Hey everyone, welcome back for our July 2024 production updates.

AltMill

As we alluded to in the last update, we were hoping to start shipping AltMills mid-June. However, we did run into a few issues:

• There were some issues in the wiring for the spindles, which were done manually in-house to get some put together. We should have new ones arriving soon
• There were some parts on hold for coating due to some scheduling issues in the plant.

However, we are excited to announce we have started shipping AltMills! Horray!

You can also find resources for AltMill here: https://resources.sienci.com/view/am-welcome/

We’ve brought on several new people to help on the production and operations side of the AltMill so that we can ramp up shipping and iron out any kinks as they come up during our production.

Also… new gSender update is expected to come in the second week of July, which will contain settings and defaults for AltMill.

Our first batch of 200 is now sold out, and we are working on stocking up on our second batch. This means that new orders will be part of the second batch. The shipping schedule will also be largely determined by the timely arrival of those parts.

AltMill/LongMill Spindle Kit

I’m not sure where I should add some extra info about the AltMill Spindle Kit, with regards to LongMill compatibility but I’ll throw that in here for now.

We have gotten a lot of requests for the AltMill Spindle Kit to be available as a separate purchase, and we have planned to have it available since the beginning. It is now available in our store.

This also brought up a lot of interest in official spindle support for the LongMill. Some of the factors why we are working on supporting a spindle option include:

• Now that we have established a spindle testing and QC procedure for the AltMill spindles, we can use the same techniques for supporting LongMill as well
• We have created a relationship with an established spindle manufacturer who can build to our specifications
• The production of the AltMill provides enough volume for us to order more spindles, which brings the overall cost down.
• The addition of the SLB greatly improves and simplifies the installation and setup between the controller and the VFD. The SLB and the VFD used in this package allow for RS485 communication which allows for control of the spindle speed directly through gcode and the gSender interface, and can allow for features such as “wait for spindle”, which allows the spindle to get up to speed before starting the cut.
• This interface also allows for additional safety features such as being able to stop due to spindle issues and shut things down in the case of an emergency (with the SLB)
• With the improved motor holding capabilities of the SLB, the Z axis is better able to support the weight of the axis.

I should include that the spindle can be used with the original LongBoard, however there are some limitations

• Speed control through code or gSender may not be supported out of the box. It is possible to control it using the PWM, but a converter from 5V PWM to 0-10V analogue may be needed. The spindle can be controlled manually and turned on and off directly from the VFD, and so it can still be used this way.
• The added weight of the spindle may be enough to cause the Z axis to come down when no power is going to the motors. To mitigate this, users can use the command “$1=255”, which causes the drivers to hold their position. However, this constantly powers the motors and use a lot of power. There is no way to control how much power goes to the motor and each motor draws full current when this setting is turned on. The SLB allows for specific current values when the machine is stationary, so power draw can be controlled.

With that in mind, to take advantage of the new features, it is strongly recommended to use the spindle with the SLB.

We have ordered 200 spindle kits to finish the first batch of AltMills, plus an additional 200 kits that will be shared between AltMill sales and separate spindle kit sales. These are expected to arrive in mid-August.

At this stage, the spindle used for the AltMill has been tested on the LongMill and work is being done to have completed support documentation for the LongMill. Users can order the AltMill spindle kit for their LongMills. At this moment, we don’t officially support it, but we will officially support it once our testing is complete.

This also brings up the consideration of potential support for non-Sienci machines. If you have a machine that you want us to consider working with for the spindle kit, let us know.

Additionally, we’ve checked the fit for the 80mm dust shoe to be compatible between the AltMill and LongMill. We are also working on supporting 2.5in hoses (the 80mm dust shoe uses 4in at the moment).

LongMill MK2.5

LongMill MK2.5 machine are shipping now. We are working to get lead times down as we work our way through our list of orders, but at the time of writing, most are shipping within 3-4 weeks. Please refer to the order page information for most up-to-date estimates.

More shipments continue to roll in for LongMill production, and we aren’t expecting much news at this moment for production.

Vortex

Vortex orders continue to ship out. We’re excited to announce that we are working on the official version of the independent 4th axis, which combines and external driver with the SLB to allow for all axis to move simultaneously. For those who don’t know, the Vortex is currently connected to the Y-axis drivers so that you can switch between either the Y axis or the rotary axis. The SLB has an external driver output which allows another driver to be used to control another motor, in this case, the Vortex.

4th axis support is already officially supported and documented on the Resources. Users can integrate their own drivers for full 4th-axis support. The new development we are doing will be a plug-and-play option for the AltMill and SLB-integrated LongMills.

Given that we are purchasing a lot of closed-loop steppers, we have decided that the cheapest and easiest way to provide the 4th-axis support was to integrate them. While this might be little bit wasteful for existing users, since it leaves one unused motor in the conversion, in the long run as we move towards the SLB adoption, it will be the simplest and best option for this application.

SLB

A new batch of SLBs has arrived and folks who have ordered them after our first batch of 500 units should be getting them now.

A small change that’s coming to the SLB will be in the design of the e-stops. Our initial version had an illuminated switch. However, we found that it would make more sense to have lights on the case itself since it offered more button choices and we wanted to have something more durable/easier to replace. Both buttons work the same and are interchangeable so there likely won’t be any particular notice on when this change happens. We have another 500 of the original version in production now and we’re getting things together to make another 1000 sets.

Sprouter

I guess one of the things that the Sprouter project has become is an exercise in understanding spindles, routers, and machining science in general.

We have received another set of prototype motors and Johann has been testing that plus the Makita router and 1.5KW spindle. Here’s a breakdown:

• At the lowest level, the Makita has the highest torque, but drops proportionally to the speed (green)
• The spindle keeps a fairly flat torque curve, but needs to spin above 20,000RPM to see power advantages over the Makita (blue)
• The BLDC/Sprouter tested in various configurations show the potential to have a higher power output between the spindle and Makita router.

It also looks like the sensorless BLDC speed response was greatly improved in the new iteration. However, because the motor was built and tested to run at 160V (to be voltage compatible for both 110V and 220V), testing at 220V gives very good results, testing at 110V does not have an acceptable response time.

The manufacturer is currently working on a new version specifically tested and manufactured for 110V use only, which in theory should have similar performance as our current version, with the compatibility for 110V.

I should include some disclaimers here:

• While this shows the maximum power output, it may not reflect real-life optimal use. For example, running the Makita at it’s highest power level can destroy itself, whereas a spindle/sprouter is designed to run at the higher level for longer.
• Power output was calculated through indirectly with speed and torque simulation under real-life cutting loads, as our dyno cannot handle this level of power at this moment, it would be

Here are some current conclusions:

• The BLDC, after all the work, shows a lot of promise. These are results that we feel are “extraordinary”, which is that because they are so good, we have to make sure they are actually true. If the results are actually accurate, then in theory, it is possible to design a spindle using BLDC that could outperform a 1.5KW spindle.
• There is also a massive potential for this technology to be used in more applications outside just our machines. Perhaps it opens up a new roadmap for the company? We feel that after all of this work and development focused in this space, we may be one of, if not the only company doing this R&D, and perhaps the only one who is in the position to put out a new revolutionary (haha, bad pun), spindle design and platform.
• This also feels like uncharted territory, so there’s a lot of business end planning we need to do to decide how we want to keep committing to this project.

In other news, the testing of the Makita clone was not very good, as the speed control was poor. After some back and forth with the manufacturer, we are expecting to have a new sample with much-improved speed control soon. This version won’t have the 5V PWM input yet, but we should know if the performance is acceptable before moving onto this next step.

Panel Computer

We have now received the new batch of fanless panel computers and they are pretty great.

There are a few things we’re trying to decide on. One of them is which operating system to use. The two main contenders are Windows 11 and Linux. These are some of the things we are trying to weigh:

• Linux is free. Windows on the other hand, is pretty configurable, but does cost money. We aren’t exactly sure what that cost would be, as some sources offer it for $10-65USD, whereas the Microsoft store sells them directly for around $139USD. For us to be able to offer Windows as an option, we would need to navigate getting the licenses legally.
• Pretty much everyone either knows how to use Windows or has used it at some point. Linux on the other hand has a pretty small following. This may make some of the support for Linux more difficult, since there are a lot of info and resources available for Windows.
• Based on initial testing, the performance of gSender on the same hardware running Linux is slightly less fast than on Windows. We are still trying to iron this out, but we suspect that due to the differences in the drivers and optimization, going with Linux may mean there needs to be some additional optimization on gSender to make it compatible.

Here are my current thoughts.

• If we can get Windows licenses at a small cost, it would be a better option because we know that it works and people are familiar with it. We would of course have to pass this cost off to the customer, but I think people would be willing to pay a little extra for being able to use the platform they are familiar with.
• We can create images for both Windows and Linux, and they can be available for anyone to download and use. They should in theory work for both the panel computers we will sell and other computers as well, but we’ll only know once we do the testing. This means that if people want to switch between operating systems, or use their own computers, they can do that.

Currently, we are:

• Looking for a way to get legitimate Windows licenses
• Testing with Windows and Linux
• Testing and designing ways to attach the panel computer to the machine

If you want to provide any extra feedback for the Panel Computer, please feel free to fill out the survey.

Hey guys, a lot of things moving forward at Sienci Labs.

So traditionally I’ve been writing production updates for pretty much every product, but I’ve realized for some things there’s not much news with some things, especially if they’ve been shipping out smoothly for a while, so I’m going to stick to writing about new stuff. If you’re looking for an update on something and you don’t see it, check out the previous blog posts for more information. https://sienci.com/category/production-updates/

LongMill MK2/MK2.5

As we brought up in the last update, we are transitioning the LongMill to include a few new updates to the machine. Things happened a lot faster than we expected, resulting in the rest of our original LongMill (LongBoard) controllers being sold out and allocated to the previous batch. We are now into Batch 9 and are taking orders for LongMill MK2.5.

The LongMill MK2.5 Kits are expected to ship 6-8 weeks from the time of order. Once we start shipping the backlog, we expect to be able to get this lead time down.

Chris has written a new post “Introducing the LongMill MK2.5”, which covers everything you need to know about the differences, changes, and pricing for the updates.

We are waiting on more SLBs to arrive in the next few weeks, which will allow us to start shipping LongMill MK2.5 and the rest of the SLB backorders. This means that we have a growing waitlist of orders, which we are currently prepping so that we can get out the door as soon as all of our parts arrive. We are tentatively expecting them to arrive around the end of this week or start of next week.

We’re also hiring some extra staff to help with the packing for AltMill and LongMill to help with the extra production workload.

SuperLongBoard

We continue to wait for the second batch of SuperLongBoards to arrive. We are expecting them to arrive in the first week of June, which will allow us to clear the remaining backlog of orders and start shipping the LongMill. We will receive a batch of 1500 in total (but spread over a few weeks), which should allow us to stay stocked for some time.

Additionally, SLB-EXT for the first 50 AltMills testing has been successful. We are updating the value of one of the capacitors and soldering the first 50 boards by hand to fix some issues with the switching circuit. I look forward to the larger rollout of the SLB-EXT, as perhaps it may offer a solid offering for not just the AltMill but for other CNC machines as well.

AltMill

A lot of progress has been made with the AltMill project, and we are getting ready for shipping. As we discussed in the last update, we are expecting a few straggler parts for us to wait on before we can ship. It appears that our last item to arrive is the wiring harnesses for the closed-loop stepper motors. We expect these to arrive in the first week of June.

Based on the team’s estimates, we expect to start shipping on the second week of June, but wanted me to let everyone know to set their expectations not too high in case we run into issues.

Additional work and supply chain continue to happen to prepare parts to arrive in July/August for the full rollout of the AltMill.

Assembly for the AltMill is in full swing, as we have mostly finished building the jigs for assembly.

Some of the test packaging has arrived and we are double-checking the dimensions of everything and doing the final touches to the graphics for all of the packaging.

Also excited to share the first look at the AltMill Spindle Kits. We will have them available for sale additionally in a few weeks, as a lot of folks have asked us to allow the purchase of them outside of the AltMill kit. We likely won’t ship them until August, so there will be a bit of a waiting period for these if you’re ordering them without an AltMill.

In addition to manufacturing for the AltMill, we’re also working on building out the assembly portion of the AltMill, and we’ve already started putting together tons of resources to help users get their machines ready to go as soon as they arrive.

Sienci Router (Sienci Sprouter)

I think I mentioned this in the last update, but Johann went to China to meet with the motor manufacturers to iron out some of the things we were continuing to work on for this project.

Here’s a couple of takeaways:

• Based on our initial assumption, we believe that we are at the edge of how accurate sensorless control can be for controlling BLDCs. That being said, Johann was able to check on the progress of testing one of the motors using sensorless control. The result seems to be usable, although there is some more improvement that can be made, as the stability on 110V versus 220V is not acceptable yet.
• We were able to talk with a manufacturer that makes Makita clones, who may be able to make the modifications to a clone to offer external speed control, better quality bearings, and ER11 collets, which would offer a slightly improved alternative to the current Makitas
• Using a larger motor will improve the speed stability, but would cost more and be larger than the 65mm form factor we are aiming for.

There were a lot of things that we learned from Johann’s trip to China, especially about the manufacturing capacities and business structure and manufacturing processes of different companies.

If you guys watched the Garrett Fromme livestream, I talked a bit about the router vs spindle, and the name thing. So a lot of people like the name “Sprouter” so I think we might go with that for now…

From Johann’s trip we were also able to connect with another manufacturer who makes the Makita clones and we are also assessing if we should also have the clones as another option. We’d focus on:

• Having ER11 collets
• An input for PWM signal to control the router with the controller
• Better, higher quality bearings
• Pricepoint similar to the existing Makita router

So this sort of brings up another question for the development, which is whether it’s worth having an “in-between” option between an improved brushed AC Makita router and a 1.5KW spindle.

Here’s our general conclusions:

• Since we already have the 1.5KW spindle in the works for the AltMill, offering it as an option for the LongMill tackles most of the users who want to make a big upgrade to their existing LongMill. Since the programming and setup is mostly done and put together as a kit, the setup should be easy enough for most people.
• We found a company that can make the Makita clones with the extra features we’re looking for. Most of the development for this can be done hands off on our end so we can let this play out and test it once it’s ready.
• The BLDC option is still going through prototype iterations and development, and will likely take a lot more time than we expected. We’re pretty close to getting the performance we want out of it, so we’ll continue to work on it and see if we can get the pricepoint down further as well. Working on the first two options provides something for people in the short term that can help us understand more about the development for the Sprouter.

Panel Computer

If you watched the podcast that I was on with Garrett the other week, you might have seen me talk about the Panel Computer.

I think some people are going to recognize some similarities to the concept here between the Masso and this panel computer thing.

There’s a couple of things we’re working on now:

• Charles (our business development manager) has a lot of Linux experience, and is taking a stab at testing the device with a different OS.
• It turns out that the chip (i5 4th gen) that is on the test computer now is only compatible up to Windows 10. I don’t think this is specifically a problem, because it does still work, but I am looking at different chipsets that support future updates. Another contender is the N5095 or N100 chip, which is slightly faster and newer. It does cost a little bit more but is potentially a better option overall.
• We did some testing and research on doing the mass-preloading for the softwares, but need to do some more setup and testing to make sure we know how difficult and how long it takes to do.
• I’m looking at different monitor mount/arm designs to see if there’s any that stand out and improve the usability of the computer.
• The test computers we have have fans in them. Chris told me that he talked with some people working with a lot of wood that the main reasons their computers fail is because of the fans. So I am working on getting a fanless version for some testing as well.

At some point, Chris took one of the two computers I had for testing, so I suspect the dev team might be using it for some other stuff as well. I did hear that there will be a UI update in the future that makes improvements to the software that may improve the touchscreen experience as well.

I haven’t had the chance to install the computer directly to the machine yet, but one of our students from last term did set up an arm and portable touch-screen monitor to test the functionality of the interface and test out the “human factors” side of things.

It turns out with the LongMill MK2 and the AltMill, since there are threaded holes that are meant for the NEMA 23 motors on the front of the machine, this serves as a really good place to mount an arm.

Here’s some thoughts about the price point. Based on our BOM costs currently, I feel like the whole package is going to end up coming in at around $400USD. I think this is not cheap enough where I feel like it should be the default option for everyone, since a lot of people probably have a spare laptop or computer kicking around that they can use.

The only way we can get this pricepoint down further is probably if we can use lower-powered hardware, like something using ARM/RISC, but at this point, gSender optimization is not quite there for the smoothest experience. There is also updates with the new architectures that are improving its performance, especially with software that wasn’t made specifically for that architecture.

Another thing to look at is how much volume of computers we should get that would bring down the price. Right now I have quotes for 200 units, but at around 1000-1500 units, we’d save about 15% off the price.

I did create a short survey to see what sort of order interest there is for the panel computers. If you want to share your feedback, please check out the survey here: https://forms.gle/m2FAHcSBq6EJZ2sT7. This will help us get a better idea on how many computers we should start making.