Qidi X-Max 3: First impressions

3D Printing
41

I never want another bed flinger. That particular hardware cost optimization was a reasonable compromise back in the days when 25mm/s was fast. The marginal (hardware) cost of a corexy over a bed flinger isn’t really that much as a percentage of total cost of the unit, and Prusa’s attempt to value-price a corexy at 2-3x the competition who are delivering good results seems unlikely to succeed. I can’t imagine even thinking of buying the new Prusa MK4. It’s $1300 for a bed flinger — with a small print volume. And then the XL starts at $2K; that’s their entry price for corexy. :open_mouth:

When did Prusa first say they were working on the corexy XL? Let’s see, it was announced November 2021 and didn’t ship first units until a year and a half later in May 2023, to the best of my understanding. It’s not like there weren’t very good open source designs for them to crib learn from. And then it’s priced very premium; how are they defending that premium?

They had to work hard to make their reputation for quality, and I respect that. They have invested in slicing, and PrusaSlicer has been a great advertisement for them and a value to everyone including their competitors. But they themselves benefited greatly from the work of others, including competitors, and that including recently (e.g. the Arachne perimeter generation engine), as well as bootstrapping them earlier. They haven’t earned the right to ignore innovation on the part of others.

I don’t understand the folks who make excuses for slow printing; I keep reading comments from people who say that the slow (by today’s standards) printing they are experiencing is fast enough for them and therefore it should also be fast enough for everyone else; that wanting to print faster is somehow a moral failure. Feels like “Stockholm syndrome” to me every time I see it.

By the time I’m done with my Qidi, I expect it will seem quaintly slow. But right now, it’s in the ballpark of the rest of the class of new, fast printers.

4 Likes
42

Yes, my only remaining bed flinger has a 500mm^3 volume but feels like a horse and buggy, now. I may make a few more big parts on it but it’s probably faster and more reliable to chop up into smaller prints for my coreXY machines and adhere the sections together afterward.

1 Like
43

I manage to sit firmly in the first camp, but totally reject the second. This is a hobby for me and the machines I’ve got do what I need. But I’ll never tell anybody else that speeding printing up is a bad thing! Quite the reverse; it’s wonderful to see the progress that is being made everywhere in the FDM world these days.

As you note, this is going to improve further and I’ll be curious to see whats on the market by the time I feel like upgrading. If I was printing commercially I’m sure I’d be a lot more aggressive. :wink:

1 Like
44

Koz Ross has been much less impressed with his X-Plus 3 and has been talking about both successes and failures on Mastodon. He got a heat creep jam while using the hardened nozzle to print PLA:

He discovered something rather unexpected about how the toolhead is held together, at least on his unit: The screws that hold it together have 1mm of engagement in plastic threads.

I haven’t disassembled mine to check whether it’s a fluke.

I haven’t printed much PLA, and I’ve printed it only with the plated copper nozzle it came with. I’ve printed mostly ABS, and haven’t yet actually printed with any of the carbon fiber or fiberglass filled filaments I bought because I keep printing things that don’t need them. :grin:

45

I wonder if he needs to add an “air assist” tube down to his hotend’s cold-end cooling fan so it can get cooler room temp air instead of chamber air which will be far hotter? Because of course you’re going to get heat creep up the transition zone when your cold-end is 40C hotter than running the chamber unheated.

46

Well, the point would be that on what is supposed to be a plug and play printer he shouldn’t need to.

The printer reminds you to leave the top and door open when printing PLA, so I doubt he’s trying to print those in a closed chamber.

1 Like
47

I jumped to the conclusion that it was not PLA but was ABS since acetone does not effect PLA and does greatly effect ABS. Therefore, he was using the enclosure all closed up and the hot air in the enclosure was/is the only air which is “cooling” the cold-end of the hot-end assembly.

Ya, it should work if shipped as a turn-key system but when it doesn’t I say ‘how can I fix this’…

48

Actually, I think I crossed wires, and that you are right. I was surprised by the acetone and thought it worked due to a mix of filaments, but no, when I went to do a search I didn’t find him saying anything about PLA after all. My mistake.

I’ve done almost all my printing in my X-Max 3 with various forms of ABS with the Qidi profile for “Generic ABS” and I haven’t seen anything like what he ran into. And I’ve been running a 100° bed and 50° chamber for all of that. I just haven’t been using the hardened nozzle for it.

49

Hot end design is a tricky job and not only is it about the shape of the filament shaft down to the heating chamber it’s also about how smooth that is. years ago I saw many posts about people who would have problems with ABS if they hadn’t first ‘seasoned’ their hot end first by putting oil down the hot end and brought it to high temp. I doubt anyone does that now but back almost 10 years ago, it was the wild west with regards to what hotend designs were. And lots of times vendors just ran the PTFE tubing right down to the hot chamber to limit the clogging but it couldn’t work with nylon and other high temp materials.

I had a custom one made with PTFE only in the cold side and the transition zone down to the nozzle was metal.

Top that off with how long the transition zone is, how hot the chamber is and how big the cold end heat sinks are it’s almost a magic formula.

50

The Qidi is an “all-metal” design. It has a thin metal heatbreak and does not run PTFE right up against a nozzle.

51

Ya, most today who state they’ll do ABS and other higher temp plastics are all metal. 5-7 years ago that was not the case. Internal designs which worked were copied and now are wide spread. E3D likely had lots to do with making what works.

52

Well… I guess I jinxed myself. :roll_eyes:

This morning I came down to an overnight print that stopped printing partway through, and it wasn’t out of filament.

It was a definite heat creep clog in ABS.

So I have a nozzle and heat break soaking in acetone now, and a replacement set mounted.

I’m going to say that now that I’ve actually changed the nozzle and heat sink on my printer, I understand completely why Koz didn’t want to change back and forth between nozzles. This printer was not designed for frequent nozzle changes. If I really want to print abrasive filament, I need to get my other printer set up to do most of the printing with non-abrasive filament.

1 Like
53

I previously had bad print adhesion using the unprinted side of the print sheet. I have had good print adhesion using the printed side, but when I print parts on top of the white lettering on the sheet, some of the white comes off on the print. I don’t like that.

When I first complained to Qidi support about the unprinted side of the print sheet not sticking, they suggested that I should re-adjust Z and re-probe the bed after reversing the print sheet. :thinking:

I finally did that since I was re-tuning after changing the nozzle (for one that wasn’t clogged, sigh), but I still had the same problem: Even with the bed at 100°C and the chamber at 50°C, ABS prints warped and didn’t stick to the unprinted side of the sheet.

This is frustrating.

I’m soaking the original nozzle in acetone to remove the clog. I’ve been removing the resulting ABS slurry a bit at a time, but I’m still having trouble getting a 0.4mm needle in the pointy end of the nozzle. I’m wondering whether I really had a clogged nozzle result in slow heat creep, rather than a heat creep block resulting in a clogged nozzle. I was printing with seven-year-old filament.

With a brand new replacement nozzle in place (I bought extras along with the printer…) I’m now printing the same print but with fresh filament that I just opened, and I’ll see whether it works better.

Changing nozzles is more trouble than I expected, out of the box. They ship with only one cooling fan that you are expected to unscrew from the cooling block previously installed in the printer, and screw onto the other cooling block for the hardened nozzle in the other assembly that shipped with the printer. Then there are two screws to hold the hot end to the carriage. And four more screws accessible only from behind allow you to remove the back of the carriage shell, under which there are at least two connectors to disconnect (one of which was glued in place at the factory), three connectors if you are changing the fan (the replacement hot ends come with replacement fans).

The shell on the back of the print carriage serves no obvious purpose, so I have just removed it and re-tuned resonance compensation with the slightly lower mass.

With an extra fan in place and the back cover permanently removed, that means removing and replacing two screws instead of eight in order to swap hot end assemblies. Then adjusting Z offset, of course. But that’s a pretty quick process.

1 Like
54

Support had trouble even understanding this problem. They do list the sheet, sold separately, as double-sided, and you are indeed suppose to be able to print on both sides, but they didn’t come up with any fix for the print failures on the un-printed side. Instead, they offered me a coupon for a free “smooth” print sheet (PEI without the manganese additive in their default “HF” sheet) which doesn’t have any logos painted on it, and so there’s no graphics to adhere to the bottom of a print.

I’ve now been able to pass a 0.4mm needle through from the tip, but there’s still lots of filament gunk left. And the needle doesn’t pass through smoothly, which makes me think that it really was clogged.

Regardless, It’s not clear to me how the heat sink is supposed to be sufficiently cooled. The heat sink fan is entirely inside a shell that doesn’t have much in the way of holes to let air flow through.

I’m wondering how long I go before I give up on staying stock and design a new cover for the hot end that improves cooling for the heat sink. :roll_eyes:

1 Like
55

My clogged nozzle approach is to take the nozzle on its own, hold with nose pliers, and hit it several times with my hot air gun set to its 650°C max. Not high enough to melt the nozzle, but high enough to ensure all plastic gets carbonized.

Then I brush them clean with toothbrushes and dental floss in the hole :wink: and finish in the ultrasonic cleaner. This hasn’t failed me yet, I saw this suggested in a random forum post years ago. Except the dude there was using a blowtorch.

Edit: I do this for heatbreaks too. And I take it easy with the floss, no pressure.

1 Like
56

I’ve fully cleaned the nozzle with another day in the acetone, and when I run the needle though it scratches like there’s a piece of dirt trapped.

This nozzle is now sitting on my mill, so that I can slice it apart and see what it is like inside. This is why I bought extra nozzles. :smiling_face:

3 Likes
57

I bought their custom webcam and installed it, thinking that it would be an improvement on what I bought.

When the printer homes in Y, it runs into the webcam. I tried twice to reinstall it to see if I had just made a mistake, but no, Qidi just sold me a camera that doesn’t work with this printer. apparently can’t write clear instructions

58

This is a sponsored video, so take it with a grain of salt…

59

There have been a gazillion videos on how baking( annealing ) your parts after 3D printing makes them stronger so why would it not make it stronger if the part was literally baked while it was being made…

Makes sense to me although the temps for good annealing are way higher than you want to run your chamber at so a hotter chamber is probably not as good as a full annealing process.
But that’s just a guess.

60

After a couple round trips with support, they said I had installed the camera on the wrong rail. Mounted on the correct rail, it does seem to mechanically fit. Haven’t had a chance to test it yet. [Edit: It works, but I’m not sure whether I’ll end up using the official one or the one I installed first. It’s not obviously better, in any case.)

The installation instructions are poorly written and are mostly a set of postage-stamp-size thumbnail images of black on black interior. The combination made it not clear that it should be installed close to bed height.

3 Likes