I was hoping this would be a hack that repurposed the unit to work standalone.
Getting the unit to work again in the Cricut "ecosystem" just means Cricut can disable it later when they catch on.
Companies that "lock" or otherwise brick viable hardware as part of the business model (I'm thinking about Sonos here, too) are really terrible. People shouldn't be rewarding them by buying their garbage to begin with.
Would be a shame if anyone did that at a large scale, invoking a PR desaster and forcing some change.
Oh and illegal. Highly illegal and probably would get you in prison for a very long time + come with crippling debt too.
So don't do that, unless you're like.. idk terminally ill and die tomorrow or something like that. Then still don't do it because it's illegal and causes real harm to real other people that didn't do anything wrong.
If anyone has an old Silhouette vinyl cutter sitting around gathering dust, there's an actively maintained open source Inkscape plugin that lets you "print" your design straight to the cutter: https://github.com/fablabnbg/inkscape-silhouette
Thanks. I DO have an old Silhouette vinyl cutter sitting around... gathering dust! I picked it up at a thrift store many years ago and just waiting for you comment I guess!
I have a silhouette cameo pro 4 (and love it), and have been wondering around similar things. The Cameo can only be controlled via the provided software, or a couple proprietary plugins for popular design software.
The provided software is clunky, and the plugins suck. I’ve been building an app for designing stickers, and what I really want is to either be able to drive the Cameo from the app directly, or output a design file in a format that can be loaded straight into the software to start cutting.
The thing I’m stuck on is… how do people reverse engineer these devices protocols, without fear of bricking the device? My understanding is that reverse engineering like this is a lot of “try shit and see if it works”. It seems to me that a few wrong commands could get the cameo in some irrecoverable state. Obviously that’s not a risk I’m willing to take on a machine I use very regularly.
I mostly deal with the possibility of bricking a device by not caring too much :) That said, I usually start by reverse-engineering the software at one of the ends, so I don't blindly "fuzz" the device. There's still some risk, but not that much.
I don't have a Cameo Pro 4 to test, but I have reverse engineered software to control some of their other devices. It's a little janky but if you want to help I'd appreciate it!
I believe there is (or once was) also an Inkscape plugin somewhere but I never tried it.
As for how I reverse engineered it, I just used a USB sniffer and wireshark. I don't think there's a big risk of bricking these devices - I don't think they even have non-volatile storage, so worst case you turn it off and on again.
I would like to reverse engineer my solar power inverter too, but that is definitely brickable (the first one they shipped me came pre-bricked!) and is a lot more expensive so I've had to resist. :/
Just a PSA to anybody considering buying a new cricut:
They are cool machines mechanically, but the software is an absolute nightmare, DO NOT BUY.
Source: i bought one. Regret sunk in the moment i tried to do anything even slightly interesting. This isn’t a bambulabs situation where it’s mostly a philosophical objection to a closed ecosystem. It’s bad software that makes it hard or impossible to use the machine except in a very narrow set of ways.
I’ve not firsthand used any of the competing machines (silhoutte, et al) but i hear they are a bit better - and i don’t know how they could be worse.
Maybe it's just me and I haven't used it much, but my fairly non-technical wife has a Cricut and the few times I've wanted to use it I had absolutely no problem creating a vector in whatever program (Illustrator or Affinity Designer, IIRC) and getting it to separate the color layers and cut at the desired size.
The Electron app felt a bit weird, but I was able to accomplish the goal of cutting out my design at a desired size, so I didn't think any more about it. <shrug>
It seemed more to me that their app leans hard into getting people to buy vector designs through them, and making it easy for those who might not even understand what a "vector" graphic is, but it didn't seem to preclude me from cutting out my own stuff.
> but my fairly non-technical wife has a Cricut and the few times I've wanted to use it I had absolutely no problem
This is this is one of those situations of knowing their audience. The machine is nice and does some impressive things for the price. Their software works fine for the people that the hardware was designed.
I too have designed things in Illustrator and imported them into their software. It took a bit of getting used to, but I've designed retail packaging with cuts and scores so I can create custom boxes that I was not able to find premade and having the custom design made was too expensive. Could the design software be better? Of course, but that software would be out of reach for their target audience. So instead, I've stopped holding it wrong, and work within its constraints.
> Their software works fine for the people that the hardware was designed.
HARD disagree. Go look at all the youtube videos of craft people finding insane workarounds to try and get it to do things outside of the narrow lane of accepted uses.
One small example. It has a pen plotter mode. But it has no way to
1) generate single-line text (like a Hershey font) - the only text mode is outlines.
Well, you need a font that is single-line by design. Then it works just fine. There are some single-line default fonts, too. For filling shapes, there is no automated cross-hatch or anything, that's true, but you could create such a pattern in a vector program.
Siser machines are as consumer as you can get but still get control over them. They are more like the commercial grade ones. I have a Juliet and love it.
It is, but I mean: Many of us do just exactly that with 3D printers.
We get a new 3D-printing machine and install our own build of Marlin on it, or move it all over to Klipper. Sometimes that involves replacing the entirety of the brain-box controls and that's just par for the course.
At its root, it's only some stepper motors, fans (more motors), sensors, and heaters that are used in very particular ways.
That's pretty similar to a Cricut machine, isn't it?
Because closed source 3d printers grew out of an open source community, it's pretty easy to do this, and you get a pretty similar experience on most printers. Unfortunately, cricuts are not like that. There are a couple projects attempting this but afaik they offer a pretty rough experience.
Nice! I've had very good luck with most (not all!) of the Biqu boards I've used, at least in printing space. I've also had great (100%) luck with warranty claims.
If you're really launching into this thing, then: I came across a set of ifixit sorta-teardowns[1] on a Cricut machine earlier today. It showed what appeared to be relatively-small 4-wire motors (often with gear trains shown, which does matter), which strongly suggests that 2-pole steppers are being used. That's easy enough.
The TMC5160 on the Biqu Rodent seems like a strange choice for this job: The 5160 is a high-current dude that's meant more for grunt than precision. The availability of grunt doesn't necessarily outlaw the ability to be precise, but it may not necessarily represent an ideal set of tradeoffs.
Perhaps a board with smaller stepper drivers may be better-suited. After all, if the motors are small and they've got gear reduction on their outputs, then: They don't need much grunt to make the bits move move in predictable ways.
The photos also showed that some/many/all of these motors have optical encoder wheels. Perhaps they're optional in a build with a custom brainbox (we do usually get along fine without that kind of feedback in printing), but their presence is certainly interesting.
And: There were nowhere near enough connector pins shown on what was described as the motherboard for all of this stuff to happen. That's concerning.
It all combined to strongly suggest to me at that time that the ICs that drive the motors might be located elsewhere in the machine, and that there may also be some servo feedback that happens in stock form.
---
That concludes my brain-dump on what I think I know about Cricut machines. Take it as you will. :)
And: The FluidNC Board Mapping Wizard is an amazing tool. I will save this for future use. Thanks!
The modern equivalent is "Just" use Fable or Sol to reverse-engineer the protocol, then write new control software that works locally the way you want to work, using the file formats you want to use.
There is no longer any reason to put up with hardware like Cricut that doesn't do what you want, or that includes artificial dependencies on remote servers. If worst comes to worst, then yes, have the clanker write a new controller.
One of the YouTubers who benchmarks vibe-hackery tried getting GLM 5.3 to reprogram an older iPod Nano. [1] Something like "I just plugged in an iPod via USB. Turn it into a GPU status display." It got pretty far but couldn't close escrow.
Possibly. But i wanted a machine to play with, not to tinker with. At that point it’d be wayyy cheaper and easier to just convert an old CR-10 printer into a vinyl cutter
silhouette has a much better open source reputation if you want this type of machine. Everybody in the crafting world seems to have the circut though, and you can buy their machines (parts, supplies) all over, while I've never seen silhouette in a local store.
I have a silhouette cameo 4 I'm pretty happy with. Silhouette studio works, it doesn't handle svgs very well so when importing I tend to render svgs as a black bitmap and then trace that in silhouette studio. Being able to print registration marks and have the cutter align itself against those is invaluable. I've done clones of paper craft kits, perfectly cut out iron on transfers forct shirts, 3d trading card art, and trading card proxies on mine
There's also Inkcut, which is the only thing that my old Silhouette will consistently respond to, but it works in Linux and it couldn't be easier to drop in an SVG and have it cut (which is my only use case).
The last time I checked, bambu was clawing back functionality from purchased devices, with things like timebomb lockouts and blacklisting TPU.
Sure, it's a matter of degree, but I'm perfectly happy with my Elegoo. If you live in California, I suggest buying one before Sacramento's open firmware ban goes into effect.
I have a bambulab printer and haven't seen anything like this. The printer doesn't even know what filament you are using unless you are using an AMS + bambulab filament. Maybe you are thinking about the fact that TPU isn't compatible with the AMS, you have to feed it in direct or it will clog.
I can load a print direct on to the printer via lan or the microsd slot without using any bambulab servers. While in theory it would be nice if it was all fully open source, but in reality there are no real restrictions unlike Cricut.
It's a vinyl cutter. I've used those. TechShop used to have them.
Theirs seem to be repurposed large-format HP plotter designs.
Vinyl cutters are good at doing one thing - cutting out patterns in vinyl or similar materials. You can make signs, stickers, and similar stuff. Some also be used as plotters. They're rather limited devices, though. At TechShop, they were not used much.
Originally, people put knife blades into HP plotters, by modifying the pen holders. Then Roland started making heavy-duty plotters purpose-built to run blades, but they spoke the same HPGL as the plotters, for compatibility.
Then everyone and their brother knocked-off the Roland designs, and of course copied the HPGL compatibility. Along the way, some special extensions crept into the dialect, for setting the downforce (the !FS; command) and maybe other stuff.
And for a long, long time, all vinyl cutters were basically the same. You could use any software with any cutter.
“Provo Craft contended that in order to understand and replicate this obscured protocol, Craft Edge had disassembled the DesignStudio program, contrary to the terms of its end-user license agreement, thereby (the company asserted) breaching copyright law.”
I wonder if the wireshark USB intercepts discussed in this thread would classify as ‘disassembly’?
That's why you don't use a Cricut to make furry stickers. It's already miserable to use a Cricut to cut them because of the fight to get proper alignment and cutting because the machine has no way to properly home materials to the cutting position.
You use a Brother Scan-n-Cut instead, since it can actually perform registration on the material you've printed out.
I was hoping this would be a hack that repurposed the unit to work standalone.
Getting the unit to work again in the Cricut "ecosystem" just means Cricut can disable it later when they catch on.
Companies that "lock" or otherwise brick viable hardware as part of the business model (I'm thinking about Sonos here, too) are really terrible. People shouldn't be rewarding them by buying their garbage to begin with.
I'm just imagining the amount of damage some determined people can do to the Cricut ecosystem because the serial numbers are guessable...
Would be a shame if anyone did that at a large scale, invoking a PR desaster and forcing some change.
Oh and illegal. Highly illegal and probably would get you in prison for a very long time + come with crippling debt too.
So don't do that, unless you're like.. idk terminally ill and die tomorrow or something like that. Then still don't do it because it's illegal and causes real harm to real other people that didn't do anything wrong.
Channel your inner Maia arson crimew
Or maybe if you live in one of the places where illegal telemarketing calls originate from but can't be stopped.
I’d expect a certain YouTuber to pick it up in due course
If anyone has an old Silhouette vinyl cutter sitting around gathering dust, there's an actively maintained open source Inkscape plugin that lets you "print" your design straight to the cutter: https://github.com/fablabnbg/inkscape-silhouette
I wrote about it a couple years ago: https://blog.bityard.net/articles/2022/June/diy-vinyl-cut-mo...
Nice write-up! I was also a Suzuki fan (1980s) and always liked the stylized S logo. Now to go find a vinyl cutter.....
Thanks. I DO have an old Silhouette vinyl cutter sitting around... gathering dust! I picked it up at a thrift store many years ago and just waiting for you comment I guess!
I have a silhouette cameo pro 4 (and love it), and have been wondering around similar things. The Cameo can only be controlled via the provided software, or a couple proprietary plugins for popular design software.
The provided software is clunky, and the plugins suck. I’ve been building an app for designing stickers, and what I really want is to either be able to drive the Cameo from the app directly, or output a design file in a format that can be loaded straight into the software to start cutting.
The thing I’m stuck on is… how do people reverse engineer these devices protocols, without fear of bricking the device? My understanding is that reverse engineering like this is a lot of “try shit and see if it works”. It seems to me that a few wrong commands could get the cameo in some irrecoverable state. Obviously that’s not a risk I’m willing to take on a machine I use very regularly.
I mostly deal with the possibility of bricking a device by not caring too much :) That said, I usually start by reverse-engineering the software at one of the ends, so I don't blindly "fuzz" the device. There's still some risk, but not that much.
I don't have a Cameo Pro 4 to test, but I have reverse engineered software to control some of their other devices. It's a little janky but if you want to help I'd appreciate it!
https://robocut.org/
I believe there is (or once was) also an Inkscape plugin somewhere but I never tried it.
As for how I reverse engineered it, I just used a USB sniffer and wireshark. I don't think there's a big risk of bricking these devices - I don't think they even have non-volatile storage, so worst case you turn it off and on again.
I would like to reverse engineer my solar power inverter too, but that is definitely brickable (the first one they shipped me came pre-bricked!) and is a lot more expensive so I've had to resist. :/
Just a PSA to anybody considering buying a new cricut:
They are cool machines mechanically, but the software is an absolute nightmare, DO NOT BUY.
Source: i bought one. Regret sunk in the moment i tried to do anything even slightly interesting. This isn’t a bambulabs situation where it’s mostly a philosophical objection to a closed ecosystem. It’s bad software that makes it hard or impossible to use the machine except in a very narrow set of ways.
I’ve not firsthand used any of the competing machines (silhoutte, et al) but i hear they are a bit better - and i don’t know how they could be worse.
Maybe it's just me and I haven't used it much, but my fairly non-technical wife has a Cricut and the few times I've wanted to use it I had absolutely no problem creating a vector in whatever program (Illustrator or Affinity Designer, IIRC) and getting it to separate the color layers and cut at the desired size.
The Electron app felt a bit weird, but I was able to accomplish the goal of cutting out my design at a desired size, so I didn't think any more about it. <shrug>
It seemed more to me that their app leans hard into getting people to buy vector designs through them, and making it easy for those who might not even understand what a "vector" graphic is, but it didn't seem to preclude me from cutting out my own stuff.
> but my fairly non-technical wife has a Cricut and the few times I've wanted to use it I had absolutely no problem
This is this is one of those situations of knowing their audience. The machine is nice and does some impressive things for the price. Their software works fine for the people that the hardware was designed.
I too have designed things in Illustrator and imported them into their software. It took a bit of getting used to, but I've designed retail packaging with cuts and scores so I can create custom boxes that I was not able to find premade and having the custom design made was too expensive. Could the design software be better? Of course, but that software would be out of reach for their target audience. So instead, I've stopped holding it wrong, and work within its constraints.
> Their software works fine for the people that the hardware was designed.
HARD disagree. Go look at all the youtube videos of craft people finding insane workarounds to try and get it to do things outside of the narrow lane of accepted uses.
One small example. It has a pen plotter mode. But it has no way to
1) generate single-line text (like a Hershey font) - the only text mode is outlines.
2) fill outlined shapes.
Well, you need a font that is single-line by design. Then it works just fine. There are some single-line default fonts, too. For filling shapes, there is no automated cross-hatch or anything, that's true, but you could create such a pattern in a vector program.
As I said:
> impossible to use the machine except in a very narrow set of ways.
Congrats, you were using it within that narrow set of ways.
Siser machines are as consumer as you can get but still get control over them. They are more like the commercial grade ones. I have a Juliet and love it.
Could you just hook new motor controllers to the motors / puncher and add an Arduino to control it?
This is a very Hacker News usage of the word "just".
It is, but I mean: Many of us do just exactly that with 3D printers.
We get a new 3D-printing machine and install our own build of Marlin on it, or move it all over to Klipper. Sometimes that involves replacing the entirety of the brain-box controls and that's just par for the course.
At its root, it's only some stepper motors, fans (more motors), sensors, and heaters that are used in very particular ways.
That's pretty similar to a Cricut machine, isn't it?
Because closed source 3d printers grew out of an open source community, it's pretty easy to do this, and you get a pretty similar experience on most printers. Unfortunately, cricuts are not like that. There are a couple projects attempting this but afaik they offer a pretty rough experience.
So a Cricut machine has more than some motors and sensors and stuff?
How much more?
Replacing the hardware with GRBL/FluidNC controller + bCNC + Inkscape migtyscape laser/engraver CAM plugin kit would be trivial.
However, modifying DMCA protected hardware is a crime in some places.
Best of luck, =3
Eh, I mean: I've done some work with GRBL in the pre-bot days. I can probably do more work with it in these post-bot days. That doesn't sound so bad.
I just don't have a Cricut or the willingness to get one (or several; stuff will break) to hack on. :)
The FluidNC ESP32 Rodent Motherboards seem a bit more modern. May grab one unless I find something better over the weekend.
https://biqu.equipment/products/bigtreetech-rodent
The firmware configuration tools are now apparently trivial for users:
https://mitchbradley.github.io/FluidNC-config-wizard/?board=...
bCNC has surface probing so is probably also better than Circut for blade wear... good use for an old laptop or Pi =3
Nice! I've had very good luck with most (not all!) of the Biqu boards I've used, at least in printing space. I've also had great (100%) luck with warranty claims.
If you're really launching into this thing, then: I came across a set of ifixit sorta-teardowns[1] on a Cricut machine earlier today. It showed what appeared to be relatively-small 4-wire motors (often with gear trains shown, which does matter), which strongly suggests that 2-pole steppers are being used. That's easy enough.
The TMC5160 on the Biqu Rodent seems like a strange choice for this job: The 5160 is a high-current dude that's meant more for grunt than precision. The availability of grunt doesn't necessarily outlaw the ability to be precise, but it may not necessarily represent an ideal set of tradeoffs.
Perhaps a board with smaller stepper drivers may be better-suited. After all, if the motors are small and they've got gear reduction on their outputs, then: They don't need much grunt to make the bits move move in predictable ways.
The photos also showed that some/many/all of these motors have optical encoder wheels. Perhaps they're optional in a build with a custom brainbox (we do usually get along fine without that kind of feedback in printing), but their presence is certainly interesting.
And: There were nowhere near enough connector pins shown on what was described as the motherboard for all of this stuff to happen. That's concerning.
It all combined to strongly suggest to me at that time that the ICs that drive the motors might be located elsewhere in the machine, and that there may also be some servo feedback that happens in stock form.
---
That concludes my brain-dump on what I think I know about Cricut machines. Take it as you will. :)
And: The FluidNC Board Mapping Wizard is an amazing tool. I will save this for future use. Thanks!
[1]: https://www.ifixit.com/Device/Cricut_Maker
The motor in the photo says HC385MG, it's a brushed DC motor with an attached encoder wheel for servo control.
Only if done over a weekend
The modern equivalent is "Just" use Fable or Sol to reverse-engineer the protocol, then write new control software that works locally the way you want to work, using the file formats you want to use.
There is no longer any reason to put up with hardware like Cricut that doesn't do what you want, or that includes artificial dependencies on remote servers. If worst comes to worst, then yes, have the clanker write a new controller.
Proposing cost to reverse engineer a cricut and cut out some pelican art as a new llm benchmark.
As long as there's a bicycle under that pelican I think we're good.
One of the YouTubers who benchmarks vibe-hackery tried getting GLM 5.3 to reprogram an older iPod Nano. [1] Something like "I just plugged in an iPod via USB. Turn it into a GPU status display." It got pretty far but couldn't close escrow.
1: https://www.youtube.com/watch?v=3TeW8L9wy-Y&t=1917s
It it less crazy that it sounds. I have recently replaced firmware on an obscure MCU to revive a flip-dot display.
https://www.bobek.cz/buse/
Possibly. But i wanted a machine to play with, not to tinker with. At that point it’d be wayyy cheaper and easier to just convert an old CR-10 printer into a vinyl cutter
Can you build a CNC machine out of nothing but a computer and a CNC machine?
You wanna know how to get a million dollars? Ok, first you start with two million dollars...
silhouette has a much better open source reputation if you want this type of machine. Everybody in the crafting world seems to have the circut though, and you can buy their machines (parts, supplies) all over, while I've never seen silhouette in a local store.
I have a silhouette cameo 4 I'm pretty happy with. Silhouette studio works, it doesn't handle svgs very well so when importing I tend to render svgs as a black bitmap and then trace that in silhouette studio. Being able to print registration marks and have the cutter align itself against those is invaluable. I've done clones of paper craft kits, perfectly cut out iron on transfers forct shirts, 3d trading card art, and trading card proxies on mine
yep, today's reminder silhouette is (quietly) a division of the larger Graphtec, and you can move up from the baby machines thru to a Graphtec cutter!
drive 'em with inkscape, drive 'em with the silhouette studio, hell, there's a bridge graphtec model that can also take silhouette studio jobs...
There's also Inkcut, which is the only thing that my old Silhouette will consistently respond to, but it works in Linux and it couldn't be easier to drop in an SVG and have it cut (which is my only use case).
The last time I checked, bambu was clawing back functionality from purchased devices, with things like timebomb lockouts and blacklisting TPU.
Sure, it's a matter of degree, but I'm perfectly happy with my Elegoo. If you live in California, I suggest buying one before Sacramento's open firmware ban goes into effect.
I have a bambulab printer and haven't seen anything like this. The printer doesn't even know what filament you are using unless you are using an AMS + bambulab filament. Maybe you are thinking about the fact that TPU isn't compatible with the AMS, you have to feed it in direct or it will clog.
I can load a print direct on to the printer via lan or the microsd slot without using any bambulab servers. While in theory it would be nice if it was all fully open source, but in reality there are no real restrictions unlike Cricut.
It's a vinyl cutter. I've used those. TechShop used to have them. Theirs seem to be repurposed large-format HP plotter designs.
Vinyl cutters are good at doing one thing - cutting out patterns in vinyl or similar materials. You can make signs, stickers, and similar stuff. Some also be used as plotters. They're rather limited devices, though. At TechShop, they were not used much.
Yup, they even speak HPGL.
Originally, people put knife blades into HP plotters, by modifying the pen holders. Then Roland started making heavy-duty plotters purpose-built to run blades, but they spoke the same HPGL as the plotters, for compatibility.
Then everyone and their brother knocked-off the Roland designs, and of course copied the HPGL compatibility. Along the way, some special extensions crept into the dialect, for setting the downforce (the !FS; command) and maybe other stuff.
And for a long, long time, all vinyl cutters were basically the same. You could use any software with any cutter.
Then Cricut came along and hosed it all up.
Try doing a run of like 10x of the same design in cricut software...
https://en.wikipedia.org/wiki/Cricut#Controversies
Also, it's a very fun article to read.
“Provo Craft contended that in order to understand and replicate this obscured protocol, Craft Edge had disassembled the DesignStudio program, contrary to the terms of its end-user license agreement, thereby (the company asserted) breaching copyright law.”
I wonder if the wireshark USB intercepts discussed in this thread would classify as ‘disassembly’?
I am very surprised that the ceo backpaddled on the limit of cloud usage that was going to force users into subscription...
I see so many of these at resale stores for next to nothing. What a shame.
He got it working, so there’s hope.
I guess if one needs furry decals badly enough you could just cut them out by hand
It would be pretty miserable when typically you want like 100+ stickers to hand out at events.
That's why you don't use a Cricut to make furry stickers. It's already miserable to use a Cricut to cut them because of the fight to get proper alignment and cutting because the machine has no way to properly home materials to the cutting position.
You use a Brother Scan-n-Cut instead, since it can actually perform registration on the material you've printed out.
No checksumming on the serial number packet makes this kind of proxy trivial. Using an RP2040 as a USB man-in-the-middle is a clever and cheap fix.