parts, or verified dimensional accuracy (bearing fit for instance) but they look usable. They actually look a bit better than they do in the pic.

The parts are a couple times more rigid than the existing CW2 which should be nice. Personally I don't believe extruders should be made from flexible plastics, such as unfilled ABS. Feel free to disagree.

I also attached a pic of a recent bed mesh, using the textured PEI sheet I've been abusing for a while. I'm finally satisfied with it.

Speaking of temperatures, I see the raspberry pi actually gets a bit toasty before I turn on the bed heater which turns on the electronics bay fans. I need to change settings so those fans just come on at startup.

From the numbers I'm seeing so far, and what the parts should be able to take that are in the box or going in the box, I think there will be no issues at 140C bed and 95-100C chamber. I'm now expecting to hit those numbers, even without the auxiliary heater I plan to install - mostly just to speed warmup. All that is without adding another layer of insulation like I had intended.

Gerald

[claudermilk]

For my electronics bay, I've actually attached the fans to the Pi MCU thermistor that I have defined in the printer.cfg. Then some delayed_gcode macros to check on the Pi temperature every 5 minutes. Now those fans turn on when the Pi hits the defined high temp and turn back off when it's cool enough (for my setup I have it at 50 high, 40 low). I also swapped those fans to 12V Noctuas to quiet thngs down.

[G_T]

Sometime I should add an extra fan just on the Pi. I'd be interested in seeing your macro!

Yesterday somehow I missed printing one of the parts for the CW2. It is printing now, along with a Stealthburner fan shroud. It's a bigger part than I've ever gotten to succeed at elevated temps, but yesterday was a good printing day, so... Anyway my settings are not PIF settings, and I'm using a 0.6 nozzle. Hopefully all these parts end up usable.

I decided to purchase new Bondtech parts rather than transfer the old ones. So this round of print head rebuild won't happen for a few days.

So I have a quandry of sorts. I intend to replace the Mosquito hot end. It is working fine and has since the last rebuild, but I don't think it will work fine when I ramp the chamber temp higher. There just isn't enough cooling fin area for air cooling in a hot chamber. It's pretty tiny actually. I've talked this over with Slice, and I'm using it beyond its rating already.

I have two options on hand. One is a Dragonfly BMO. I like that this has a tapered interface between the heat break and the heat sink. It takes pretty common nozzles with lots of options, which is a big plus. It's also cheap enough and uses standard heaters and thermistors, and is rated for high enough temp to suit my applications. I want AT LEAST 400C!

The second is the Voron heatsink with E3D Revo High Temp or High Precision hotends. There is enough of a variety of nozzles available now for that system, for abrasives and high temp. But they are pricier, and limited to E3D. That I don't like. On the plus side, no chance of slow plastic leak around the threads or other oddities that might happen at high temp. And if I mess up a heat break, then I just swap the nozzle and I'm back to printing. No rebuild required. At least it no longer uses proprietary heater and thermistor setup.

Thoughts?

Gerald

[G_T]

PS - One advantage to the E3D solution would be to prevent a battle I'm fighting at the moment with my Mosquito. In spite of the nozzle having been hot tightened, a very slow oozage occurs through the threads at these higher temps. It flows down the nozzle in a fairly thin film, but thickens into a blob at the cooler end of the nozzle. Then of course at some point the blob detaches and makes a lump that the nozzle then bounces over for a while. Or worse, it then causes a filament detachment and things go downhill from there. It also somewhat coats the bottom of the hot block. It took me a while to figure out how that was happening. It happens so slowly it's like watching grass grow.

I'm trying to print the stealthburner fan shroud, round 2. Round 1 died in the above pattern. Round 2 I paused the print to scrub the nozzle at about the same point in time - roughly an hour into the print.

Gerald

[claudermilk]

That is definitely one of the design features of the Revo. If you are getting oozing from the top of the nozzle/heatbreak, you've got other problems. 

While a Revo nozzle is more expensive than just a V6 style nozzle, if you add on the matching heatbreak--which is part of a Revo nozzle--then the pricing is comparable. I looked at that back when picking a hot end. I don't run abrasive material, so my OG 0.4 brass nozzle is still going strong 3 years on. Though it isn't on the printer all the time--I take full advantage of the easy swap and change sizes all the time.

The new high temp heater shows it's rated to 500C, so that covers your requirements.

Here's my fan config for keeping the Pi cool. It's probably not the most elegant solution available, but it's working for me.

[G_T]

Thanks for the fan stuff! I'll check it out soon. Right now, too busy printing. The oozing seems to only be a problem with polycarbonate. I never had it with PLA or PETg, and it doesn't look like it's happening with nylon, at least not so far. Of course PC prints at a fair bit higher temp than the PA12 I'm printing at the moment.

The supports in the fan shroud were a real PITA to get off/out. I used needle nose pliers and a utility knife to pry/break off lots of little pieces. It took me a while.

Currently printing the mount for Voron Revo HT, in Fiberlogy PA12-GF. The nylon can take a bit more heat than any PC.

Ideally this should be printed in something like Ultem 9085. That way it is self-extinguishing in case of a fire, still doesn't conduct, and the fumes from burning it are not human harmful - just in case the hot end goes crazy.

I'm hoping once the printer can handle hotter materials that I'll reprint the hot end one more time, for the (hopefully) final version. Bootstrapping... It is not likely to be a stalthburner at that point. I'm doing stealthburner now for expediency.

The printer is hot enough for some prints that I have to be a bit more careful loading and unloading filament, or messing with the nozzle. It's only going to get worse when I crank the heat up to the next level.

Gerald

PS - Not quite PIF quality and not PIF settings (it's overbuilt) or approved material, but it will probably do.

Edited June 26 by G_T

[atrushing]

8 hours ago, G_T said:

The nylon can take a bit more heat than any PC.

I just came across a video about some new Polymaker Fiberon filaments. Their PPS-CF10 is some interesting (but expensive) stuff! They claim it has a heat-deflection of over 250° C so it might be worth looking into. It doesn't look like they have it available on their site yet though.

[G_T]

That PPS should be nice stuff. There are other sources as well. ThermaX PPS for instance. Don't know if it comes in a CF version. It is within the range of what I could print easily at the moment, depending on whether I could come up with a good print profile for it. It's among the materials that interest me, in this case due to temperature range and chemical resistance.

I printed the more complex half of the printhead for Revo Voron in Phaetus NexPA-GF25. It came out usable from the looks of it. I gave up on the Fiberlogy PA12-GF. I kept having issues so I changed filament. I'll try again later. It shouldn't be moisture since it had spent weeks under vacuum when not in the hot print chamber.

The Phaetus filament is certainly a bit more abrasive on the outside. That won't matter for this application. What I found interesting is that it would air extrude a nice clean string at 300C (low end for that filament) whereas the Fiberlogy would put out a very kinky string. Interestingly I found the Phaetus printing better at the lower end of its recommended temperature range. So I used 75C to 70C bed, 300C nozzle, 55C chamber.

I ordered a spool  of Matterhacker's house brand PA-GF to try out later. I'm looking for a PA-GF that prints nice looking low warpage parts.

Printing the other side of the printhead failed with firmware shutdown. Twice. At the same point in the print. I'm thinking there is a gcode error in there. I may try again tomorrow but starting up a fresh instance of SuperSlicer and going from there.

Gerald

[G_T]

I still can't get nice enough prints out of PA-GF. They're pretty coarse. I found increasing layer thickness helped, but then the hot end couldn't quite keep up melting the filament! That seemed to be the reason for the firmware dropouts. Sigh...

I could of course print in the CF-PC I've been using. That would still be better than ABS, but not as good as high temp nylon.

So for now I ordered SLS PA12 printed parts. It's what I've been using, just for a different heatsink.

Gerald

[G_T]

I've been working on assembling the replacement PC-CF CW 2 extruder from the parts I printed. I should have done this a long time ago!

IF YOU BUILD OR WANT TO BUILD A CW2, YOU'LL WANT TO READ THIS POST

Since this is my 2nd one, I'm not exactly following the assembly manual. i think it misses some important stuff.

I'm using new Bondtech IDGA (HIGHLY recommend) and new thumbscrew.

I first assembled the guider arm or tension arm, without the roller installed. First thing I determined is the thumb screw spring washer is being used as a retainer for packaging and shipping. To do that, they essentially make the washer the wrong size. The ID is too small. It needs to be able to slide over the threads or the whole spring mechanism doesn't work. So I enlarged the inside of the nylon washer until the threads slid through fine. So now that tongue piece can actually move and return, using the spring as a spring.

Next I put the guider arm on the base plate in position, and moved it through its motion range. It was way too tight. I slowly filed/scraped away a little plastic until it would move smoothly all the way to the hard stop.

Next I took some fresh Bowden tubing and checked the fit at the top of the extruder. It was too tight. So I used a drill bit that was a few thou undersized in my bench drillpress and hand-turned the chuck as I fed the bit in. I stopped when it quit going easily.

Now the tubing fit in without being squished but was still retained. Squish can cause extra filament drag. Extra filament drag is bad.

Next I took a piece of smooth filament (PC in my case since it was first at hand) and fed it through the filament path. It was too tight. Sure, it would have worked, but by making the extruder work harder. And forget about flexibles.

So I took a very small very fine rattail file that had a diameter just very slightly larger than 1.75mm filament. I used it sort of like a reamer to clean up the filament path. Now the filament would feed through smoothly without all that excess drag.

After this I did the usual assembly up through attaching the guider arm and latch.

The latch... is made from PC-CF. It's fairly rigid. This is one part where straight PC would have been a better choice. A little bit of spring in the latch is probably a good thing. I had to remove perhaps 10 thou from the latch to keep it from being like clamping vice grips!

It is still quite solid. Actually the whole mehanism is rediculously solid compared the the flexy handle-with-care ABS version. The bolts that the instructions warn about not overtightening? Well you can firm them up. These new parts don't bend under the load.

On the original PIF CW2 I had to do a lot of plastic removal to get everything to sort of work. Very little was required this time around.

Now with the mechanism latched shut (reminds me of vice grips but not quite as aggressive) it only takes an ounce or two to pull filament through the extruder. But if I block the large gear with my  thumb I can't move the  filament. The internal drag is a fraction of what it was with the old extruder. It is so much smoother.

Without the motor mounted, if you hang the extruder by a lenght of filament and give it a push, it will progress down the filament a ways before it stops. This is with it clamped on a fairly hard filament tighter than you would probabaly actually run it! The filament comes out almost looking threaded at this tension.

Filament loads more easily too, and moving the filament by turning the big white gear is fairly buttery. So I highly recommend doing the bit of tuning that I mentioned above.

With the tension set at this level, I used a little allen wrench to back out the limit screw until it made contact, then advanced it a touch. That little access hole in the guider arm is quite convenient.

That's as far as I've gotten. Recovering from COVID.

Gerald

[G_T]

Replacement CW2 complete.

I used a high temp LDO motor and LDO's version of the two piece breakout board since that is what is in the printer at the moment. Note, since this PC-CF should be slightly conductive, I flush cut the pins on the back of the board. I also shortened the motor cable to just what is needed. And as you can see I don't use a cover. I think of it as extra weight and a heat trap.

It's funny how a cellphone pic takes fairly flat black prints and bumps the micro contrast sky high giving an appearance like it was made of sand or concrete. Sorry about the pics.

I'll likely install it this evening.

I think this just leaves the RamaLama2's to be replaced, and the Bento Box to be reprinted. Then I think everything inside the printer should be good up to roughly 120C chamber temp, and most things in there good for quite a bit more than that. Likely the fans die around then, and I'm not sure quite how much heat the Bondtech plastic gear can take. I wish it was metal.

Gerald

[G_T]

Everything is reassembled.

I did find, just like with the original PIF stealthburner parts, that I had to adjust the fit of the boards so they lined up right. The default just bolt it together alignment is off a mm laterally. Not looking when assembling could possibly get the pins wrong, probably leading to fried parts.

I filed down and shifted the fan board, and replaced the plastic standoff for the other board with two slightly shorter pieces of Bowden tubing. Now the boards line up right.

It turns out the nice CPU cooler I picked up won't work for my installation. It interferes with a little daughter board. So I've ordered a copper plate heat sink. It should be here in a few days. I'll look at installing it as part of the next round of upgrades that require going into the electronics bay.

I have some 60w heater cartridges arriving in a few days. It is already pretty obvious that a 50w unit can't keep up towards 330C and a 0.6 nozzle and I'm intending to go to about 400C. It won't even get there with a 50w unit. It might not with a 60w.

I'll probably print PC-CF mounts for the Dragonfly BMO that I have NIB. It would make a good breakin test for the replacement parts. It will get a 60w heater and a Pt1000.

Gerald

[G_T]

The mounts are printed and installed. These were the first parts printed with the new extruder. The improvement is considerable. The only real remaining print artifacts are some edge defects on some edges which I think are slicer related. I'll have to work on that sometime to figure out what SuperSlicer setting(s) are causing the issue. But for the parts I'm making to go in the printer it doesn't matter.

All the parts in the all-new print head practically snapped together. They all fit each other perfectly, and they fit the bearings, bolts, and Dragonfly BMO perfectly. This is way better than the original parts.

The BMO got a Pt1000 and a 60w heater. It definitely heats faster than the 50w in the Mosquito. I ran a PID calibrate but will need to re-run. I'm going to be putting a 1/8" ceramic insulation piece between the heat block and the heat sink to reduce heat transfer when I really crank up the temp. That should happen tomorrow.

Now everything in the hot end except for the Bondtech gear is rated for 170C. I have no idea of the gear rating unfortunately. I guess I'll find out if I melt it. If so, it will get replaced with a metal gear.

Hmmm, still some blue bits in the printer. That ABS stuff can't take the heat. More parts to print...

Gerald

[G_T]

I've put a 1/8" thick ceramic fiber paper piece between the heat block and the heat sink of the Dragonfly. This fills the gap, and should reduce heat transfer a fair bit. I figure that will be important for higher temperature printing while still using chamber air for cooling.

So far it seems to be doing fine, after close to 20 hours printing. However I haven't given it a detailed inspection. YMMV, use at your own risk, ...

Gerald

[G_T]

I've switched to printing Fiberlogy PA12-GF15 for a bit. This is the replacement Bento Box v2 that should be able to take a little more heat than the commercial PETg unit it is replacing.

I'm using a textured PEI sheet, and VisionMiner's Nano-Polymer adhesive. I had used Polygoo PA first but the brin came completely loose. The nono adhesive works slightly better.

I have a Garolite sheet prepared specifically for printing nylon. But I found my light bed clips couldn't quite reach far enough. And the heavier clips are probably a bad idea since I have a full-diameter bed heater. I don't want to pressure the heater!

I'll solve the problem likely by making some clips from heavy gauge solid copper wire and a little hammering. Copper is wonderful stuff. But in the mean time I've been printing. I could print this material a lot faster, at least twice as fast, but since the profile is working I decided to just run with it.

I'm almost out of the green, but am in the process of drying a spool of red. I hope to get all the outside done in the green except for the top grill. Then the internals and top grill in red. I think I like this filament. It prints well enough just has some warpage and adhesion issues (it's nylon, doh). I completely gave up on the HTN+GF30. I went through half a roll and couldn't get a single acceptible print.

In the Bento Box running in the printer lately, I've been using Noctura fans. They are quiet, but they don't move enough air for the filtering to work well. I'm not using them this time around.

All the parts are the same color. But the lighting inside the printer is more blue, and the light outside the printer is more orange (old bulbs). Really the color is a little more of a pastel light green, not quite so fluorescent.

Gerald

 

PS - It seems that my old SuperSlicer has problems with the activated carbon grate pieces. It loses the holes. So I might have to stick with those pieces being PETg for a bit.

 

Edited July 10 by G_T

[G_T]

One more bit of ABS has been removed - the wire guide that mounts on one of the motor mounts. It is now PC-CF. I had printed it quite a while ago but not gotten around to swapping it. Now it is swapped.

Pictures are with the printer's insulated sides removed. It is easier to work on that way.

I ran into issues printing the Bento Box parts in PA12-GF. Warpage of the base layer... That resulted in rocking fit of the big middle section that holds the filter media. I will be reprinting that part but I don't have enough filament left. So another spool of the green is on the way.

The bottom piece having some curve on its bottom didn't matter in this case. I have a strip of VBS tape under it to help secure it and reduce vibration transfer from the motors to the enclosure. That makes it quieter.

I also ordered a Lightyear G10 build plate. It is G10 with a magnetic backing steel layer so it can be used just like a regular PEI sheet. Nylon should have no issues sticking well to G10 so I expect that will greatly reduce my warping issues.

If I get the chance I'll print the internal media holder and the top cover from the red Fiberlogy PA12-GF I have on hand. I don't think warping would be an issue with those parts. We'll see.

In the mean time I went ahead and mounted the motor and duct section of the new Bento. I needed the stronger fans for adequate filtration. The Noctura fans I'd put into the commercial Bento just didn't cut it for the job.

The parts above the fan mount will be the commercial PETg parts until they get reprinted.

By the way, if anyone has any comments or questions, feel free to post! If not, I'll just keep on documenting like I've been doing.

Gerald

Edited July 12 by G_T

[G_T]

I have swapped out the x-y joint cable bridge with one printed in CF-PC. I have also completed reprinting and replacing the Bento Box parts, now in PA12-GF15. I've finally got that filament working fairly well.

To get the nylon to print well I had some problems to solve - bed adhesion, and warping.

I have a garolite sheet but couldn't get it to work well. It behaved almost as if it were contaminated with some oil at some point? Anyway the nylon would not stick to it.

I ended up using a textured PEI sheet, with VisionMiner Nano-Polymer adhesive. I had to clean the PEI sheet well, but then this worked. PROVIDED I upped the bed temperature! I ended up printing the various parts with bed temperature between 125 and 135C depending on the part. 135C is too high for the Fiberlogy filament since it causes the color dye to brown a little. 130 was marginal, so I'd say 125C is probably best for bed temperature.

When it came to warping, I made a few changes. Eliminated any use of cooling fan on perimeter. I had been using a very light fan on exterior. That's a no-go. I also added a draft sheild to trap heat better. Understand my chamber temp in these prints - depending on the settings I was testing - was in the range of low 70's to low 80's C. That's out of range of most non-pro printers.

Nozzle temp ended up being 255C first layer 250C otherwise, which is at the low end of recommended. But with the hotter chamber and the draft sheild the plastic has longer to cool.

This makes solid blocks of plastic. There is no weakness along layer lines. The skirt requires cutting off for instance.

Even with these profile changes, warpage still exists unless the parts are thin. Across the bottom edge of the taller Bento Box parts, there would end up being about 1/16" curve when the part was cooled and removed. I'd guesstimate a chamber temp of around 120C would be required to essentially eliminate the warpage.

About the Bento Box 2, I did not use the original pattern for the filter media grates. I thought the holes were too small and therefore overly restrictive of air flow. So I made them bigger. Even so I think the air motion through the filter is borderline at best. I have some higher static pressure fans on order that should be here soon. I'll swap them in and see if that helps.

Gerald

Gerald

[G_T]

I've replaced the Bento Box motors with the 4028 higher flow motors sold as upgrades by Voxel. They are as thick as one could fit in. They are also 24v, so I need to change a voltage jumper on the Octopus. Since I need to get into the bottom of the printer, I might go ahead and put on the better heat sink to the Pi. We'll see... depends on whether the existing heat sinks come off easily or not!

I could tell the old motors weren't getting the job done, by the thin layer of filament dust in the bottom of the printer. Wiping with a damp paper towel made it pretty evident. I'm hoping the new motors improve the situation.

Gerald

 

PS - Turns out I need longer mouting screws for the Pi with the heatsink change. Hopefully they arrive tomorrow. I'll report on the heatsink once it is functional. Also the new fans.

Edited August 11 by G_T

[G_T]

The heat sink is properly mounted and tested. It reduces the Pi's temperature by 5C. Not a lot, but useful. It is a passive heat sink so no fan to fail, and it connects to the chips on the top of the board, not just the CPU. The plate has some thermal mass so the Pi doesn't heat up as quickly. That probably keeps the peak temp down under brief high load situations. Just keep in mind if you use one of these you'll need 2mm self tapping screws at least 14mm and at most 16mm to mount it and the Pi to the usual plastic mounts. The heat sink is from Amazon - Geekworm 4mm Heatsink H402 for Raspberry Pi 4.

Very important! Since we don't use the standoff screws be very careful not to torque down on the 2mm self tapping screws we use for mounting! You could break your Pi. You want good contact. You don't want to be bending it!

I also included a pic of the next attempt at fans for the Bento Box. These at least manage to move a little bit of air through the filter and that does make a difference. But really I think the base should be redesigned to use a fan type which intrinsically has higher static pressure. I think these new fans work better than the others I've tried so far. They would be adequate with just the HEPA filter, or just the filter media. But the combination of both is a bit much for this sort of fan to handle. At least these new fans are not annoyingly loud like the originals!

Gerald

[G_T]

GRRRRR...  Ok, I think the Bento Box could do a good job. However I think the usual selection of fans are unsuitable for making the the thing actually work. The fans need to move air through a small HEPA filter which requires static pressure not CFM per-se. It also has to move the air through 5 grills and three layers of filter media. So there is a lot of resistance to air motion. There needs to be a lot of static pressure to force the air through all of this. Axial fans are really not ideal for this.

I'm tired of swapping and testing fans. The current ones, from Voxel, IMHO lack sufficient static pressure to really do the job. They are not however the worst I've tested so far, do manage to push A LITTLE air through the filters, and are not particularly noisy.

I've ordered a pair of Sanyo 9GAX0412P3S001 fans. Each costs more than a pair of the other sorts of fans I've tried before. But the static pressure is at least 3x higher and the temperature rating is also a fair bit higher. Hopefully the noise won't be too bad. The heavy insulated walls of my printer do manage to muffle sounds a fair bit. I'll report once the new fans are installed and tested.

I figure the Bento Box will actually be doing its job when the bottom of the printer stays fairly clean, rather than accumulating dust. We'll see. The only time the bottom stayed fairly clean was when I was running TWO filter boxes with quite whiny fans. I'm sure they were disturbing the neighbors. They were disturbing me! Too bad I no longer have those fans and don't know what they were.

Gerald