So... I finally got a round tuit (I don't have enough tuits to go around, and most of them are square). It's only been 5 months since delivery!
Bear in mind that mine was a middle-to-late Kickstarter backer order, so although production was ramping up it is probably not totally representative of what you would get if you ordered a U1 now. It is a little annoying that the price hasn't really gone up much from my Kickstarter pledge, but I think it is about to – and anyway, c'est la vie.
In fact, backing the project on Kickstarter did not absolutely guarantee me a printer at the end of it, but my assessment was that Snapmaker had always played fair in the past and would be dragged through the mud if they defaulted now.
Unpacking
It is beautifully packaged. The packaging is strong and the padding is high-density, with plenty of transit tape. The printed Quick Start Guide is also very high quality, and although mine was supplied with two amendment sheets I noticed a few inadequacies in the instructions which are largely addressed in the on-line download from Snapmaker Support, and I imagine units ordered now will arrive with an updated Guide.
In particular: although the carton shows it as a two-man lift on the outside, the Quick Start Guide illustrates lifting the unit out of the carton by one person! Let me be clear on this: it is possible (I did), but I would not recommend it. The crate itself is just about OK to lift by one person with care (it has hand holes for the purpose), but getting the contents out requires reaching over a very large box, grabbing the polythene bag containing the contents, lifting it out while trying to leave the box behind (the box wants to lift too, because of the suction), and lifting the contents high enough to get them out of the box – hopefully onto a waiting table. This would all be easier, and particularly much safer (both to you and to the printer – at risk of being dropped), with another person or even two people to help. Yes, I was very, very careful, and I am quite strong, but I can't say I wasn't worried.
Then the printer itself needed to be lifted so the packaging underneath could be removed. Through all this, the printer is much heavier than you might expect: the accessories and a set of 4x500g filaments are packed inside it!
That said, I don't suppose any 12"-class enclosed printer is much easier. I've thought about how packaging can be arranged so that it is easier to unpack, and I don't have any practical ideas. This is a one-off job, be prepared (unlike me) get some friends involved.
The next question is what to do with the packaging. Yes, it would be useful to keep in case the printer needs to be relocated at some point in the future, or even needs to be returned, but it's big! I'm a squirrel, but even I baulk at keeping a box that big. There seems to be no choice than to ditch it (although rather than take it all to the tip I will probably salvage the large cardboard panels as sheet material for future projects).
Commissioning My Snapmaker U1
Moderator: CrazyIvan
Re: Commissioning My Snapmaker U1
.
Assembly
Yes, considerable assembly is required. Even after removing all the packaging and transit tape, removal of three screws holding the bed down (not easy to get to) and another 6 securing a pair of guards (I think they held the internal boxes in place for transit) was required. And then the instructions mentioned some transit tape on the bed levelling screws underneath the bed, as an afterthought and only after the user would have run a test print! Not that it would necessarily matter, but why?!!
Assembly involves:
The instructions are detailed to the point of being ponderous, but there are several places you might trip up if you don't pay attention: fitting the spool holders, installing the print heads, and fitting the cables. Don't rush, and make sure you understand the instructions – particularly the orientation of the spool holders before you press them home (people say they take too much force, but I bet they've not offered them in the right orientation), and the screws for the cables (the head end must use the shorter screw, and not be done up tightly – it only needs to stop the cable falling out. Driving the screw in too far damages the print head assembly).
Installing the print heads can easily go wrong. It's easy once you understand what's going on, but easy to get wrong and the indicators should tell you it's wrong, but it might not be obvious why it's wrong. It does not help that the instructions talk about "inserting dowel rods" when actually you are manoeuvring the print head onto the dowel rod (translation issue, perhaps). The key thing to understand is that, having ensured the head swap mechanism is in the right place, the dowel pin needs to locate into the small hole not the large one.
Something I think would be an asset is a checklist in the Quick Start Guide, to tick off each stage as you've done it and ensure nothing has been skipped.
Assembly
Yes, considerable assembly is required. Even after removing all the packaging and transit tape, removal of three screws holding the bed down (not easy to get to) and another 6 securing a pair of guards (I think they held the internal boxes in place for transit) was required. And then the instructions mentioned some transit tape on the bed levelling screws underneath the bed, as an afterthought and only after the user would have run a test print! Not that it would necessarily matter, but why?!!
Assembly involves:
- Fitting the automatic feeders (2-off);
- Fitting the feed tubes (4mm) from the automatic feeders (4-off);
- Fitting the spool holders (4-off);
- Tensioning the drive belts;
- Fitting the poop tray;
- Installing the print heads (4-off);
- Fitting the cables from the chassis to the print heads (4-off, 16 screws);
- Fitting the feed tubes (6mm) and supports to the print heads (4-off);
- Fitting clips to secure the cables to the feed tubes (12-off)
The instructions are detailed to the point of being ponderous, but there are several places you might trip up if you don't pay attention: fitting the spool holders, installing the print heads, and fitting the cables. Don't rush, and make sure you understand the instructions – particularly the orientation of the spool holders before you press them home (people say they take too much force, but I bet they've not offered them in the right orientation), and the screws for the cables (the head end must use the shorter screw, and not be done up tightly – it only needs to stop the cable falling out. Driving the screw in too far damages the print head assembly).
Installing the print heads can easily go wrong. It's easy once you understand what's going on, but easy to get wrong and the indicators should tell you it's wrong, but it might not be obvious why it's wrong. It does not help that the instructions talk about "inserting dowel rods" when actually you are manoeuvring the print head onto the dowel rod (translation issue, perhaps). The key thing to understand is that, having ensured the head swap mechanism is in the right place, the dowel pin needs to locate into the small hole not the large one.
Something I think would be an asset is a checklist in the Quick Start Guide, to tick off each stage as you've done it and ensure nothing has been skipped.
Re: Commissioning My Snapmaker U1
.
First Switch-On
It is rather amusing to watch as the unit goes through its initial calibration cycle, and you have to be present to remove and then replace the PEI build plate on request (and be ready with the off switch if it all goes tits up). The unit senses Z=0 by nozzle contact with the bed, X & Y by contact of the carrier with the frame (in the front left corner), picks up each print head by dead-reckoning and checks Z for each nozzle, and calibrates X & Y offset for each nozzle by waggling the nozzle in a cup in the bed (exposed by manually removing the print plate).
The noisiest thing is the fan that blasts a jet of air across the chamber at nozzle height, and the calibration heats and cools the nozzles in this jet. It's pretty loud, but should be suppressed a bit when I get a top cover (coming soon!).
I skipped the networking and registration, and loading filaments for now, I wanted to see how this thing performs as-delivered. With the initial factory calibration done, the bed is in a high position (or you can raise it with the menus) to remove any transit tape from the levelling wheels (although I had raised the bed by hand before turning on, and there wasn't any tape). The instructions don't mention this until later, after the first trial print (when, actually, the bed has retreated into the bottom and would have to be raised).
First Switch-On
It is rather amusing to watch as the unit goes through its initial calibration cycle, and you have to be present to remove and then replace the PEI build plate on request (and be ready with the off switch if it all goes tits up). The unit senses Z=0 by nozzle contact with the bed, X & Y by contact of the carrier with the frame (in the front left corner), picks up each print head by dead-reckoning and checks Z for each nozzle, and calibrates X & Y offset for each nozzle by waggling the nozzle in a cup in the bed (exposed by manually removing the print plate).
The noisiest thing is the fan that blasts a jet of air across the chamber at nozzle height, and the calibration heats and cools the nozzles in this jet. It's pretty loud, but should be suppressed a bit when I get a top cover (coming soon!).
I skipped the networking and registration, and loading filaments for now, I wanted to see how this thing performs as-delivered. With the initial factory calibration done, the bed is in a high position (or you can raise it with the menus) to remove any transit tape from the levelling wheels (although I had raised the bed by hand before turning on, and there wasn't any tape). The instructions don't mention this until later, after the first trial print (when, actually, the bed has retreated into the bottom and would have to be raised).
Re: Commissioning My Snapmaker U1
.
First Print
Obviously I was gonna print the baby dragon that comes pre-loaded, and it's four-colour using the filaments supplied with the printer. I wasn't going to consult any instructions at this point, I expected it to work intuitively.
I didn't know which head to load each colour on, but having read a bit about the RFID tag reading on this printer I hoped it would sort itself out. I simply loaded the reels onto the spool holders in any order, and it was very satisfying when just poking the end of the filament into its respective auto-feeder port grabbed it and fed it all the way to a couple of inches from the head, and displayed the filament type and colour for each head on the control panel. TBH I view RFID as a bit of a gimmick, mainly because the filaments I'll be using won't be tagged and I expect to be able to declare filaments manually.
Nonetheless, it was easy and neat, and when I ran the print the correct filament got used for each element of the print and the dragon was coming out the right colour.
Something like 3 hours later, I had this (OK, I printed two – see later):
.
.
It's 60mm high. The block on the left is the prime tower (there were two of those too, one for each print), and is the equivalent of the pre-print wipe line on a single-filament print, so that each nozzle has somewhere to wipe at the current Z when it comes out of being parked. Amusingly, there is still a wipe line at the front of the plate in the first filament used!
It's a skewed example perhaps, with such a small model, but the model itself weighs 22g and the prime tower is 21g. The control panel predicts a total of 43.6g, which is well within the accuracy of my scales.
Granted this is a fine-tuned model file to showcase the capability of the printer, but look at the quality. The unsupported overhang under the dragon's chin is astounding:
.
.
Having done a first print without any snags, I wanted to check that wasn't a fluke and set off another, then went to bed confident I would see another dragon ready by the morning. It wasn't! About 2 hours into the print, there had been an error and the print was paused.
When the carrier picks up each head in turn, it makes a three-wire electrical connection to run the fan located in the carrier, I prersume including a tacho signal. The connection is made with spring-loaded "pogo" pins pressing onto pads on the carrier. These are gold-plated. If a fault is detected in the fan circuit, the controller takes that as meaning the head has not been picked up correctly (which could be catastrophic if it is badly misaligned) and stops with an error report ("swapper error"). Visual inspection indicated no problem, so I pressed play to resume the print, and it did.
Not long after, it did another one, but then competed the print successfully after that and I can't see the join (although the dragon model is designed to hide defects).
Is the fault detection giving up too early, with not enough retries? The pogo pins are expected to need cleaning and eventual replacement, but after only a few hours' printing?? I have yet to access the diagnostic logs, but I think both times this was associated with head 2... and when the filaments were loaded and purged there was some foreign filament in head 2 (it pooped red when I was loading white) whereas the other three heads didn't. Perhaps head 2 has some milage on it before it was dispatched to me!
I shall now update the firmware and have another go – perhaps the fault recovery system has been improved.
First Print
Obviously I was gonna print the baby dragon that comes pre-loaded, and it's four-colour using the filaments supplied with the printer. I wasn't going to consult any instructions at this point, I expected it to work intuitively.
I didn't know which head to load each colour on, but having read a bit about the RFID tag reading on this printer I hoped it would sort itself out. I simply loaded the reels onto the spool holders in any order, and it was very satisfying when just poking the end of the filament into its respective auto-feeder port grabbed it and fed it all the way to a couple of inches from the head, and displayed the filament type and colour for each head on the control panel. TBH I view RFID as a bit of a gimmick, mainly because the filaments I'll be using won't be tagged and I expect to be able to declare filaments manually.
Nonetheless, it was easy and neat, and when I ran the print the correct filament got used for each element of the print and the dragon was coming out the right colour.
Something like 3 hours later, I had this (OK, I printed two – see later):
.
- SnapDragon.jpeg (65.51 KiB) Viewed 220 times
It's 60mm high. The block on the left is the prime tower (there were two of those too, one for each print), and is the equivalent of the pre-print wipe line on a single-filament print, so that each nozzle has somewhere to wipe at the current Z when it comes out of being parked. Amusingly, there is still a wipe line at the front of the plate in the first filament used!
It's a skewed example perhaps, with such a small model, but the model itself weighs 22g and the prime tower is 21g. The control panel predicts a total of 43.6g, which is well within the accuracy of my scales.
Granted this is a fine-tuned model file to showcase the capability of the printer, but look at the quality. The unsupported overhang under the dragon's chin is astounding:
.
- SnapDragon Details.jpg (216.83 KiB) Viewed 220 times
Having done a first print without any snags, I wanted to check that wasn't a fluke and set off another, then went to bed confident I would see another dragon ready by the morning. It wasn't! About 2 hours into the print, there had been an error and the print was paused.
When the carrier picks up each head in turn, it makes a three-wire electrical connection to run the fan located in the carrier, I prersume including a tacho signal. The connection is made with spring-loaded "pogo" pins pressing onto pads on the carrier. These are gold-plated. If a fault is detected in the fan circuit, the controller takes that as meaning the head has not been picked up correctly (which could be catastrophic if it is badly misaligned) and stops with an error report ("swapper error"). Visual inspection indicated no problem, so I pressed play to resume the print, and it did.
Not long after, it did another one, but then competed the print successfully after that and I can't see the join (although the dragon model is designed to hide defects).
Is the fault detection giving up too early, with not enough retries? The pogo pins are expected to need cleaning and eventual replacement, but after only a few hours' printing?? I have yet to access the diagnostic logs, but I think both times this was associated with head 2... and when the filaments were loaded and purged there was some foreign filament in head 2 (it pooped red when I was loading white) whereas the other three heads didn't. Perhaps head 2 has some milage on it before it was dispatched to me!
I shall now update the firmware and have another go – perhaps the fault recovery system has been improved.
Re: Commissioning My Snapmaker U1
Ha! Updating the firmware didn't go well. Originally at 0.9.3, the latest is 1.5.1 so I installed that, only to find the dragon wouldn't print any more. There have been complaints about 1.5.0 and 1.5.1, so I downgraded to 1.4.1 and all was well.
The dragon printed again, without any swapper errors.
The dragon printed again, without any swapper errors.
Re: Commissioning My Snapmaker U1
That's not entirely true. There was an error message on the screen after the print completed, which seems to imply there was a swapper error (or more) but the unit recovered (perhaps the legacy of 1.4.1). Ditto for another run.
Apparently the problem I had with 1.5.1 was because it did not re-scan the print file, it only relied on the summary from a previous scan (under 0.9.3) and didn't like what it saw. Apparently if I had forced it to re-scan it would have been OK. Or so I'm told. I have no immediate plans to re-install 1.5.1, there are various complaints about it and I'm not sure I need anything it has to offer. Indeed, I'm more likely to install the Paxx12 custom firmware (currently based on 1.4.1).
Re: Commissioning My Snapmaker U1
The wipe line is the first thing printed. Then the elements of the first layer in that particular filament, and then it starts the prime tower.CrazyIvan wrote: Wed Jul 08, 2026 1:13 pm Amusingly, there is still a wipe line at the front of the plate in the first filament used!
Re: Commissioning My Snapmaker U1
Wow thanks for the detailed log of the Snapmaker U1.
You finally did it and unboxed it lol !
I have a new toy myself, not a 3D printer but a Thermal Camera (admittedly I got it free for test).
You finally did it and unboxed it lol !
I have a new toy myself, not a 3D printer but a Thermal Camera (admittedly I got it free for test).
Re: Commissioning My Snapmaker U1
Thermal cameras are interesting to use on FDM printers!
(1)
Re: Commissioning My Snapmaker U1
About 4 weeks in and... wow.
Only having used PLA before, PETG is a game-changer. Prints just feel so much more "professional". They are a little more heat resistant, sure (not much), but I designed a spool adapter so the U1's 55mm (internal diameter) spool holders will take filament spools with larger centre holes, and the prints (in PETG) look and feel like off-the-shelf plumbing parts. Very strong. PETG is cheap too, I got 4x 1kg from Amazon for £32.
.
.
Multi-material is a game-changer. A friend wanted spacer blocks printed to replace some (commercially moulded) aged and broken ones (secondary double-glazing fittings). These are small, and have tiny features (the slot gap is of the order 1mm), but by printing in PETG but using PLA as support, the support Z gap can be reduced to zero and the support just pulls away cleanly, because PETG does not stick to PLA.
In fact, the slicer (I'm using vanilla Orca rather than Snapmaker's version of Orca) offers to print the "support interface" in a different material (in my case PLA, but could be soluble filament), so the bulk of the support is printed in the main material (PETG) and only the places where support is in contact with the actual model are printed in PLA as a liner.
The actual part is too small to show up well in a photo, so here is the drawing:
.
.
All the advice is that PETG has to be dried before printing, but I've left my stocks out for a couple of weeks and not seen significant problems. Perhaps that's because I'm making functional parts rather than decorative ones, and don't care about a little stringing or the occasional surface blemish.
I had a big tussle with the "Toolhead Swapper Anomaly" reports. I wanted to know what exactly is going on, and carried out experiments (and spammed Snapmaker's support department) to find out, in the process reporting errors in their test file (which is a print file designed to just keep on swapping toolheads). I was printing dragons as a known test piece, and ended up with loads of them (some have found good homes, where they act as firelighters).
.
.
When the toolhead is picked up by the carrier, if an electrical contact fault is detected in the three "pogo" pins (which transfer power from the toolhead to the part fan built into the carrier), the U1 firmware parks the toolhead and tries again up to three times. This process can be seen in the console log (which can be monitored using the web control interface Fluidd), and it confused me that these retries were happening, and even giving up after three attempts, without triggering a Toolhead Swapper Anomaly report.
To cut a long story short, it turns out that even a failed sequence of retries does not comprise an anomaly unless the part fan was due to be running and the fan fails to reach the desired RPM. If the fan wasn't supposed to run, a pogo pin contact fault has no impact on the print, and can be ignored. If the fan was supposed to run but doesn't, version 0.9 firmware paused the print for user intervention, but version 1+ just logs it and carries on (which is better). There is a good chance that the temporary loss of the part fan (which might well work next swap) has insignificant effect on the final print. At the end of the print, a Toolhead Swapper Anomaly report gets displayed on the LCD.
I'm not entirely happy with the way this works (currently): it is possible to click through from the report to a stats screen showing how many anomalies were recorded for each toolhead during that print run, but the stats then reset for the next print, and are only accessible via the end-of-print message (if there is one). These figures need to be cumulative, and accessible through the menus. A particular toolhead accumulating anomaly counts requires cleaning or investigation, and if it's not specific to a particular toolhead it's the carrier that needs looking at.
Only having used PLA before, PETG is a game-changer. Prints just feel so much more "professional". They are a little more heat resistant, sure (not much), but I designed a spool adapter so the U1's 55mm (internal diameter) spool holders will take filament spools with larger centre holes, and the prints (in PETG) look and feel like off-the-shelf plumbing parts. Very strong. PETG is cheap too, I got 4x 1kg from Amazon for £32.
.
- U1SpoolAdapter.jpeg (99.59 KiB) Viewed 132 times
Multi-material is a game-changer. A friend wanted spacer blocks printed to replace some (commercially moulded) aged and broken ones (secondary double-glazing fittings). These are small, and have tiny features (the slot gap is of the order 1mm), but by printing in PETG but using PLA as support, the support Z gap can be reduced to zero and the support just pulls away cleanly, because PETG does not stick to PLA.
In fact, the slicer (I'm using vanilla Orca rather than Snapmaker's version of Orca) offers to print the "support interface" in a different material (in my case PLA, but could be soluble filament), so the bulk of the support is printed in the main material (PETG) and only the places where support is in contact with the actual model are printed in PLA as a liner.
The actual part is too small to show up well in a photo, so here is the drawing:
.
- SpacerBlock.jpeg (55.44 KiB) Viewed 132 times
All the advice is that PETG has to be dried before printing, but I've left my stocks out for a couple of weeks and not seen significant problems. Perhaps that's because I'm making functional parts rather than decorative ones, and don't care about a little stringing or the occasional surface blemish.
I had a big tussle with the "Toolhead Swapper Anomaly" reports. I wanted to know what exactly is going on, and carried out experiments (and spammed Snapmaker's support department) to find out, in the process reporting errors in their test file (which is a print file designed to just keep on swapping toolheads). I was printing dragons as a known test piece, and ended up with loads of them (some have found good homes, where they act as firelighters).
.
- Dragons!.jpeg (92.99 KiB) Viewed 132 times
When the toolhead is picked up by the carrier, if an electrical contact fault is detected in the three "pogo" pins (which transfer power from the toolhead to the part fan built into the carrier), the U1 firmware parks the toolhead and tries again up to three times. This process can be seen in the console log (which can be monitored using the web control interface Fluidd), and it confused me that these retries were happening, and even giving up after three attempts, without triggering a Toolhead Swapper Anomaly report.
To cut a long story short, it turns out that even a failed sequence of retries does not comprise an anomaly unless the part fan was due to be running and the fan fails to reach the desired RPM. If the fan wasn't supposed to run, a pogo pin contact fault has no impact on the print, and can be ignored. If the fan was supposed to run but doesn't, version 0.9 firmware paused the print for user intervention, but version 1+ just logs it and carries on (which is better). There is a good chance that the temporary loss of the part fan (which might well work next swap) has insignificant effect on the final print. At the end of the print, a Toolhead Swapper Anomaly report gets displayed on the LCD.
I'm not entirely happy with the way this works (currently): it is possible to click through from the report to a stats screen showing how many anomalies were recorded for each toolhead during that print run, but the stats then reset for the next print, and are only accessible via the end-of-print message (if there is one). These figures need to be cumulative, and accessible through the menus. A particular toolhead accumulating anomaly counts requires cleaning or investigation, and if it's not specific to a particular toolhead it's the carrier that needs looking at.