Tronxy X5sa Pro modified with NF Sunrise, enclosure, and Klipper 

as you said, because it's a flawed printer. People who don't like to tinker don't buy a printer that needs tinkering, and people who like to tinker buy a printer that works first of all, and then tinker with it

@@HitLuca94 you’re right. I think everyone should own a Prusa as their first machine (Mk3 or the latest-and-greatest Mk4) - or maybe even the Bambu Lab machines - yes, they are expensive to buy, but they work, and they give you a baseline experience and the ability to print parts which you can then use to modify some other flawed, cheaper printer like the Tronxy. To put it another way, everyone needs a small (reliable) printer before they take on a large (unreliable) printer :) My Tronxy in the end has printed parts the others cannot. It’s been worth it, but it would never have worked as my only machine. It also printed all the parts for my Voron 2.4, which has proved to be a fine replacement for the Tronxy but again, could never have worked unless I’d had another working printer first.

Hello - I remember the fan only just clearing the Y-axis carriage, I think my favourite was the Prusa-style fan mount/duct that you can get from here: www.thingiverse.com/thing:5150173

Ah yes, I remember looking at that model (from TwoTrees) before buying the Tronxy. They seem a nice machine in many ways, linear rails and a strong-looking frame. I reckon you should be able to get good results. The standard hotend looks horribly cheap. I think some versions have the Z screws linked by a belt, and some don’t. Seems best to link them. If it’s still using a remote extruder/Bowden tube then I reckon start with the direct extruder (like the NF Sunrise), I think that makes printing a lot more reliable (gets rid of all the stringing and gaps or blobs, clogging, etc.) Then tackle the mechanicals - belts tensioned and idlers running smoothly, that sort of thing - and finally run it with Klipper. Gives better results for the same hardware, as the print moves are all calculated pre-emptively (it looks ahead) - can use input shaper tuning to cancel out resonances.

Hello, please see www.thingiverse.com/thing:5240585 - thanks

I think you’re better off going down the Voron path which will get you to 200mm/s easily in a 350x350x350 size, but otherwise you need to build a VZ-bot. There is now a kit for the VZ-bot - approx. US$1250. I suspect the only way to get 400x400x400 will be to start with a regular Tronxy X5sa 400 (non-Pro version without the metal wheels) and fit linear rails, VZ-bot parts, new control board, Klipper etc. The modifications I’ve shown here will only get you to about 150mm/s max. Good luck with your build :)

Going down the voron path... go Vzbot. It started as a modified x5sa... it's the fastest printer on the planet of its size. And yes, I get well over 200mms on several modified x5sa. Go linear rails, simple hotend, move the z rods apart as fast as you can, and move them in to the frame a little so you can enclose the frame easy. Also I've got 350x350x420, in the standard x5sa frame. My highly modified x5sa gets 20k acceleration and 1000mms travel speeds. I can print up to 40mms³/s throughout on my hotend, at whatever speed I like. 100. Not a vz, but my own design. Of course with a highly tuned slicer profile.

I’m sorry, I’ve only just started using Superslicer with my Voron 2.4, haven’t tried setting up the Tronxy yet. I reckon, though, that as long as your printer.cfg is safe and correct (in Klipper, in the Pi), the slicer settings won’t take you long to get right. Set retraction to 1mm and 40mm/s first… then set all the speeds to 80mm/s, and work your way from there - I don’t reckon it will take you 300hrs to get it working this time around, as you’ll remember what works and what doesn’t :) but I do sympathise, it’s very unsettling when all those carefully-refined settings are gone!

Pretty well - Klipper shows a Heightmap where you can see the shape it’s seen. If it’s a saddle-shape, check that the top frame sides aren’t over tightened down onto the smooth rods - that will bow the frame sides and make the bed levelling think the middle of the bed is ‘high’. Add washers to take up the gap between the smooth rods and the top frame. Pre-warm the printer for half an hour before probing the bed. Adjust the bed manually to get it as flat as possible on the visualisation. Check the vertical scale - 0.15mm variance is OK, the auto-levelling will take care of that, and shouldn’t need re-probing before each print

Hello - I’m using Gates-branded 2GT belts; I think they are genuine. Good-quality belts are flexible enough to wrap around 20mm pulleys without creating a load of friction, while also having the teeth surfaces treated with a fabric (in this case a brown material) to improve their wear resistance. Make sure to order enough length! a.aliexpress.com/_m02BG1O

Fair enough, I’ve actually just finished building my own Voron 2.4. It’s much neater, works generally better… but of course, it was more expensive to buy :) consider the Trident too

Thanks - yes that’s fair I suppose, it’s the equivalent of a Prusa or Voron setup, with about 20mm of Teflon tube between the filament feeder and the top of the heatbreak. It would be hard to get it any closer

Thanks. I don’t have time to make a video about the Klipper install, but I suggest you do a basic setup of the Raspberry Pi OS (Lite, without desktop) and then use KIAUH (the Klipper Installation And Update Helper); just go to the KIAUH page on your PC, then copy and paste the supplied commands into a PuTTY terminal window that’s logged in to your Raspberry Pi. From KIAUH, you can then install Klipper and Mainsail (or Fluidd if you prefer) really easily. For setting up the printer, see gist.github.com/cab404/b7bcbb0cd592a14515493694719de59b

Hello Marcus, thank you for your comments! I feel that you will definitely get better results with Klipper, especially with input shaping and pressure advance. But anyway, I would suggest the MGN12H rails for you as they are probably lighter than Tronxy’s rails. I will say, though, that the worst linear rails are sometimes gritty, where the Tronxy rails are relatively smooth. Anyway, I will be rebuilding mine with Hiwin rails (the best type). I hope you get nice results from your large 400-size machine!

That was the Tronxy TR sensor (black), which came with the coated glass bed. It was a capacitive type (detects anything). The blue sensor that came with the printer was an inductive type (detects aluminium, not glass). I’ve built a VZBot now - I just use a Z-stop microswitch on the frame and level the bed manually. No probe at all :)

@@alexshepherd as a mainly PETG user there is real no good solution for this stuff haha, cannot use PEI, glass you have to use gluestick or hairspray and other bed materials either grip way too little or are far too expensive, maybe i will have to get one and see if it works with my glass bed.

Thank you, wanted to share what’s worked for me :) designs for the parts I used are www.thingiverse.com/thing:5240585 and from there you can browse my other items, not sure what you mean by ‘loops’ sorry!

@@alexshepherd Alex, thanks for the link, I meant door hinges on the leaves, this is the sixth photo on the list. Probably do it yourself, but if there is a ready-made solution, I will be grateful to you. Shake your hand.

@@alexshepherd Alex, I'm sorry, I did not immediately pay attention to your description, I found these at the link. Thanks again.

Thanks! 8mm thick www.bunnings.co.nz/twinwall-1200-x-610-x-8mm-clear-panel_p0165356

@@alexshepherd Appreciate the reply, still getting round to it! how tall are those side pieces of twinwall, not sure how much clearance to give

Cheers, I appreciate it :) I have no need to push for ‘likes’ and ‘subscribes’, will never have enough views or followers for RU-vid to care. All I like to know is that someone, somewhere might get better results out of their Tronxy :) In original form they’re not really much good, but with the three main modifications shown here (the sync belt, the direct-feed extruder, and Klipper firmware) they work much better. Then, an enclosure makes it possible to print in ABS (or large prints in PETG) without warping, and suddenly it’s a useful machine!

Or 48$ for 4x8 abs panels from a local plastic company, or online 😏

I looked up the photos on the scales for you - 112g off the X rail, 16g off each of the three carriages (one X and two Y) so about 160g saved all-up by drilling holes

Direct extruder motor, with original hotend, bracket and fan was 377g, NF Sunrise with fans is 258g (so another 119g saved by changing to the NF Sunrise)

The holes are bigger than they look because they are drilled from the back, with a step drill - so the holes just touch each other

Very good! The only thing I would suggest beyond what I showed in the video is to consider a newer hotend design - the OMG H2 looks clever because it can bolt onto the same mounting holes, but you do need an extruder as well, so the NF Sunrise might still be the neater solution. The Tronxy rails are still working fine and I think without spending lots of extra money, your setup will get you good quality for large prints up to a speed of about 100mm/s. The frame bracing aluminium corners are the most effective, also consider doubling up the vertical rod bearings as I have (it is cheap and easy) as it stops the bed from wobbling. My next move is going to be a VZ-bot build, I just haven’t got around to it yet. There are many parts to order - lightweight rails (if you want smooth movement they are expensive), new controller board, better stepper drivers, so when you look at the double-cost to go that little bit faster, I think you will find that the standard X5sa Pro with the three modifications I said in the video - NF Sunrise, sync belt, and Klipper - will be a good-value setup for you. Running Klipper and Mainsail makes a big difference to the usability of the printer, it makes it just like a Voron to control. I really recommend that. You don’t have to use a touchscreen, you can control it through the Web interface from a tablet or phone (it seems odd at first to have a printer with no control panel, but it does work). I have a Voron 2.4 as well as my Tronxy. You can never have too many printers. The Voron is good but it is not really any faster than the Tronxy nor any better print quality, because the moving parts are heavy. It’s only advantage is that the first layer levelling is a bit more reliable with the quad gantry setup and heavy, flat buildplate (but again, more expensive as you know). I am running a Rapido hotend in the Voron which heats fast and is easy to change the nozzle on. But I think the most important is to use good-quality plated nozzles, like you can get from the Mellow Store (AliExpress) where you find the NF Sunrise. Consider that with a large printer you are probably optimising for prints with a 0.6 nozzle (strength and speed) - PETG or ABS, enclosing the printer is necessary, but part cooling is not as important - for detailed small prints, you have your Prusa-Mini clone. Thanks for your comments :)

@@alexshepherd Nah, thank you. Lots of food for thought. 👍

Well, seems i was unlucky and lucky at the same time. The seller sent me a X5SA-400 Pro instead of the X5SA Pro i ordered. I already constructed a nice table with wheels, bought and cut plastic glass sheets, so this larger one should go back, but the seller is making it complicated. I may have to use Ebay to get this fixed. At the same time i saw the X5SA Pro at Amazon for 250€, comes with the Z synchronization kit too. Bought it, should be here soon. I may end up with a spare X5SA-400 Pro if Ebay decides it so, but in that case i figure i can sell it myself, got it for 320€. 😅 Since there isn't much to do right now i keep printing flexi rexis for christmas on the Prusa Mini clone and read up on Klipper, extruders and hotends.

@@citronski I think you’re making the right choice. Giant printers are a nuisance in three ways; the expense of the print surface (and heating a large bed), the lack of frame stiffness, and the space they take up in your workroom, garage, etc. I hope you manage to return it, if it was not as described - it shouldn’t be up to you to resell it

@@alexshepherd The seller was uncooperative, i went to Ebay for help. Ebay decided to just give me my money back. When i asked what to do with the wrong printer, they suggested to throw it away! So now i have two, a X5SA-400 Pro and a X5SA Pro, for 250€! I'll keep the 400 for a few weeks, in case the seller wants it back, but after that I'm selling it. 😅 What a muppet that seller was! So i built the X5SA Pro, no problems so far, if you have built a printer before it's easy. I think even a newbie could do it. Printed a benchy, it's acceptable, far from what i get on a Prusa Mini clone, like what i got from the Ender 3 back in the day. The long bowden tube is a real problem. Klipper is next on the agenda. I have decided to go for a Volcano clone hotend and a Mellow Sherpa Mini direct extruder, both are ordered, as well as a bunch of nozzles and different heatbreaks. That's it from me for now, Cheers. 😄

You’re very kind; it’s hard to make a good video in one take when there is no editing possible :) I think the maximum practical acceleration is similar to the Voron 2.4, about 5000m/s^3

- Tronxy lattice glass and TR sensor - Gates belts - Additional LM8UU bearings for Z rods - Tronxy Z-sync belt kit with pulleys - Mellow NF Sunrise, extension wiring for extruder motor - 5015 fan, also replacement 4010 fan for the controller box (use ball bearing fans) - Raspberry Pi 3B+ back when they cost NZ$80, 7” HDMI display - several sheets of ‘twinwall’ polycarbonate (sold as verandah roofing) - various printed parts and brackets to relocate spool holder etc. which are listed on my Thingiverse page www.thingiverse.com/alexgsofnz/designs That’s about all I can think of right now

Hello, I can’t reply with a photo here, but facebook.com/groups/169728600253557/permalink/1057215468171528/ - I used a step drill, holes are several sizes with the larger holes on the back. Weight of the X carriage rail went from 591g to 480g, total weight saving 19g off X moving mass and 162g off Y moving mass. You would save more weight by changing the design to use linear rails, but it is a wait for them to arrive :)

@@alexshepherd yeah, i planned to change into extrusion and linear rail instead. But it need to wait cause i want to change my extruder to nf sunrise first, then change my board to btt octopus, and after that i change my xy movement to linear rail.

There’s only one? It’s running on a Pi Zero 2W The interface is Mainsail and Klipperscreen I recommend KIAUH - the Klipper Installation And Update Helper. Go to the website, copy the command into a PuTTY terminal window (that’s logged into your Pi) and it’s pretty easy from there. The printer-specific config details can then be searched for and found (e.g. search for Tronxy X5sa Klipper)