So far, this printer is a hunk of junk. I just started the return process. Idk how I’m going to take it apart I don’t have the patience for it, I want to just smash it. Let’s start with the bed, it’s uneven, way too textured, and has destroyed a few prints. I have to print thin backing which is the reason I bought this printer. Even messing with the settings and distance it’ll destroy part of the print because it’s so adhered it rips it to shreds or it’s so loose it’s like a thousand string banjo. It adheres so hard I have been scraping thr finish off the printer to get my print off. OK fine maybe it’s too thin and this printer is meant for big projects. I start a print, halfway through runs out of filament. After feeding air for several hours… great a whole new spool of material straight into the trash. Start a few more prints, inconsitiant adherence to the exact same model. Some are great some are dogshit. And the one that made me return it, the nozzle will no longer feed filament. After a few smaller prints that came ok enough I started another print on a larger model that was less than a full spool and could deal with not being perfect. Just randomly stopped halfway and not feeding any filament. I made no adjustments, just stopped. I can force feed it through by hand but other than that the gear spins and nothing comes out. I messed with the knob and was able to stick a tool in the print head to clear it but nope, still won’t feed anything. I’m not gonna take apart a brand new print head to clear a problem that shouldn’t be there. I’m sending it back. I hope whoever designed this got fired. What a hunk of junk. 1300 bucks, hours of wasted time, material, now I have to take the piece of shit apart to mail it back wasting more time. After this experience I feel sovol owes me money for beta testing.
You don’t mention what filaments you are using, steps you have taken to tune the profiles, etc. Sovol could certainly be a little more proactive in providing pre-tuned profiles, better customer service, longer support after sales and so forth to smooth the out-of-box experience, but then, they’d have to charge like Bamboo. The MAX is best thought of as a pre-assembled Voron kit: fit and finish is up to the user to complete. Or think of it like a musical instrument – it can only sound as good as the person playing it.
There are lots of examples around Youtube etc. showing the MAX can lay down a good single-layer 500 x 500 sheet of plastic. If it’s tuned properly. While I have had some failed prints along the way, I get good results by walking through each of Orca’s calibration routines in order.
Given your inconsistent first layer results, I’d start by checking the Eddy Current sensor mounting bolt. I’ve had it vibrate loose and confuse the heck out of the printer by giving inconsistent readings.
It’s always the same story. It’s always the printer’s fault, never the user’s.
Often, the problem lies somewhere between the chair and the printer.
I printed for years, I know how printers work and I understand them, Im also a machinist so I know how to take things apart and troubleshoot. It appears the filament feeder does not work, I’ve taken the entire printer head and filament sensor apart. It’s junk, poor quality control and poor engineering. There’s plenty of people making the same complaints as me. Maybe you should worry about keeping what’s between your chin and the empty space in your head closed.
For starters it’s a simple tinkercad file. My other printers print it no problem I’m just able to print multiple on this one. Or should be able to. It’s elegoo pla, and it can’t even print the stock benchy it starts with. I’ve taken it all apart and put it all back together and nearest I can figure the filament sensor is broke. It’ll feed material through the print head if I force it, it won’t pull any through by it’s own print head power tho. The extruder gear is fine the print head seems to be in working order which it should be brand new. The filament feeder tho seems to fault out and not be able to feed filament when it’s supposed to. I can’t disable it because the printer cant pull through all the ptfe. It’ll print no problem then just stop feeding any filament. No clog, nothing. There is no possible way to lay down a single sheet on this printer. My print is 1.2mm thick and it SHREDS it pulling it off I don’t know what the bed is made out of but it’s awful. Give me standard pei plate any day.
If you are not satisfied with this printer, why not contact Sovol by email at info@sovol3d.com, providing your order number?
They will respond within a week.
The complaints about the “buffer” are widespread. I don’t own a Max but follow this (and a couple other) forum.
If I had a Max the first things I’d do:
- Ensure the tube exiting the buffer is free to move up/down to allow the telescoping section in the buffer to extend/retract.
- Put a stack of 6 to 10 M6 nuts around the tube at the buffer outlet (be sure to put a clip under the release ring) to provide “gravity assist” for the (too weak) spring inside the buffer.
Ok, first of all let me mention that I’m not a half bad machinist myself. And as you know, a good machinist can get tolerances down to tenths even on a machine with several thou of slop. Sometimes it’s as much art as science.
It sounds like you have a couple of issues here, most of them pretty familiar to the rest of us. While most of Sovol’s hardware is well made, that auxilliary feed buffer is less than perfect as released, but there are several options: stretch the spring, replace the spring with a stronger one, add nuts for more weight or just delete it. I personally replaced the spring with a stronger one and I use DKEU (Demon Klipper Essentials Unified) from Github to disable clog detection on prints I can supervise.
Next is that long Bowden tube. No way around the length really; it’s a big printer. But Sovol supplied tube with a 2.5mm ID. 3mm PTFE tube is a much better choice because a) it’s larger diameter reduces friction and b) PTFE itself offers less friction. Just be forewarned: there are vendors of Bowden tube claiming their product is PTFE when it isn’t. Real PTFE cannot be made clear, so any claims it can are BS and a sure sign that product is made of something else. Buy from a reputable source.
Third is your print bed. First of all, Sovol’s plate is real, textured, PEI. But it needs to be washed with a gentle detergent then rinsed with 99% ethyl alcohol before first use and periodically thereafter to remove any residue. Plastic won’t adhere to dirt, dust or oils. That massive print bed also comes with a price: preheating. It’s sitting on top of a comparatively thick piece of aluminium which takes some time to heat soak. During that time, it’s expanding and changing shape. One of the DKEU macros helps by sampling the bed plate at the area of greatest change (as determined by the user for their specific machine) and monitoring the changes with repeated sampling. It won’t start printing until the bed stops changing.
On that note, the bed is held down by 8 screws. One set of 4 is near the centre and should be tight. The other set of 4 is closer to the outside corners and they are best left “snug”. This allows the plate to expand outwards rather than up and down, keeping the bed better levelled.
On to software issues: Sovol’s software sucks. No doubt about it. They provide a custom fork of Klipper that has not been maintained and that darned buffer software is incompatible with the current version of mailine Klipper. Their default acceleration settings have been fairly described as “insane” – 40,000 mm/sec! Driven by the marketing department striving for a fast Benchy, no doubt. I found layer shifts were almost inevitable at anything over 10,000 and personally have things set to 8,500mm/sec as the absolute maximum.
Fortunately there is a very active MAX community centered around DKEU, which fixes most of the problems through custom macros.
Using either DKEU or Orca slicer, you can adjust the textured PEI plate offset. Here’s the issue: the eddy current sensor detects the metal plate. The tap detects the top of the texture’s peaks, not the bottom of the valleys. So the wrong offset leaves the first layer in partial contact with the bed. Adjusting the offset slightly forces the filament down into the valleys where it (PLA in particular) sticks like glue until it cools.
Speaking of temperature, correct nozzle temp is essential! Next problem with Sovol’s software: they’ve screwed up the conversion between thermistor voltage and actual temperature. I did extensive testing and others have also noted that temperatures must be set 10 - 15°C higher than normal to achieve the desired nozzle temp. I re-wrote the printer.cfg section to correct the readings on mine based on my measurements. I can’t guarantee that all machines have the same precise error – Sovol may have purchased sensors with a broad range of manufacturing tolerances. But both my original and my spare extruder have the same error so it’s worth investigating.
On the topic of extruders, there can be slight differences between eddy current sensors. It is imperative to run Sovol’s eddy current compensation routine before first print and any time one changes extruders. I even run it when changing a nozzle, just in case.
Another issue: Sovol’s filament change routine can itself cause problems when filament is re-loaded. It retracts the filament immediately from the hot nozzle which can lead to long hair-like strands in the Bowden tube. This causes jamming and filling of the extruder with plastic debris, especially with the stock 2.5mm tubing. One can manually retract (NOT unload) the filament, disconnect the tube at the extruder, remove the filament and snip it off before retracting. Or DKEU’s replacement macro will retract the filament a bit then wait for it to cool below the stringing temperature before unloading if you prefer a fully automated unload.
There is one mechanical issue with the extruder: the darn roller wheel floats and can shift sideways so it doesn’t grip the filament properly. Again, the community has a simple fix that can be printed in 10 minutes and keeps the roller where it needs to be.
In conclusion, then: the MAX is less than polished. The out-of-box experience is sadly not on par with Bamboo. Sovol should be embracing and incorporating the improvements developed by the extensive community that has sprung up around the MAX rather than continuing to ship the same thing they provided us in the pre-order phase. That is not the same as saying the printer is “a hunk of junk”. With a bit of time and care this can be a great machine! I am regularly printing large objects like equipment cases and speaker cabinets where the first layer is every bit as – or even more – important than when printing thin panels. One printed overnight while I was away. I’m printing another as I type this (at 50cu.mm/sec, I might add!. My buffer is firing off about once per second and will be for the next 18 hours):
I have to chime in, again, on expectations. R&D is expensive! The R&D cost has to be factored into the price of the printer with the formula being total R&D budget/units sold. Sovol cannot spend like Bamboo does and hit their price targets.
@MoldyKumquat You bought the low cost option, and then complain you got what you paid for. Bamboo won’t build a printer the size of the Max because the market segment is too small. The other players in this size class are 2x to 5x more money. Logic dictates Sovol will need a little post-sale R&D. The community is actively crowd sourcing that R&D. You have to be willing to apply the fixes for the issues you are facing.
I have to chime in, again, on expectations. R&D is expensive! The R&D cost has to be factored into the price of the printer with the formula being total R&D budget/units sold. Sovol cannot spend like Bamboo does and hit their price targets.
Sovol doesn’t have to. The community has done most of it for them. What they need to do is a) update or release the code for the buffer so it can be incorporated into mainline Klipper and then b) adopt all the other improvements made by the community instead of ignoring them. For example, I provided them with all the data I collected regarding the thermistor issue many months ago. Demon has provided them with lots more fixes and requested the buffer code so he could fix it. The buffer has been a problem since day 1. All ignored.
The hardware is basically solid. A little more attention to early adopter feedback especially on the firmware/software side and they could be a solid competitor for Bamboo at their current price point. People who buy Bamboo products are not the tinkerers and innovators, but that’s the largest market share and will continue to expand exponentially. Sovol needs to wake up and smell the coffee.
This is an actual pain point that @sovol3d needs to address. Not making their Klipper fork public is a GPL copyright violation and inhibits crowd source developmenf.
My guess is Sovol contracts their firmware creation to a 3rd party (MKS?) with no maintenance in the contract. Sovol has no one on their support team is able to access the source code. Firmware updates would require a new contract with the actual developer.
Their PLR system has been broken, with an “illegal” file system write inside the motion planner loop. Results in “timer too close” errors. I’ve posted a patch and got a heart from @sovol3d but no official action.
I did, I got one reply and that was it. The reply was not helpful at all they asked for a video. Mf how am I supposed to show you a mistake that happens sometime during a print it’s not like it happens at a specific time.
I had a similar issue where I couldn’t record a video.
I recommend you follow up with Sovol and let them know that you can’t record a video.