Maslow 4 - The next generation of Maslow

[bar]

I’ve been working on the next version of Maslow for a while now and hinting at how it works, but I want to come out of the shadows a bit more in light of the Cubiio kickstarter that is very similar.

I’m confident that I’ve disclosed all the key bits publicly enough times already that they are securely in the public domain and will remain open source, but I haven’t shown them off all together.

The next iteration of Maslow is called Maslow 4 for the obvious reason that it uses four belts, but also because it’s the 4th (depending on how you count) version. The first era had the two solid anchor points, the second had the ring, the third era was when other companies took over producing kits and sold their own versions, and then this is the fourth.

I plan to launch a kickstarter soon-ish with the goal of selling a kit for roughly the original $350 price tag…although with inflation that might not be an option and they might be more like $400. We’ll see how cheaply I can do it.

Move

Maslow 4 moves all of the motors and electronics onto the sled and replaces the chains with belts for easy storage.

Each of the four belts connects to one of four belt drive units, each of which is free to rotate around the router.

Having the belt run directly to the router provides a very very rigid connection because the force is transmitted directly.

Pivot Demo

Each belt drive unit has a gear reduced motor which drives a spool holding the belt, and a pair of pinch rollers with an encoder to measure the belt length precisely.

The controller board also incorporates current feedback allowing the controller to control and measure the length of the belt (using the encoder), and also the tension of the belt (using the current feedback).

The frame holds no electronics, only provides anchoring points for the other end of the belts.

This setup improves over the last version in that it is faster, more compact (ie because the frame is just any flat surface you can use a wall or the floor), and more precise (especially close to the edges of the sheet).

Faster Move

I’ve been trying to quit my job to get back to focusing on Maslow for a few months now, but when I saw the Cubiio project I let work know that for real this is my last week so I should be working on it full time again ASAP.

[RoBotX]

I’m sold. When can I order one?

[TimS]

Very impressive. Is your plan to ship all the 3D printed parts or just the hardware and files to print your own? Assembly looks like a pain. Would it still be open source or are you going to patent it and start a business? Sorry about you having to quit your current job. Put me down for an order as well.

[bar]

This is very much a prototype, I haven’t been showing it off because I was hoping to show off a bit more polished version. The final version will be injection molded. It will be open source so anyone who wants could print their own, but injection molded plastic can be so much stronger.

The assembly isn’t too bad. It’s about as tough as putting together a Prusa 3D printer. I can put one together in about an hour at this point, but I’ve done it a lot of times

[digitydogs]

Sign me up! Also expect to see a large number of Maslow 1-3’s flood the for sale marketplaces shortly after launch

[Orob]

Bar,

I have a ton of specific questions about it. Here are the first couple: Is it running the esp32 board? What software are you using to send to it? When and where can I learn more?

Thanks for sharing!

[jonatpridesleap]

Take. My. Money.

[bar]

It is running on the ESP32 board. The interface is web based, it will either connect to your wifi network, or if you don’t have wifi where you want to use the machine it creates an access point that you can connect to. USB and Bluetooth are also supported, although I haven’t messed with those much. The software is based on FluidNC which is a fork from GRBL with a custom (in progress) UI.

In terms of where you can learn more…here! I haven’t done a nice documentation write up. The files are all public, but I haven’t organized them in any way that would be intelligible to anyone else

[mooselake]

@bar

It is running on the ESP32 board.

That answered my first question

With current feedback you can do homing, you can tell when the infeeding belt stalls, and when you need to let out more belt on the other 3 drives.

I’m guessing at that price there’s no need to have an upgrade kit from the original version. Sign me up, too!

Kirk

[idougal]

Sounds like a great Idea. I’ll support it too!

[pdw125]

Looks great and a definite interest - is it going to suit a full 8x4 sheet or is the idea to be closer to 6x3…? And how do you plan to keep dust out of the mechanism…?

UK availability would be great too - most of the other vendors want a premium to ship outside the US.

[hdbham]

Looks amazing! Sign me up

[md8n]

Are the belts reinforced in some way to mitigate stretching? (It kinda looks like that from the closeups, but I’m not sure.

Also, if you want someone to help review the code, give me a ping.

And lastly - sign me up for one as well

[md8n]

Also, any thoughts on supporting a proper spindle instead of “go buy your own router”?

Router choice here is very poor, and controlling them adequately is entirely dependent on the specific router. So being able to order a ‘good’ spindle (something with proper speed control, overload detection)
as a part of the kit would be super beneficial

[Alex_Caminade]

Hi there, I discovered the Malsow CNC Project just a few days ago ! for quite some time I’ve been looking for an affordable way to build or buy a CNC…
I think i found what i was looking for just here ! :
Open Source !
Fonctionnal !
Modifiable !
Seamingly quite precise !
Cheap !
Easy to build with not much more than a hand saw and a screwdriver… !

I havent been testing it yet so i don’t know much about the precision issues, and all the things that make it “not perfect”.

Nevertheless, I’m used to building custom tools and developping (or adapting) mechanical systems.

It feels to me that this new version is going to be great ! I love the engineering
More Precise on the sides, quicker, easier to store, etc !

But it also feels to me that it’s also loosing a bit of the very essence of the project in the way that it seems more complex and less adaptable to everyone’s needs.
For example, how can i use a different motor than the one you are using if all the parts are injection moulded. or in an other way, how can i make a custom set of “pignons” if i don’t have access to a 3D printer ?

Sorry if my comment is a bit of a bug, i know way too well how much time, effort, thinking, testing, re-thinking… and how much commitment is involved in such a project. I respect all that !
Sometimes, i hate when people get in to my projects, but sometimes, it helps to think “out of the box” !

[Alex_Caminade]

Oh and just an other question… does this version 4 still have to be tilted or can it be flat ?

Thanks very much !

[bar]

@mooselake

With current feedback you can do homing, you can tell when the infeeding belt stalls, and when you need to let out more belt on the other 3 drives.

Exactly . I’m doing automatic homing and stall detection, but the actual belt lengths are set using the position feedback system from the encoders. Stall detection is just there to shut things down if something goes wrong.

@pdw125

is it going to suit a full 8x4 sheet or is the idea to be closer to 6x3

The goal is to be able to cut the full 4x8. Because the Maslow work area isn’t really a rectangle it’s hard to put a number on the work area size, but the goal is to be able to cut a sheet of plywood in half both hotdog and hamburger style as well as diagonally. I have not actually done that yet, although I have tested the performance near the edges of the sheet and it is dramatically superior to the two chain version. Right now the belt lengths are just computed from the target position, but I would like to add some adaptive behavior to put less tension on the lower belts when cutting high on the sheet and more tension on the lower belts when cutting lower on the sheet which I think will improve performance there even more.

@pdw125

UK availability would be great too - most of the other vendors want a premium to ship outside the US.

UK availability will absolutely be a thing. Having an international community is part of the fun! We won’t charge anything extra to ship internationally above what shipping + tariffs cost us.

@md8n

Are the belts reinforced in some way to mitigate stretching?

Yes, they are Kevlar reinforced. I found the steel reinforced ones still stretched too much

@md8n

Also, any thoughts on supporting a proper spindle instead of “go buy your own router”?

It’s designed pretty much around the Dewalt D611 which I’ve found to be the best compact router out there. The Makita and it’s cheap clones run really hot (like the spindle gets too hot to touch under no load, and that heat transfers into the router bit causing it to wear out more quickly). The D611 is also pretty widely available. A proper spindle would be nice, and I might put together an adapter kit for one if there is enough interest, but I’ve heard pretty mixed reviews about the spindles that are under about $500. Dewalt has done a pretty amazing job of building a good router for the price that is tough to beat…but I’m open to looking at alternatives.

@Alex_Caminade

But it also feels to me that it’s also loosing a bit of the very essence of the project in the way that it seems more complex and less adaptable to everyone’s needs.
For example, how can i use a different motor than the one you are using if all the parts are injection moulded. or in an other way, how can i make a custom set of “pignons” if i don’t have access to a 3D printer ?

That’s a great point. It will always be open source and I plan to sell all the bits independently so you could do a mix and match of just buying the parts that you want for a custom version. That being said, my main goal is to make big CNC routing accessible to as many people as possible, so I think that there are a lot of folks out there who are willing to trade some customization for something which is faster, easier to use, more reliable.

@Alex_Caminade

Oh and just an other question… does this version 4 still have to be tilted or can it be flat ?

Great question. It can be flat or nearly vertical.

[MartinJ]

<< TODO: insert shut up and take my money gif >>

When can we expect a kickstarter to go live? I’m not talking about when orders might get fulfilled. I’m sure you’re going to need seed money to use for design and initial production (not to mention maybe something to survive on since you lost your primary income).

[bar]

@MartinJ

When can we expect a kickstarter to go live? I’m not talking about when orders might get fulfilled. I’m sure you’re going to need seed money to use for design and initial production

I’d like to let the Cubiio folks finish their kickstarter before I launch one so I don’t steal people from their campaign. I’m thinking maybe 6 or 8 weeks?

For anyone interested, here’s the new control board. It’s a 5 axis controller (4 servo axis with current feedback + 1 stepper motor axis). It’s got wifi and bluetooth, the fancy trinamic stepper driver, 4 gigs of integrated storage (for files), and a break out serial connection for future expansion.

[TimS]

Looks nice. Will there be aux pins for z zeroing, router/spindle on/off, or spindle speed control?